Children with externalizing
Transcription
Children with externalizing
Children with externalizing behavior problems Risk factors and preventive efforts Jill Thijssen Children with externalizing behavior problems Risk factors and preventive efforts Jill Thijssen ISBN: Cover design: Printing: 978-94-6299-317-4 Guyon Muijres & Jill Thijssen Ridderprint BV - www.ridderprint.nl © Jill Thijssen, Maastricht, 2016 All rights are reserved. No part of this book may be reproduced or transmitted in any form or by any means, without written permission from the author or, when appropriate, the publisher of the article. Children with externalizing behavior problems Risk factors and preventive efforts Proefschrift ter verkrijging van de graad van doctor aan de Universiteit Maastricht, op gezag van de Rector Magnificus, Prof. dr. L.L.G. Soete volgens het besluit van het College van Decanen, in het openbaar te verdedigen op vrijdag 22 april 2016 om 12.00 uur. door Jill Thijssen Promotores Prof. dr. C. de Ruiter Prof. dr. P. Muris Beoordelingscommissie Prof. dr. T. van Amelsvoort (voorzitter) Prof. dr. M. Dekovic (Utrecht University) Dr. L. van Domburgh (VUmc) Dr. J. Roelofs The research presented in this dissertation was supported by The Netherlands Organisation for Health Research (ZonMw, grant number 157001022), Innovatiefonds, VSB fonds, and Kinderpostzegels Nederland. CONTENTS Chapter 1 General introduction 7 PART I: Parent Management Training – Oregon model 21 Chapter 2 The effectiveness of Parent Management Training - Oregon Model 23 in clinically referred children with externalizing behavior problems in The Netherlands Chapter 3 Treatment fidelity as a determinant of the effectiveness of Parent 47 Management Training – Oregon model in The Netherlands Chapter 4 Initial validation of the Dutch translation of the Caregiver Wish List, 63 an interview-based scale for measuring parenting practices PART II: Callous-unemotional traits 81 Chapter 5 Emotional memory for central and peripheral details in children 83 with callous-unemotional traits Chapter 6 Emotional true and false memories in children with callous- 95 unemotional traits Chapter 7 General discussion Summary Samenvatting Valorization addendum References Dankwoord (acknowledgments) Curriculum Vitae 105 123 129 135 143 157 163 General introduction 1 This chapter is a translated and extended version of the following chapter: Thijssen, J., de Ruiter, C., & Albrecht, G. (2008). Preventie van antisociaal gedrag bij kinderen: Parent Management Training Oregon. [Prevention of antisocial behavior in children: Parent Management Training Oregon]. In J.R.M. Gerris en R.C.M.E. Engels (Eds.), Vernieuwingen in jeugd en gezin: Beleidsvisies, gezinsrelaties en interventies (p. 125 - 140). Assen: Van Gorcum. General introduction | Chapter 1 The case of Tim Tim is eight years old and has a difficult temperament. His parents are divorced and he lives with his mother Kate. Tim spends every other weekend with his father. Kate does not have a lot of financial means and she feels depressed which makes her less patient with Tim. Kate gets easily irritated by his disobedient behavior, while Tim is frustrated because he feels he cannot do anything right. Positive behaviors do not meet with compliments, which makes Tim use negative behaviors to attempt to obtain what he wants. The negative behavior between parent and child becomes a standard in the lives of Kate and Tim. Because Tim regards negative behavior as normal, he also starts to demonstrate negative behavior at school. His friends reject him because of his aggressive behavior, which makes Tim develop low self-esteem. Over the years, the negative interactions between Kate and Tim become more ingrained and the emotional distance between mother and son keeps growing. Kate does not have a grip on Tim anymore and she has no idea what he is doing in his spare time. At age 13, Tim starts hanging out with older delinquent boys who encourage him to commit crimes. Because Tim does not feel supported at home and his delinquent friends do not think school is important, he drops out of school at the age of 16. Regularly, he gets into trouble with the police. At the age of 20, Tim gets arrested on suspicion of dealing drugs. Externalizing behavior problems In the Netherlands, 13.6% of the children between 4 and 11 years of age show externalizing behavior problems (de Looze et al., 2014). Externalizing behavior problems refer to aggressive (fighting, destroying), oppositional (disobedient, running away), and delinquent behaviors (lying, stealing, substance abuse). Aggressive behavior is part of the normal development of children (Tremblay, 2000). Between the age of 2 and 3, most children start to show physical aggression as a natural way of expressing anger. After their third birthday, when they learn to talk, there is a decline in the level of physical aggression (Alink et al., 2006). However, some children continue to show a high level of aggression and are at risk of developing a chronic pattern of physical aggression and delinquency (Broidy et al., 2003; Nagin & Tremblay, 1999). This can result in the development of disruptive behavior disorders, such as Oppositional Defiant Disorder (ODD) and Conduct Disorder (CD; American Psychiatric Association [APA], 2013). ODD consists of a persistent pattern of negativistic, defiant, disobedient, and hostile behavior towards authority figures. Children with ODD annoy other people, refuse to comply with requests or rules of adults, blame others for their own mistakes, are easily irritated and often angry (APA, 2013). CD comprises a recurrent pattern of behavior in which the basic rights of others or major age-appropriate societal norms or rules are 9 1 Chapter 1 | General introduction violated. Children with CD steal, vandalize, cause physical harm to other people or animals, and lie (APA, 2013). ODD is often seen as a precursor or less severe form of CD (Loeber & Burke, 2011). Both disorders have a high co-morbidity with Attention Deficit Hyperactivity Disorder (ADHD; APA, 2013). Early signs and consequences Several longitudinal studies have demonstrated that already during toddlerhood there are behavioral characteristics which predict later antisocial behavior (Coté, Vaillancourt, LeBlanc, Nagin, & Tremblay, 2006; Simonoff et al., 2004). Indeed, the earlier externalizing behavior problems begin, the higher the risk of serious problems during adulthood (Moffitt, 1993; Patterson, DeBaryshe, & Ramsey, 1989). In toddlers, behavioral problems are characterized by behaviors such as disobedience, temper tantrums and hitting. During the school years, new forms of antisocial behavior develop, such as fighting, stealing, and cruelty to animals, while substance abuse and robbery occur in adolescence (Patterson, Dishion, & Yoerger, 2000). When externalizing behavior problems in childhood are not addressed properly, about 50 to 75% will continue to show serious problem behaviors up to six years later (Nixon, 2002). Externalizing behavior problems are a major social and financial burden on the community (Bennett et al., 1999). Scott, Knapp, Henderson, and Maughan (2001) examined the financial costs of children with disruptive behavior problems for public services. They followed 142 10-year-old children from inner London up to adulthood (until age 28). Based on their childhood profile, the 142 children were divided into three groups: children with no problems, children with some disruptive behavioral problems, and children with conduct disorder. They found that by age 28, the costs for individuals with conduct disorder (£70 019) were 10 times higher than for children with no problems (£7 423) and 3.5 times higher than for children with mild behavioral problems (£24 324). The highest costs were incurred by crime, followed by the provision of extra education, and foster and residential care (Scott et al., 2001). Romeo, Knapp, and Scott (2006) found that the mean annual costs for children with severe antisocial behavior in the UK were £5 960. The National Health Service, education and volunteer agencies were the most commonly used services. However, over three-quarters of the total costs fell on the family itself (e.g., additional household tasks, house repairs due to the child’s destructiveness, taking days off work because the child was sent home from school). Thus, the financial costs of externalizing behavior problems are large and involve several agencies (Romeo et al., 2006; Scott et al., 2001). 10 General introduction | Chapter 1 Parenting A large number of risk factors have been reported as being associated with the development and persistence of externalizing behavior problems, such as low serotonin level, low resting heart rate, impulsivity, difficult temperament, associations with deviant peers, and living in a disadvantaged neighborhood (Frick, 2006). Probably the most studied and established risk factor for the development of externalizing behavior problems is ineffective parenting (DeBaryshe, Patterson, & Capaldi, 1993; Forgatch, Bullock, & Patterson, 2004; McCoy, Frick, Loney, & Ellis, 1999; Nix et al., 1999; Patterson, Forgatch, Yoerger, & Stoolmiller, 1998). In particular, harsh and inconsistent discipline practices, poor monitoring and supervision, and low positive involvement have been associated with externalizing behavior problems in a number of studies (Bierman & Smoot, 1991; Cunningham & Boyle, 2002; Ehrensaft et al., 2003; Frick et al., 1992; Nicholson, Fox, & Johnson, 2005). Persistent ineffective parenting can result in the development of overt forms of externalizing behavior problems, such as hitting, temper tantrums and noncompliance. Children learn and maintain the overt forms of externalizing behavior by negative reinforcement of antisocial behavior by family members (Snyder & Patterson, 1995). In a well-functioning family, the child learns to use prosocial skills during conflicts. However, in a dysfunctional family, the payoffs are higher for coercive (negative) responses than for prosocial reactions (see the case of Tim presented above). Coercive processes usually become automatic with little or no cognitive awareness (Forgatch et al., 2004). Prosocial behaviors in such families remain unnoticed and are not rewarded. The antisocial behaviors learned at home can generalize to other settings, such as school and sports clubs (DeBaryshe et al., 1993). Noncompliance with teachers puts these children at risk of poor school achievement (Ledingham & Schwartzman, 1984; Prior, Smart, Sanson, & Oberklaid, 1999). With peers, the antisocial behavior of the child can lead to rejection by prosocial peers (DeGarmo & Forgatch, 2005; Patterson, 1986). School failure and rejection by peers can lead to low self-esteem and depressed mood, but also to associations with deviant peer groups (DeGarmo & Forgatch, 2005; Snyder, 2002). Through interaction with deviant peers, children learn covert forms of antisocial behavior, such as lying, stealing and substance abuse (Forgatch, Patterson, DeGarmo, & Beldavs, 2009). In deviant peer groups it pays off to show deviant behavior. However, parents also contribute to covert antisocial behaviors by coercive parenting and poor monitoring. The latter enables the youngsters to wander away from home and to engage in undesirable activities. Eventually, this can lead to delinquency. Thus, overt antisocial behavior is maintained through negative reinforcement by family members, while covert antisocial behavior is mainly maintained by positive reinforcement by deviant peer group members (Forgatch & Patterson, 2010). Youngsters who combine 11 1 Chapter 1 | General introduction overt and covert forms of antisocial behavior problems are at highest risk to become career offenders (Patterson, Forgatch, & DeGarmo, 2010). As outlined above, externalizing behavior problems in childhood can have major social and financial consequences. Therefore, it is important to intervene as soon as the child shows signs of externalizing behavior problems. Many interventions for childhood behavior problems focus on increasing effective parenting practices and decreasing the use of ineffective rearing behaviors. Results of several studies support the theory that an improvement of parenting practices is the most important mechanism that contributes to more adaptive behavior of the child (DeGarmo & Forgatch, 2005, 2007; Forgatch & DeGarmo, 1999; Martinez & Forgatch, 2001; Ogden & Amlund-Hagen, 2008). Social Interaction Learning theory The Social Interaction Learning (SIL) theory describes the mechanisms that could stimulate or prevent antisocial behavior. Figure 1 shows that the child is surrounded by two layers of context. The inner layer represents the parents, while the outer layer refers to the background context. On the left side of the inner layer are the effective (positive) parenting practices (discipline, monitoring, positive involvement, skill encouragement, and problem solving). On the right side are the coercive (negative) parenting practices (negative reciprocity, escalation, and negative reinforcement), which hinder a healthy social-emotional development of the child (Forgatch et al., 2004). The layer of the parenting practices is surrounded by contextual factors. These contextual factors influence the quality of parenting. For example, when a father experiences a lot of stress from work, he may become more irritable and less patient with his child at home. Therefore, these contextual factors indirectly also have an effect on the development of the child. On the other hand, parents also influence the wider social context, for instance, they can make sure the child grows up in a safe environment (Forgatch et al., 2004). The parenting skills will be briefly explained. Discipline decreases the problem behavior by appropriate and consistent use of mild sanctions, such as giving a time out or taking away privileges (Patterson, 1986). Monitoring involves knowing the child’s friends and keeping track of the child’s activities to protect the child against associations with deviant peers (Snyder, 2002). Positive involvement concerns the many ways in which parents give their child loving attention, spending time together and engaging in fun activities (Forgatch & DeGarmo, 1999; Forgatch, Patterson, & DeGarmo, 2005b). Skill encouragement stimulates the development of prosocial behaviors by means of scaffolding techniques (e.g., breaking behavior into small steps, prompting appropriate behavior) and positive reinforcement (e.g., giving compliments when the child behaves well; Forgatch & DeGarmo, 1999). Finally, the parenting practice problem solving helps parents to negotiate during arguments and to clearly determine rules and consequences when the rules are violated (Forgatch et al., 2005b). 12 General introduction | Chapter 1 1 䘀愀洀椀氀礀 猀琀爀甀挀琀甀爀攀 䔀猀挀愀氀愀愀漀渀 匀漀挀椀愀氀 猀甀瀀瀀漀爀琀 一攀最愀愀瘀攀 爀攀椀渀昀漀爀挀攀ⴀ 洀攀渀琀 Figure 1 | The Social Interaction Learning model (Forgatch et al., 2004) Coercion is a pattern of social interaction which hinders the development of prosocial behaviors in the child and stimulates problem behavior. Coercive parenting is characterized by irritable commands and negative attributions about others, but also by other ineffective techniques such as inconsistency, threatening but not following through, and high emotional intensity sanctions (Martinez & Forgatch, 2001; Patterson et al., 2010). Negative reciprocity arises when parents pay more attention to what the child is doing wrong than to what the child is doing right and when parents frequently start arguing with their child. This often leads to escalation. When the parent does not stop the argument in time, emotions will get the upper hand. Negative reciprocity creates irritation in the parent and anger in the child. At the moment the escalation reaches its peak, the child starts screaming and throwing things around, for example, and the parent yells to the child to go upstairs. After the peak, the escalation decreases and everything becomes relatively calm again. However, the confrontation has not been solved in a satisfactory way, which means there is high risk of a new negative reciprocity and escalation. Through negative reinforcement children learn that negative behavior pays off in the short term. For example, a child and the parent go to the supermarket and the child wants some candy but does not get it. However, the child has learned from previous experiences that when he screams long and hard, the parent will eventually give in and buy what the child wants. In this way, parents develop the feeling that they do not have 13 Chapter 1 | General introduction any control over their child’s negative behavior (Snyder, Edwards, McGraw, Kilgore, & Holton, 1994). Negative contextual factors, such as an unsafe neighborhood, low socio-economic status, stressful life events (e.g., divorce), substance abuse, and mental health problems can cause an imbalance between coercive and positive parenting practices. For example, when the father loses his job, he may become more irritable which in turn can lead to increased coercive parenting. Biological factors are also included among the contextual factors. Several studies have shown that, for example, having a difficult temperament or ADHD can lead to externalizing behavior problems (Bates, Pettit, Dodge, & Ridge, 1998; Chronis et al., 2007; Johnston & Jassy, 2007; Simonoff et al., 2004). When families have to deal with such negative contextual factors, adequate parenting practices become even more important. Positive parenting practices ensure that the child is able to cope with the negative influence of negative contextual factors, thereby acting as a buffer (Forgatch et al., 2004). PMTO Parent Management Training - Oregon model (PMTO) is a training for parents of children with serious externalizing behavior problems such as ODD and CD, but also ADHD. It is based on the SIL model. PMTO focuses on children between 4 and 12 years of age, because this type of preventive parent training is most effective with young children. Reasons for this are that their behavior is still malleable and for children of this age parents are still the main socializing agents (McCart, Priester, Davies, & Azen, 2006). The basic assumption of PMTO is that not only the child is the problem, but the social environment of the child is involved as well. Changing the behavior of aggressive children means that the social environment in which the child is living also needs to be changed (Patterson, Reid, & Eddy, 2002). The parents or caretakers play a very important role in maintaining the quality of the social environment of their child. A core principle of the PMTO training is the balance of five positive reinforces against one negative one (Forgatch et al., 2004). This means that for every limit set, five compliments or encouragements need to be given. Parents who have children with externalizing behavior problems generally pay attention to the problem behaviors, while ignoring the positive, prosocial behaviors. Teaching parents to encourage their children forces them to focus more on their child’s positive behavior. The 5:1 principle teaches parents to recognize and reinforce positive behaviors, which makes the child better able to accept limit setting (Forgatch et al., 2004). 14 General introduction | Chapter 1 Content of the PMTO intervention PMTO is an individual therapy of 15 to 25 weekly one-hour sessions in which the therapist works with the parent(s) of one family. The children are not present during these sessions. The PMTO therapy aims to teach the parents how to reduce coercive parenting practices and to replace them by the five effective parenting practices. Usually, the therapy takes place at the site where the therapist works, but sometimes at the parents’ home as well. Role play is an important mechanism in the PMTO sessions to teach the parents effective parenting practices. Simply telling parents what they should do, does not lead to change. The new parenting practices have to be practiced repeatedly until they become automatic (Forgatch et al., 2004). From other parenting therapies it was learned that there was a lot of resistance among parents to change their behaviors, but by means of the frequent use of role play this resistance tends to decrease (Patterson, 2005). The therapist uses the role play to demonstrate good and bad examples of parenting. One of the parents plays the role of the child and directly experiences the positive and negative effect of these behaviors. Afterwards, the therapist asks the parents to indicate which specific behaviors stimulated or hindered a positive outcome. The role play can also be used as a diagnostic instrument to examine the strengths and weaknesses of the parents, to determine what the parents have already learned, and to see which skills need extra attention (Forgatch et al., 2004). As soon as the parents have sufficiently mastered the pertinent parenting behavior, the therapist and parent select a specific situation for the upcoming week in which the parent will try to apply the newly acquired skill at home. This increases the chance that the parent will actually use the skill in daily life. During that week, the therapist calls the parent to offer support and to answer questions. In the subsequent session, the therapist evaluates the success of the newly acquired parenting skill and discusses how this could be further improved (van Leeuwen & Albrecht, 2008). One of the parenting practices parents learn with PMTO is stimulating their children by encouragement. The therapist uses this skill during the sessions with the parents as well. When the parent says or does something right, the therapist gives a compliment or a reward in the form of Scooby Loops. Scooby loops are elastic wrist bands which the parents put on their wrist. The effect of the wrist bands for the parents is the same as for the child: the confidence and the appropriate behaviors, in this case the adequate parenting practices, will increase. Parents also use the wrist bands to remind themselves to compliment their child. They start the day with, for example, five wrist bands and each wrist band represents a compliment to the child. At the end of the day, all wrist bands must have been used. The earned wrist bands can also be exchanged for a bigger reward for the child, such as playing on the iPad for half an hour. 15 1 Chapter 1 | General introduction Integrity of PMTO The training to become a PMTO therapist encompasses 18 days. After the first nine days of training, the therapist-in-training starts with his first PMTO-family. All therapy sessions are recorded on DVD. The supervisor uses the DVD’s to give feedback, but also for evaluating treatment fidelity. During the last nine days of the training, the therapistin-training brings the DVD’s of the therapy sessions for further coaching. A therapist has to treat at least three families before the certification procedure can start. The therapist is being rated on the basis of four DVD’s. When the therapist performs the PMTO principles correctly, he receives the PMTO license. The duration of the licensing process varies per therapist, because it is fitted to individual learning processes. In general, the training takes about 13 to 15 months (Berger & van Everdingen, 2006). After finishing the PMTO training, there will still be regular supervision meetings and the therapist has to recertify every year to retain the PMTO license. The DVD’s of the therapy sessions are judged by using the Fidelity of IMPlementation rating system (FIMP; Knutson, Forgatch, Rains, & Sigmarsdóttir, 2009). This rating system determines to what extent the therapist conducted the PMTO intervention according to its principles. Research has shown that high FIMP scores are good predictors of improvements in parenting practices. Therefore, being loyal to the PMTO principles predicts a positive therapy outcome (Forgatch et al., 2005b). During supervision the FIMP is used to see which FIMP elements the therapist is performing well and which elements could be used more and/or better. With certification and recertification, the FIMP can also help to make a decision. Hence, the FIMP is an important part of the PMTO training. Effectiveness of PMTO The past four decades, a lot of research has been conducted on the effectiveness of PMTO. In the 1970’s, a few small randomized controlled trials were performed which showed that parent training is more effective with aggressive boys than a placebo condition (Walter & Gilmore, 1973) and more effective than a waiting list control condition (Wiltz & Patterson, 1974). Later, larger-scale trials were carried out which replicated these results (Bank, Marlowe, Reid, & Patterson, 1991; Patterson, Chamberlain, & Reid, 1982). PMTO appeared not only to be effective in complete families with two biological parents. The effectiveness of PMTO has also been established in divorced families (Forgatch & DeGarmo, 1999). A sample of 238 recently divorced mothers were randomly assigned to PMTO or a no intervention control condition. After 12 months, the researchers found that in the PMTO condition the effective parenting practices had improved compared to mothers in the control condition. In a follow-up study, it was found that the parenting practices of these mothers changed within the first 12 months, 16 General introduction | Chapter 1 followed by changes in the behavior of the children and eventually a significant decrease in the depression level of the mother after 30 months (DeGarmo, Patterson, & Forgatch, 2004). Even at 9 years follow-up, there was a significant difference in delinquency and the number of police arrests between the boys in the PMTO condition compared to the control group (Forgatch et al., 2009). The effectiveness of PMTO has also been investigated in the context of stepfamilies (Forgatch, DeGarmo, & Beldavs, 2005a). The study consisted of 110 remarried biological mothers and stepfathers who were randomly assigned to either PMTO or a no intervention control condition. After 12 months, PMTO was found to produce clear improvements in parenting practices and a significant reduction of externalizing behavior problems of the child (Forgatch et al., 2005a). However, this effect was less clear-cut at 24 months. This was mainly due to the fact that the quality of the parenting practices in the control group had increased. Thus, at first, a remarriage can be problematic for the parenting of the stepfather, but eventually his skills also improve during the natural course of stepfamily formation (DeGarmo & Forgatch, 2007). Finally, the effectiveness of PMTO was also examined in foster families. A permanent placement is the best choice for foster children. To examine if PMTO is effective in increasing the chance of a permanent placing, 90 children between 3 and 6 years old were randomly assigned to either PMTO or treatment as usual. After 24 months, the researchers found success rates of permanent placings of 64% for treatment as usual and 90% for PMTO. Therefore, it can be concluded that PMTO increases the chance of success of a permanent placing (Fisher, Burraston, & Pears, 2005). The first randomized controlled trial (RCT) on the effectiveness of PMTO conducted outside of the US was completed in Norway, in which 112 clinically referred boys and girls aged between 4 and 12 years were randomly assigned to either PMTO or CAU (Ogden & Amlund-Hagen, 2008). Results indicated that PMTO was superior to CAU on the post-treatment outcome measures relating to effective discipline, obedience of the child, child-initiated negative behaviors and externalizing behavior problems. The effect of PMTO appeared to be moderated by the age of the child: that is, the intervention proved to be more effective in children below 8 years of age than among older children (Ogden & Amlund-Hagen, 2008). However, at one-year follow-up, the differences between PMTO and CAU on child behavior problems and parental discipline were no longer significant (Amlund-Hagen, Ogden, & Bjørnebekk, 2011). A highly similar RCT was conducted in Iceland by Sigmarsdóttir, Thorlacius, Guŏmundsdóttir and DeGarmo (2014), who also allocated clinically referred children with behavior problems aged 5 to 12 years (N = 102) to either PMTO or CAU. PMTO was found to be more effective than CAU in improving general child adjustment at post-treatment, although the only significant effect was documented for social skills. Surprisingly, this study did not obtain 17 1 Chapter 1 | General introduction support for the idea that PMTO has a positive effect on parenting skills (Sigmarsdóttir, DeGarmo, Forgatch, & Guðmundsdóttir, 2013). Callous-unemotional traits A subset of children with externalizing behavior problems is characterized by socalled callous-unemotional (CU) traits. CU traits resemble the emotional detachment component of psychopathy in adult forensic samples (Frick, 2006; Frick, Cornell, Barry, Bodin, & Dane, 2003a). Children with CU traits display low fear in combination with high impulsivity, are not so much concerned with others’ feelings, and typically do not feel bad or guilty when breaking rules. Furthermore, children who are characterized by CU traits have an earlier onset of offending and higher levels of aggression, delinquency, and police contacts compared with antisocial children without these traits (Andershed, Gustafson, Kerr, & Stattin, 2002; Frick et al., 2003a; Frick, Stickle, Dandreaux, Farrell, & Kimonis, 2005; Frick & White, 2008; Lawing, Frick, & Cruise, 2010). McMahon, Witkiewitz, and Kotler (2010), for example, found that CU traits in young adolescents with conduct disorder improved the prediction of antisocial outcomes, in terms of general delinquency and police arrests, with very low false-positive rates. A number of studies have demonstrated that CU traits are relatively stable over time, either assessed by selfreport or parent-report (Burke, Loeber, & Lahey, 2007; Dadds, Fraser, Frost, & Hawes, 2005; Frick, Kimonis, Dandreaux, & Farell, 2003b; Obradovic, Pardini, Long, & Loeber, 2007). Therefore, the presence of CU traits seems to designate a distinct group of children with conduct disorder who show a more severe and chronic pattern of externalizing behavior problems and delinquency. As outlined above, studies demonstrate that improving parenting practices is the most important mechanism contributing to the decrease of externalizing behavior in children (DeBaryshe et al., 1993; DeGarmo & Forgatch, 2005; McCoy et al., 1999; Nix et al., 1999). However, in children with CU traits, previous research has shown that ineffective parenting seems to be unrelated to the frequency and severity of antisocial behavior problems (Hipwell et al., 2007; Oxford, Cavell, & Hughes, 2003; Viding, Fontaine, Oliver, & Plomin, 2009; Vitacco, Neumann, Ramos, & Roberts, 2003; Wootton, Frick, Shelton, & Silverthorn, 1997). For example, Wootton and colleagues (1997) found that the relation between ineffective parenting and antisocial behavior problems was moderated by the presence of CU traits. Ineffective parenting was related to behavior problems only for children without CU traits, while children with CU traits showed antisocial behavior independent of the quality of parenting. One of the main reasons why children with CU traits are less responsive to effective parenting is that these children tend to be less sensitive to punishment (e.g., house arrest, privilege removal), while previous studies found that effective discipline is the most 18 General introduction | Chapter 1 important parenting practice in reducing behavior problems in the child (DeBaryshe et al., 1993; Nix et al., 1999; Ogden & Amlund-Hagen, 2008; Oliver, Guerin, & Coffman, 2009). According to the operant conditioning principle, children should learn to inhibit undesirable behaviors in order to avoid the punishment (Dadds & Salmon, 2003). However, children with CU traits have a specific deficit in processing negative stimuli (Frick, 2006). For example, they have more difficulty in identifying fearful and sad facial expressions (Blair, Colledge, Murray, & Mitchell, 2001; Stevens, Charman, & Blair, 2001; Woodworth & Waschbusch, 2008) and are less physiologically responsive to distressing and threatening pictures, films and words (Anastassiou-Hadjicharalambous & Warden, 2008; Blair, 1999; Kimonis, Frick, Fazekas, & Loney, 2006; Loney, Frick, Clements, Ellis, & Kerlin, 2003). It appears that children with CU traits respond with insufficient anxiety when parents confront them with their misbehaviors. This deficit in processing negative emotional stimuli could also affect memory. In general, emotional events are remembered better than neutral ones (Buchanan, 2007; LaBar & Cabeza, 2006). However, because children with CU traits have difficulties processing negative emotional material, they might not benefit from an enhanced memory for emotional events. This effect has already been demonstrated in adults and adolescents with psychopathic traits (Christianson et al., 1996; Dolan & Fullam, 2005, 2010). Therefore, children with CU traits are less likely to learn from a negative emotional experience. As they are less capable of remembering the experience of being punished after their misbehavior, they are less inclined to inhibit their behavior in future situations in order to avoid the punishment. Aim and outline of the thesis The aim of this thesis is to examine the effectiveness of PMTO in The Netherlands and to identify possible risk factors for the effectiveness of parent management trainings in general, and PMTO in particular. It consists of two parts. Part 1 concerns the effectiveness of PMTO in The Netherlands in reducing externalizing problem behavior in children. Part 2 concerns emotional memory in children with callous-unemotional traits. PART I: PARENT MANAGEMENT TRAINING – OREGON MODEL Improving effective parenting is seen as an important mechanism in reducing externalizing behavior problems in children (McCart et al., 2006). PMTO had been proven effective in America, Norway and Iceland (e.g., Forgatch et al., 2009; Ogden & Amlund-Hagen, 2008; Sigmarsdóttir et al., 2014). However, this does not guarantee that PMTO also works in the Netherlands. Therefore, the effectiveness of PMTO in the Netherlands was examined. Families of clinically referred children (N = 146) with 19 1 Chapter 1 | General introduction externalizing behavior problems aged between 4 and 11 years were assigned to either PMTO or CAU. In Chapter 2, the results on behavior problems, parenting practices, and parental psychopathology are presented. Research has shown that there is a positive association between treatment fidelity and the effectiveness of PMTO (Forgatch et al., 2005b). The aim of Chapter 3 is to examine whether this relation is also true for Dutch PMTO therapists. Hence, the treatment outcomes of the 86 PMTO families from the effectiveness study as described in Chapter 2 were related to the FIMP-scores of the PMTO therapists. Research has revealed the crucial importance of adequate parenting for a healthy social-emotional development of children. Thus, the need for valid assessment of parenting quality cannot be overestimated, particularly in clinical settings. The Caregiver Wish List (CWL; Hodges, 2002) is a new 53-item interview scheme with Likerttype scales for measuring parenting practices as defined by the SIL-model. Chapter 4 describes a validation study of the Dutch-language version of the CWL. We examined the factor structure, reliability, and validity of this instrument in a sample of 348 parents of children aged between 4 and 11 years. PART II: CALLOUS-UNEMOTIONAL TRAITS Children with callous-unemotional (CU) traits are less responsive to parenting than children without these traits. Children with CU traits have a specific deficit in processing negative stimuli, which make them respond with insufficient anxiety when confronted with their misbehavior (Frick, 2006). Therefore, we examined emotional memory of children with CU traits. In Chapter 5, we used a self-developed peripheral memory test to test the memory for central and peripheral components of neutral and emotionally negative pictures in 8-to-12-year-old non-clinical children. The deficiency in emotional processing in children with CU traits could also target incorrect remembering (i.e., false memories). The study described in Chapter 6 examined false memories in children with CU traits using neutral and negative word lists of the Deese-Roediger-McDermott (DRM) paradigm (Deese, 1959; Roediger & McDermott, 1995). Finally, in Chapter 7 the findings of the studies in the preceding chapters are summarized and discussed. Implications for clinical practice and recommendations for future research are presented. 20 P a r e n t Management Training - Oregon model PART I The effectiveness of Parent Management Training-Oregon model in clinically referred children with externalizing behavior problems in The Netherlands 2 Thijssen, J., Vink, G., Muris, P., & de Ruiter, C. (2015). The effectiveness of Parent Management Training - Oregon Model in clinically referred children with externalizing behavior problems in The Netherlands. Manuscript accepted with revisions Chapter 2 | Effectiveness of PMTO ABSTRACT The present study examined the effectiveness of Parent Management Training - Oregon model (PMTO) as a treatment for children with externalizing behavior problems in The Netherlands. Clinically referred children (N = 146) aged 4 to 11 years and their parents were included in this research of which 91 received PMTO and 55 Care As Usual (CAU). Families were assessed at four time points: at pretreatment, and after 6, 12, and 18 months. Results showed that both PMTO and CAU were effective in reducing child externalizing behavior, parenting stress and parental psychopathology, with no significant differences between the two treatment conditions. PMTO and CAU interventions also produced some improvements in self-reported parenting skills, but not in observed parenting skills. Finally, the effect size of PMTO on parent-reported externalizing behavior problems as found in the present study was comparable to that found in previous studies evaluating PMTO as an intervention for this type of child psychopathology. 24 Effectiveness of PMTO | Chapter 2 INTRODUCTION Ineffective parenting is a well-established risk factor for the development of externalizing behavior problems in children (DeBaryshe, Patterson, & Capaldi, 1993; Forgatch, Bullock, & Patterson, 2004; McCoy, Frick, Loney, & Ellis, 1999; Nix et al., 1999; Patterson, Forgatch, Yoerger, & Stoolmiller, 1998). The role of parenting in the emergence and maintenance of problematic child behavior is cogently explicated in Patterson’s Social Interaction Learning (SIL) model (Forgatch et al., 2004). Briefly, the SIL model assumes that contextual factors, such as socio-economic disadvantage and parental psychopathology, have a negative impact on parenting quality. Further, ineffective rearing practices have a direct detrimental influence on the behavior of the child, thereby hindering its healthy social-emotional development. More precisely, persistent coercive parenting – which is characterized by hostility and holding power over children via punitive or psychologically controlling means – can promote overt forms of externalizing behavior problems, such as noncompliance, temper tantrums, and verbal and physical aggression, which in turn are maintained by negative reinforcement of the parents (Snyder & Patterson, 1995). When children become more negative in their behavior, they are harder to discipline, which leads to parents using even more aversive strategies (Patterson, 1986). In this way, families become entangled in a downward spiral of negativity. The antisocial behaviors acquired at home also tend to generalize to other social settings, such as school and sporting clubs (DeBaryshe et al., 1993). Within the peer group, the antisocial behavior can lead to rejection by normal, prosocial peers. In turn, this can lead to associations with deviant peer groups (DeGarmo & Forgatch, 2005; Patterson, 1986; Snyder, 2002) in which it pays off to show negative behaviors like lying, stealing, and vandalism (Forgatch, Patterson, DeGarmo, & Beldavs, 2009). However, parents also make a contribution to such deviant behavior by poor monitoring of the whereabouts and behaviors of their children outside the home environment. It enables youngsters to wander away from home and to engage in, for example, drug use and criminality (Patterson, Forgatch, & DeGarmo, 2010). These antisocial behaviors in childhood may take the form of an Oppositional Defiant Disorder (ODD) or Conduct Disorder (CD; American Psychiatric Association, 2013), which have been shown to be possible precursors of Antisocial Personality Disorder in adulthood (Loeber & Burke, 2011). The fact that externalizing behavior problems in children can have significant negative long-term consequences, underlines the importance of early intervention programs. Many of these programs focus on the improvement of parenting practices and there is indeed evidence showing that the enhancement of positive and more effective parenting is an important mechanism that promotes children’s prosocial 25 2 Chapter 2 | Effectiveness of PMTO behavior (DeGarmo & Forgatch, 2005, 2007; Forgatch & DeGarmo, 1999; Martinez & Forgatch, 2001; Ogden & Amlund-Hagen, 2008). A good example of an intervention that is based on the key principles of the SIL model is Parent Management Training Oregon model (PMTO). The program is especially developed for the parents of children between 4 and 12 years of age showing the severe behavior problems associated with ODD or CD and aims to teach parents how to reduce coercive parenting practices and to replace these with five effective parenting practices: encouragement (i.e., stimulation of prosocial behaviors in the child by using scaffolding techniques and positive reinforcement), effective discipline (i.e., consistent use of mild sanctions like giving a time out), monitoring (i.e., knowing the child’s friends and keeping track of its activities), problem solving (i.e., responding effectively to rule-breaking behaviors and settling arguments with the child), and positive involvement (i.e., giving love and warm attention and engaging in fun activities with the child; Forgatch & DeGarmo, 1999; Forgatch, Patterson, & DeGarmo, 2005b). Initial studies conducted in the Unites States (US) have demonstrated that PMTO is an effective intervention for reducing externalizing child behavior problems (e.g., Bank, Marlowe, Reid, & Patterson, 1991; Patterson, Chamberlain, & Reid, 1982). For instance, in the study by Forgatch and DeGarmo (1999), 238 recently divorced mothers were randomly assigned to PMTO or a no intervention control condition. After 12 months, it was found that in the PMTO condition the effective parenting practices had significantly improved compared to the control condition. At a long term follow-up, 9 years after the PMTO intervention, there was still a significant difference between the boys in the PMTO condition and the control group with the former showing lower levels of delinquency, criminal activities, and convictions (Forgatch et al., 2009). Furthermore, PMTO has also been shown to be effective in newly formed families consisting of biological mothers and stepfathers: again, parenting practices improved and behavior problems of the child decreased, as compared with newly formed families who did not receive an intervention (Forgatch, DeGarmo, & Beldavs, 2005a). Finally, in foster families, researchers found a success rate of permanent placements of 90% for PMTO versus 64% for Care As Usual (CAU) at an assessment which took place at 24 months after the interventions. PMTO was also significantly associated with reductions of stress for both the children and the foster parents (Fisher, Burraston, & Pears, 2005). The first randomized controlled trial (RCT) on the effectiveness of PMTO conducted outside of the US was completed in Norway, in which 112 clinically referred boys and girls aged between 4 and 12 years were randomly assigned to either PMTO or CAU (Ogden & Amlund-Hagen, 2008). Results indicated that PMTO was superior to CAU on the post-treatment outcome measures relating to effective discipline, obedience of the child, child-initiated negative behaviors and externalizing behavior problems. The effect 26 Effectiveness of PMTO | Chapter 2 of PMTO appeared to be moderated by the age of the child: that is, the intervention proved to be more effective in children below 8 years of age than among older children (Ogden & Amlund-Hagen, 2008). Further, at a one-year follow-up, the differences between PMTO and CAU on child behavior problems and parental discipline were no longer significant (Amlund-Hagen, Ogden, & Bjørnebekk, 2011). A highly similar RCT was conducted in Iceland by Sigmarsdóttir, Thorlacius, Guŏmundsdóttir and DeGarmo (2014), who also allocated clinically referred children with behavior problems aged 5 to 12 years (N = 102) to either PMTO or CAU. PMTO was found to be more effective than CAU in improving general child adjustment post-treatment, although the only significant effect was documented for social skills. Surprisingly, this study did not obtain support for the idea that PMTO would have a positive effect on parenting skills (Sigmarsdóttir, DeGarmo, Forgatch, & Guðmundsdóttir, 2013). The present study evaluated the effectiveness of PMTO in The Netherlands. Onehundred-and-forty-six families of clinically referred children with externalizing behavior problems aged between 4 and 11 years were assigned to either PMTO or CAU. Effects of PMTO and CAU were examined by means of measures of child externalizing behavior, parenting skills, and parental stress and psychopathology, which were administered at baseline, and three follow-up measurements after six, 12, and 18 months. Parents’ treatment satisfaction was also evaluated after 6, 12, and 18 months. In addition, effect size and clinically significant change in children’s externalizing behavior problems was examined and compared across both treatment conditions, and several possible moderators of the effects produced by PMTO were explored (i.e., child, parent, and family variables). The following hypotheses were tested: (1) PMTO will result in greater improvements of children’s behavior problems, parenting skills, and parental stress and psychopathology than CAU; (2) PMTO will be associated with higher treatment satisfaction of parents and fewer dropouts as compared to CAU; and (3) PMTO will show a larger effect size and greater proportion of clinically significant change than CAU. With regard to moderator effects, predictions were less obvious, although it can be hypothesized that PMTO is more effective in families displaying the characteristics that are the target of this intervention (i.e., poor parenting skills) or that facilitate the application of the newly acquired skills in daily life (e.g., higher educational level of parent). 27 2 Chapter 2 | Effectiveness of PMTO METHOD Participants Participants were 146 children and their parents, who were recruited at five child service agencies across The Netherlands. Of these children, 104 (71.2%) were boys and 42 (28.8%) were girls. At baseline, the age of the children ranged between 4 and 11 years, with a mean age of 7.13 years (SD = 1.75). Based on the Diagnostic Interview Schedule for Children (DISC; Shaffer, Fisher, Lucas, Dulcan, & Schwab-Stone, 2000), 75.4% of the children met the DSM diagnostic criteria for ADHD, 67.3% for ODD, and 11.6% for CD. The mean age of the main caregiver was 37.39 years (SD = 8.09). The vast majority of the main caregivers was female (90.5%), had the Dutch nationality (89.7%) and was employed (76.0%). One-hundred-and-thirteen of the 146 main caregivers (77.4%) were living with a partner, which in the majority of the cases (n = 106) was the other biological parent of the child. Five of the main caregivers were the adoptive parent of the child. Sixty-eight percent of the main caregivers had a college or university degree, 30% had completed high school, while the remaining 2% had only finished elementary school. To be included in the study, the child had to reveal a T-score of 60 or higher on the externalizing subscale of the Child Behavior Checklist (CBCL; Achenbach, 1991; Dutch version: Verhulst, van der Ende, & Koot, 1996) and the child had to be residing at home with at least one biological or adoptive parent. Exclusion criteria were: severe intellectual disability or psychopathology of the parent(s) as this that would interfere with participation in treatment, sexual abuse in the family, and a child IQ lower than 70. Procedure Families were included in the period between June 2009 and January 2014. As soon as families were referred to the child service agency, it was checked whether they met the inclusion criteria for the study. Families who met the criteria received information about the study and its procedure and were invited to participate. When parents agreed, they were asked to give their written consent. The study protocol was approved by the Medical Ethics Committee of Maastricht University Medical Centre. Allocation to the treatment conditions (PMTO and CAU) was random at three of the five child service agencies. At the other two agencies no randomization took place. After the study had started, one child service agency decided it would no longer offer CAU and thus only recruited families for the PMTO condition. The fifth agency was specifically included in the study to compensate for the missing CAU families, but unfortunately this agency was less successful in recruiting participants for the study. Eventually, this resulted in unequal sample sizes for the two treatment conditions, with 94 families receiving PMTO and 61 families receiving CAU. In the PMTO condition, 17 families (18%) dropped out, of which two families never started. In the CAU condition, 10 families 28 Effectiveness of PMTO | Chapter 2 (16%) dropped out of which seven families never started. No data could be collected for the families that never started. These families were discarded from the data analysis, leaving a final sample of 91 PMTO families and 55 CAU families. Enrollment (N = 155) Allocated to PMTO (n = 94) Allocated to CAU (n = 61) Received PMTO (n = 91) Received CAU (n = 55) Received PMTO (n = 91) Baseline (T0)Received CAU (n = 55) IQ screening DSM-IV diagnosis Child behavior problems Parenting skills Parental stress and psychopathology Parental stress and psychopathology After 6 months (T1) Child behavior problems Parenting skills Parental stress and psychopathology Parent satisfaction After 12 months (T2) Child behavior problems Parenting skills Parental stress and psychopathology Parent satisfaction After 18 months (T3) Child behavior problems Parenting skills Parental stress and psychopathology Parent satisfaction Figure 1 | Flowchart of the present study. PMTO = Parent Management Training – Oregon model; CAU = Care As Usual. 29 2 Chapter 2 | Effectiveness of PMTO Figure 1 gives an overview of the assessments that were carried out during the course of this study. As can be seen, assessments were performed at four time points: at baseline (T0), and at six-months (T1), 12-months (T2), and 18-months (T3) followup. IQ tests, interviews and video observations were mostly conducted at the child service agency (with a few exceptions at the families’ home), while questionnaires were administered to parents and teachers through a web-based system that could be approached by a computer in the home or agency environment. The assessments were conducted by trained research assistants who were not involved in the treatment of the families. The parent questionnaires and interviews were completed by the main caregiver, who was the parent who spent most time with the child. If present in the child’s family, the second caregiver was also assessed during the video observations. Participating families received a small financial compensation in the form of gift vouchers for the three follow-up assessments (i.e., €10 at T1; €20 at T2; and €30 at T3). The five child service agencies were independent therapy clinics for children. Therapists involved in the present study were all licensed clinicians working at the participating agencies. The researchers were not involved in the treatment of either PMTO or CAU. Therapists providing PMTO within the context of this study were not allowed to give CAU to families included in the study. In addition, CAU therapists were not allowed to be trained or to be in training as a PMTO therapist. Assessment Descriptive characteristics IQ measurement. To have an indication of the intellectual abilities of the children, a standardized IQ test was administered. When children were younger than 6 at baseline, the complete Wechsler Preschool and Primary Scale of Intelligence, 3rd edition (WPPSIIII; Wechsler, 2002) was used for this purpose. For older children, the short form of the Wechsler Intelligence Scale for Children, 3rd edition (WISC-III; Wechsler, 1991) was employed, which consists of two verbal (i.e., similarities and vocabulary) and two performance (i.e., picture arrangement and block design) subtests. The Dutch versions of the WPPSI-III and the WISC-III are both reliable and well-validated instruments that were current at the time that the study was conducted (Hamilton & Burns, 2003; Kaufman, Kaufman, Balgopal, & McLean, 1996; Minshew, Turner, & Goldstein, 2005). DSM-IV diagnosis. The Diagnostic Interview Schedule for Children (DISC-IV; Shaffer et al., 2000) is a highly standardized interview schedule to identify the presence of DSM-IV diagnoses in children. For the present study, the Disruptive Disorder module (module E) was used, which assesses three disorders: Attention Deficit Hyperactivity Disorder (ADHD), ODD, and CD. The interview was administered and scored by trained clinicians. 30 Effectiveness of PMTO | Chapter 2 Treatment outcome assessment Child behavior problems. The Child Behavior Checklist (CBCL) and Teacher Report Form (TRF) are widely used rating scales for assessing behavioral and emotional problems of children aged 6 to 18 years (Achenbach, 1991; Dutch version: Verhulst et al., 1996). Each scale consists of 120 items that are scored on a 3-point Likert scale (0 = not true; 1 = somewhat or sometimes true; 2 = very or often true). Items can be allocated to narrowband or DSM-based scales which represent specific types of problems (e.g., somatic complaints, social problems, affective problems, and conduct problems), or to the more general broad-band scales of internalizing (emotional) and externalizing (behavioral) problems, which can also be combined to a total score. For children aged 4 or 5 years, the 1.5–5 year version of the CBCL was used. The CBCL and TRF are widely used instruments that have been demonstrated to possess good reliability and validity (Achenbach, 1991; Verhulst et al., 1996). In the present study, the internalizing, externalizing, and total problems scores were used, which in the current study at baseline had Cronbach’s alphas of respectively .83, .84, and .90 (CBCL) and .84, .93, and .96 (TRF). The Parent Daily Report (PDR) is a reliable 34-item checklist covering the range of mild (e.g., complaining) to severe (e.g., stealing) problem behaviors (Patterson et al., 1982). The checklist is first administered face-to-face with the main caregiver to assess which specific problem behaviors of the child have occurred during the past six months. Next, these items are administered again via telephone on three consecutive days to examine whether these behaviors have occurred during the past 24 hours. The number of symptoms endorsed on the three consecutive days is summed to obtain the total PDR score. The internal consistency coefficient of the PDR at baseline was .80. Parenting practices. The Dutch translation of the Caregiver Wish List (CWL; Hodges, 2005; Hodges, de Ruiter, & Thijssen, 2009) is an interview-based instrument consisting of 53 items questioning the parent about his/her parenting skills. The interviewer reads the questions to the parent, who has to indicate the most applicable response option using a 5-point Likert scale. Items are allocated to six domains of parenting skills: providing direction and following up (4 items), encouraging good behavior (5 items), discouraging undesirable behavior (24 items), monitoring activities (13 items), connecting positively with child (3 items), and problem solving orientation (4 items). Each domain score should be regarded as a dimension with weak parenting skills on one end and strong parenting skills on the other. In the current study, the reliability of the CWL was not particularly strong: Cronbach’s alphas of various subscales were .22 for providing direction and following up, .44 for encouraging good behavior, .80 for discouraging undesirable behavior, .50 for monitoring activities, .55 for connecting positively with child, .37 for problem solving orientation, and .76 for the CWL total score. 31 2 Chapter 2 | Effectiveness of PMTO Structured Interaction Tasks (SITs) were used to observe the parenting behaviors during a series of seven structured tasks that had to be performed by parents and their child: planning a fun activity for the weekend (3 min.), problem solving, in which the family members discuss a topic chosen by the parent where they regularly have arguments about (5 min.), drawing a picture of their own house (7 min.), a snack-break to observe the family members when they do not have an assignment (5 min.), problem solving for a topic chosen by the child (5 min.), teaching tasks, which consist of two homework-like assignments of which one is slightly difficult for the child’s age and/or intellectual ability to evoke frustration (9 min.), and monitoring, in which parents had to interview their child about a moment when there was no supervision by an adult (5 min.). The SITs are designed to elicit parenting practices. The second problem solving task and the teaching tasks are performed with the main caregiver. When present, the drawing of a house is performed with only the second caregiver. The tasks were videotaped and later coded by three trained, independent raters (psychology Master’s students) using an adapted version of the Coder Impressions developed by researchers at Oregon Social Learning Center (Forgatch, Knutson, & Mayne, 1992). Briefly, the videotaped SITs were employed to score a number of items referring to domains of effective parenting behaviors of positive involvement (12 items), encouragement (20 items), problem solving (27 items), discipline (25 items), and monitoring (5 items) as well as coercion (16 items), child compliance (8 items), and interpersonal atmosphere (24 items). For each SIT domain, item scores are accumulated, with higher scores reflecting better parenting skills. Cronbach’s alphas were poor for monitoring (mothers .28; fathers .06) and coercion (mothers .62; fathers .48). The other parenting dimensions had good internal consistency (range .74 - .91). All videotapes were coded by three trained and calibrated psychology graduate students blind to treatment condition and assessment point (i.e., T0, T1, T2, or T3). Two independent raters coded a random selection of 103 SITs (29.4% of the coded SITs). Interrater reliability was examined by means of a two-way mixed, consistency, averagemeasures Intraclass Correlation Coefficient (ICC; McGraw & Wong, 1996). The ICC’s for the parenting dimensions were fair to excellent, with the exception of ICC’s for discipline and compliance, which were poor. Parental stress and psychopathology. The Nijmeegse Ouderlijke Stress Index (NOSI; de Brock, Vermulst, Gerris, & Abidin, 1992) is an adaptation of the Parenting Stress Index (Abidin, 1983) and measures stress experienced by parents in the relationship with their child. The NOSI comprises 123 items that have to be rated on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). The items refer to parent (e.g., competence, social isolation, health, relationship with spouse) as well as child (e.g., hyperactivity, demandingness, mood) characteristics. Ratings on all items can be summed to create 32 Effectiveness of PMTO | Chapter 2 a total stress score, with higher scores reflecting higher levels of perceived stress by the parent. The NOSI has adequate reliability and validity (de Brock et al., 1992). In the present study, Cronbach’s alphas were .93 for the child domain, .94 for the parent domain, and .96 for the total score. Psychological symptoms of the parents were measured by the Dutch version of the Symptom Checklist-90 Revised (SCL-90-R; Arrindell & Ettema, 1986). This version of the questionnaire is based on the SCL- 90-R of Derogatis (1977). The 90 items are rated using a 5-point scale (1 = no problem to 5 = very serious) to indicate the extent to which the parent has experienced the listed symptom during the previous week. In contrast to the original version of the SCL-90-R, the Dutch version comprises eight instead of nine subscales: Anxiety, Agoraphobia (in original version: Phobic Anxiety), Somatic Symptoms, Depression, Inadequacy of Thinking and Acting (in original version: Obsessive-Compulsive), Distrust and Interpersonal Sensitivity, Hostility, and Sleeping Problems. In the Dutch version of the SCL-90-R, the subscales Interpersonal sensitivity and Paranoid ideation (and three items from the original Psychoticism subscale) are combined into Distrust and Interpersonal Sensitivity due to insufficient discrimination between these dimensions (Arrindell & Ettema, 1986). Higher scores on the SCL-90-R indicate more serious psychopathology. In the present study, only the total score was used, which had an internal consistency coefficient of .97 at baseline. Parent satisfaction. The short form of the Working Alliance Inventory (WAI-S; Tracey & Kokotovic, 1989) was used to assess the quality of the parent-therapist alliance. The WAI-S comprises 12 items that can be allocated to three subscales of four items each: (a) agreement between parent and therapist on the goals of the therapy; (b) agreement that the tasks of the therapy will address the parent’s problems, and (c) the quality of the bond between the parent and the therapist. Normally, the items of the WAI-S are rated using a 7-point Likert scale. However, in the present study, a 5-point Likert scale was used for practical reasons. Ten items are positively worded and two items (items 4 and 10) are negatively worded. The scores on the negatively stated items are recoded, so that all scores can be summed to obtain a total score. Higher total scores indicate a better parent-therapist working alliance. In the present study, Cronbach’s alpha for the total score was .71. Interventions PMTO. PMTO is a therapy consisting of weekly sessions in which the therapist works with the parent(s) of one family. The children are not present during these sessions. The PMTO therapy aims to replace parents’ coercive parenting practices by the five effective parenting practices as defined by the SIL model. Role play is an important mechanism in the PMTO sessions to teach and extensively practice these effective parenting skills. The 33 2 Chapter 2 | Effectiveness of PMTO therapist uses the role play to demonstrate good and bad examples and to determine which parenting skills need extra attention (Forgatch et al., 2004). As soon as the parent has sufficiently practiced the parenting skill, the therapist and parent choose a specific situation for the next week during which the parent will try to apply the newly acquired skill at home. In between sessions, the therapist calls the parent for support and to answer questions. Usually, the therapy takes place at the agency, but occasionally sessions are given at the parents’ home as well. Treatment duration depends on the family’s needs and progress throughout the therapy, but typically takes between 15 and 25 weekly sessions. Parents in the present study received PMTO from 25 certified therapists. All therapists had completed the full PMTO training program of approximately 24 months. During this training period, therapists had to treat at least three families with PMTO before they were allowed to take part in the official PMTO certification procedure. This procedure involved treating another family with PMTO. On the basis of videotaped sessions of this therapy, it was determined whether or not the therapists received their license to carry out PMTO in clinical practice. Following the completion of the training program, therapists were regularly monitored on their treatment fidelity, leading to annual recertification of their license. The association between treatment integrity and treatment outcome is addressed in a separate study (Thijssen, Albrecht, Muris, & De Ruiter, 2015). Of the children for which medication use was documented, 14.8 % used additional ADHD medication. CAU. CAU-treatments were treatments that were available at the child service agencies for children with externalizing behavior problems and included family therapy (n = 31), psychiatric intensive home care (n = 10), parent therapy (n = 9), or other treatments (n = 6). In 9 CAU-families (17.6%), children received ADHD medication in combination with one of the mentioned therapies. Two families in the CAU condition received more than one treatment, which explains the higher number of CAU therapies than CAU families. Missing data The percentage of missing values in the dataset ranged from 0 % for the demographic variables to approximately 50 % for scores on the TRF. For many variables the missing values could be considered as bonafide because no score can be observed if the variable is not applicable (e.g., no observations for the second caregiver when only the main caregiver was present). We applied an intention-to-treat design by conducting multiple imputation (Rubin, 1987) to handle the missing data through a chain of conditional regression models (fully conditional specification; Van Buuren, Boshuizen, & Knook, 1999). We used predictive mean matching (PMM; Little, 1988; Rubin, 1986) for the scale variables, a custom version of PMM for scale variables that contain bonafide missings, logistic regression 34 Effectiveness of PMTO | Chapter 2 for dichotomous variables and polytomous regression for ordered categorical data. All computations were carried out with Mice (Van Buuren & Groothuis-Oudshoorn, 2011) in R (R Core Team, 2014), with 150 iterations for the algorithm to converge and 25 multiply imputed datasets, using available and custom imputation routines in Mice. The outcomes over the 25 datasets were combined into a single inference using Rubin’s rules (Rubin, 1987, p. 76). Analytic strategy Overall effectiveness was examined for all outcome measures by using repeated measures ANOVA, with time (outcome measures at T0 through T3) as the within subjects factor and treatment condition (PMTO vs. CAU) as the between subjects factor. Because previous research found that child age and gender were significantly associated with treatment outcome (Ogden & Amlund-Hagen, 2008; Sigmarsdóttir et al., 2013), we checked whether these variables had any influence on the outcome measures in our study. When this appeared to be the case, we controlled for the pertinent variable by performing an ANCOVA. Furthermore, clinically significant change in externalizing behavior problems was examined using the Jacobson-Truax Reliable Change Index (RCI), since this is the most widely-used and recommended method (Bauer, Lambert, & Nielsen, 2004; Jacobson, Roberts, Berns, & McGlinchey, 1999). This method consists of two steps. First, a cutoff point needs to be established to determine whether the child has moved from the dysfunctional to the functional range. The second step is to calculate the RCI to determine if the child’s change from pretreatment to follow-up is not the result of measurement error. When these two criteria are combined (cutoff and RCI), the children can be classified as Recovered (i.e., passed both criteria), Improved (i.e., passed RCI criterion but not the cutoff ), Unchanged (i.e., passed neither criterion), or Deteriorated (i.e., passed RCI criterion but worsened; Jacobson et al., 1999). RESULTS Baseline comparisons Families in the PMTO and CAU condition did not differ significantly on any of the demographic variables. It should be noted, however, that there was a trend towards significance for ADHD, with fewer children having this diagnosis in the PMTO condition [χ2(3) = 7.33, p = .06]. Then, we checked whether there were differences in outcome measures between the PMTO and CAU group at the baseline assessment. Only one significant difference was found: children from the CAU group displayed higher levels of behavioral problems on the PDR than children in the PMTO group [t(119) = 2.28, p = .05]. 35 2 Chapter 2 | Effectiveness of PMTO Treatment attendance The number of sessions families received was better documented for PMTO than for CAU. Reports on treatment attendance were available for 61 PMTO families and 18 CAU families, and these showed that PMTO families received more treatment sessions than CAU families. Families in the PMTO condition received on average 23.85 (SD = 9.86) treatment sessions, while families in the CAU condition received a mean of 20.50 (SD = 10.67) sessions. This difference was not significant [t(77) = 1.24, p = .22], and therefore we did not have to control for treatment attendance in our effect analyses. Effects of PMTO versus CAU Child behavior problems Mean scores (and standard deviations) of children in the two treatment conditions on various CBCL and TRF scales and the PDR are shown in Table 1. First, we compared the effectiveness of PMTO and CAU on externalizing behavior problems because this was the main outcome variable. For CBCL externalizing and the PDR, a significant main effect of time was found [F(3, 68.87) = 14.75, p < .001 and F(1.99, 315.02) = 13.17, p < .001, respectively]: post-hoc comparisons with Bonferroni correction showed that in both conditions, symptom levels significantly decreased from T0 to T1 after which they remained fairly stable at T2 and T3. No interaction effects of treatment condition and time were found, indicating that there were no significant differences in effectiveness between PMTO and CAU on these outcome measures [F(3, 112.49) < 1 and F(1.99, 897.57) = 2.37, p = .09, respectively]. For TRF externalizing, neither a main effect of time [F(3, 33.87) = 1.79, p = .17] nor an interaction effect of treatment conditions and time could be documented [F(2.66, 182.89) < 1]. Second, treatment effects on parent and teacher rated internalizing and total problems were analyzed. The pattern of results resembled that found for the externalizing behavior problems. That is, for CBCL internalizing as well as total behavior problems, a significant main effect of time was found [F(2.62, 223) = 14.14, p < .001 and F(2.82, 74.87) = 15.10, p < .001, respectively]. Pairwise comparisons using Bonferroni correction again only revealed a significant decrease in behavior problems between T0 and T1. No interaction effects of treatment condition and time were found [F(2.62, 348.94) < 1 and F(2.82, 126.29) = 1.33, p = .27, respectively]. For TRF internalizing and total behavior problems, neither significant main effects nor interaction effects were found. 36 PDR 53.75 (16.54) 58.61 (17.42) 14.14 (11.79) b 60.43 (9.40) 14.92 (15.87) b 58.81 (19.08) 55.31 (20.06) 12.77 (14.51) b 62.05 (15.48) 58.88 (14.54) 58.01 (10.51) 21.92 (12.13) 63.32 (10.79) 62.32 (11.59) T2 M (SD) T3 M (SD) 16.00 (16.77) 61.75 (10.35) 60.01 (12.80) 56.76 (12.21) 15.53 (13.52) 59.76 (16.01) 58.52 (19.83) 53.19 (18.41) 15.52 (19.21) 62.87 (15.31) 60.23 (15.93) 56.58 (18.21) 65.80 (11.61) b 64.42 (11.83) b 64.32 (11.00) b 65.80 (10.90) b 65.35 (11.18) b 66.57 (10.46) b 57.62 (11.69) b 59.81 (14.22) b 57.91 (13.21) b T1 M (SD) CAU 13.17* 2.12 1.79 1.76 15.10* 14.75* 14.14* Main effect F 2.37 .32 .53 .49 1.33 .61 .48 Time X Treatment F Note. T-scores are presented for the internalizing, externalizing and total scales of the Child Behavior Checklist (CBCL) and the Teacher Report Form (TRF). Means with different subscripts indicate significant difference at p < .05 (Bonferroni corrected). PDR = Parent Daily Report. * p < .05 62.13 (10.00) 17.64 (14.57) a TRF Total 60.38 (12.66) 56.56 (12.83) 56.95 (10.74) 62.92 (10.27) TRF Int TRF Ext 61.06 (13.10) b 60.97 (13.77) b 69.32 (6.18) a 68.60 (6.71) a CBCL Total 71.23 (6.53) a 70.90 (6.32) a CBCL Ext 63.07 (9.59) b 62.66 (9.38) a 65.32 (10.70) b 62.99 (13.60) b 63.90 (12.93) b T0 M (SD) 63.62 (9.01) a T3 M (SD) 57.86 (13.05) b T2 M (SD) 58.39 (11.93) b 58.71 (17.19) b T1 M (SD) CBCL Int T0 M (SD) PMTO Table 1 | Means, standard deviations, and treatment effects for child behavior problems Effectiveness of PMTO | Chapter 2 2 37 Chapter 2 | Effectiveness of PMTO Parenting practices Parenting practices were assessed using self-report (CWL) and structured observations (SIT). Since only the CWL subscale ‘discouraging undesirable behavior’ and the total score on this self-report measure displayed acceptable internal consistencies, only these scores were used in subsequent analyses. For discouraging undesirable behavior, a significant main effect of time was found. Parents in both conditions reported a significant increase in their employment of discouragement of undesirable behavior during the first six months [F(2.79, 53.21) = 7.21, p < .001]. For the CWL total score, also a significant main effect of time emerged [F(2.77, 79.07) = 11.05, p < .001] whereas the interaction of treatment condition and time did not attain significance [F(2.77, 978.85) = 2.15, p = .10]. Thus, in general (i.e., irrespective of treatment condition, parents reported to use more effective parenting skills over time during the first 12 months. Using the SIT data in relation to the main caregiver, no significant main effects of time or interaction effects of treatment condition and time were found. For the second parent, a significant Time x Condition interaction was found for interpersonal atmosphere [F(3, 6942.45) = 2.74, p = .04]. Pairwise comparisons using Bonferroni correction showed that this interaction was apparent between T2 and T3: The second parent in the PMTO condition showed an increase in positive interpersonal atmosphere between T2 and T3, while the second parent in the CAU condition showed a decrease. No significant effect was found for child compliance [F(3, 18) = 1.47, p = .26], which was probably due to the low base rate of child problem behavior during the SITs. The mean number of coded problem behaviors during the seven tasks at baseline (T0) was .90 (SD = 1.54), indicating that on average children showed problem behavior in fewer than one of the seven SITs. Parental stress and psychopathology The mean scores (and standard deviations) of the NOSI and SCL-90-R are presented in Table 2. To assess treatment effects on parenting stress, the NOSI was used. On the parent domain, child domain, and the total score of the NOSI, a significant main effect of time was found [F(2.42, 45.73) = 9.60, p < .001, F(2.81, 64.14) = 12.82, p < .001, and F(2.55, 79.5) = 21.37, p < .001, respectively]. Post-hoc comparisons with Bonferroni correction showed that in both treatment conditions, parenting stress significantly decreased from T0 to T1 and then remained stable at T2 and T3. No interaction effects of treatment condition and time were found, which indicates there were no significant differences in effectiveness between PMTO and CAU on parenting stress. For psychopathological complaints, as measured by the SCL-90-R, there was a significant main effect of time [F(2.26, 145.71) = 9.71, p < .001]. Pairwise comparisons, using Bonferroni correction, showed that for both treatment conditions the level of psychopathology significantly decreased from T0 to T1 after which no further change was observed. Again, no significant interaction between treatment condition and time was found [F(2.26, 259.43) < 1]. 38 111.42 (35.22) b 125.86 (41.96) a 109.66 (31.85) b 117.87 (56.27) b 184.33 (75.17) b 113.04 (36.25) b 9.71* 185.75 (70.20) b 12.82* 9.60* .63 1.72 .83 1.03 Main Time X effect Treatment F F Note. Means with different subscripts indicate significant difference at p < .05 (Bonferroni corrected). NOSI = Nijmeegse Ouderlijke Stress Index (Parenting Stress Index); SCL-90-R = Symptom Checklist-90 Revised. * p < .05 SCL-90-R 129.81 (46.29) a 115.73 (34.89) b 118.82 (64.35) b 207.41 (50.92) a 183.19 (65.87) b 127.24 (53.43) b T3 M (SD) 358.46 (82.68) a 300.50 (104.72) b 287.22 (107.35) b 283.89 (132.06) b 353.32 (95.63) a 310.31 (112.58) b 310.08 (113.89) b 313.08 (118.80) b 21.37* 168.65 (66.32) b 135.35 (60.44) b T2 M (SD) NOSI total stress T1 M (SD) 148.21 (49.84) a 128.56 (46.95) b T0 M (SD) 207.94 (45.69) a 176.85 (64.05) b 173.67 (91.32) b 115.97 (51.19) b T3 M (SD) NOSI child domain T2 M (SD) 148.93 (45.42) a 123.12 (47.43) b 125.74 (54.22) b T1 M (SD) CAU NOSI parent domain T0 M (SD) PMTO Table 2 | Means, standard deviations, and treatment effects for parental stress and psychopathology Effectiveness of PMTO | Chapter 2 2 39 Chapter 2 | Effectiveness of PMTO Parent satisfaction To examine whether there was a difference in treatment satisfaction between PMTO and CAU, the scores of the WAI-S at T1 were analyzed using an independent samples t-test. Results showed that the difference between PMTO (M = 43.95, SD = 4.67) and CAU (M = 42.94, SD = 4.75) was not significant [t(142) = 1.05, p = .16]. Effect size and clinical significance Cohen’s effect size d was calculated for the main outcome variable of the present study, i.e., parent-rated child externalizing behavior problems, in order to compare our findings to those of previous European studies.1 As can be seen in Table 3, the effect size of PMTO in the present study was comparable to the effect sizes for PMTO in the Norwegian studies and even somewhat higher than the effect size for PMTO as found in the investigation conducted in Iceland. With the exception of the effect size found for aggressive behavior for the control condition in the Norwegian follow-up study (Cohen’s d = .63), the effect size for CAU in the present study (Cohen’s d = .55) was generally higher than that obtained for the treatment control conditions in the Norwegian and Icelandic studies (Cohen’s d’s between .22 and .43). Our analyses indicate that both PMTO and CAU produced statistically significant decreases of externalizing problem behavior. However, to examine whether the improvement in child behavior was also clinically significant at the individual level, the RCI was calculated for both the PMTO and CAU condition using the CBCL externalizing behavior problems scores at T0 and T3. Based on the Jacobson-Truax method (Jacobson et al., 1999), children could be classified as recovered, improved, unchanged, or deteriorated. The percentages of children in each category for PMTO and CAU are presented in Table 4. In total, 45.8% of children in the PMTO condition improved against 42.8% of the children in the CAU condition. In the PMTO group, 16.9% of the children recovered compared with 9.7% in the CAU group. The distribution over the four categories did not differ significantly between PMTO and CAU [χ2(3) = 1.60, p = .66]. 1 40 We limited our comparison to European studies because these studies were similar to our study in terms of design, sample, and culture. The Netherlands Present study Iceland Norway CBCL ext CBCL ext Outcome measure Post treatment CBCL ext 1 year after post treatment CBCL agg CBCL del Post treatment 18 months after baseline Follow-up assessment .47 .85 .70 .73 .73 Cohen’s d PMTO Note. CBCL = Child Behavior Checklist, ext = externalizing, agg = aggressive, del = delinquent a Based on intention-to-treat (ITT) analyses. bBased on raw scores not presented in the original paper, but requested from the authors. Sigmarsdóttir et al. (2014)b a Amlund-Hagen et al. (2011) Ogden & Amlund-Hagen (2008) Norway Country Study Table 3 | Cohen’s d in four European PMTO effectiveness studies for parent reported externalizing behavior problems .39 .63 .22 .43 .55 Cohen’s d control group Effectiveness of PMTO | Chapter 2 2 41 Chapter 2 | Effectiveness of PMTO Table 4 | Percentages of reliable change based on parent rated externalizing behavior problems PMTO CAU Recovered 16.9 9.7 Improved 45.8 42.8 Unchanged 25.5 33.6 Deteriorated 11.4 14.1 Moderators To examine whether PMTO works better for certain families than for others, we tested whether factors could be identified that moderate the effect of PMTO on the main outcome variable (i.e., CBCL externalizing). Children who were classified as recovered and improved based on the RCI (n = 58) were compared with children who were classified as unchanged or deteriorated (n = 34). We examined child variables (i.e., severity of problem behavior at baseline, IQ), parent variables (i.e., age, ethnicity, education level, level of parenting skills and parenting stress at baseline, job status), and family variables (i.e., single parent household, number of siblings). Only one significant result was found for the CBCL externalizing subscale at T0 [t(333) = 2.41, p < .001]. Children who showed reliable improvement in parent rated child externalizing behavior problems showed significantly more severe externalizing behavior problems at baseline (M = 72.19) as compared to children who did not show reliable improvement (M = 68.69). DISCUSSION The present study compared the effectiveness of PMTO and CAU in Dutch children who had been referred to child care organizations because of externalizing behavior problems. It was hypothesized that children in the PMTO condition would show a greater reduction in externalizing behavior problems relative to children in the CAU condition. Furthermore, parents who received PMTO were expected to show greater improvements in effective parenting skills in comparison to parents who received CAU. Finally, we predicted that parents in the PMTO condition would show less parenting stress and psychological complaints than parents in the CAU condition. In contrast with our expectations, the results revealed no statistically significant differences in effectiveness between PMTO and CAU on the primary treatment outcome measures of parent-reported externalizing behaviors. That is, children in both conditions showed a significant decrease in CBCL externalizing and PDR scores within the first 6 months of treatment, after which symptom levels remained fairly stable. For parentrated internalizing and total behavior problems, a similar pattern was found: in both 42 Effectiveness of PMTO | Chapter 2 treatment conditions significant decreases were found during the first six months, but no evidence was obtained that children in the PMTO condition fared better than those who received CAU. The fact that internalizing symptoms were also reduced following interventions which essentially target externalizing problems, suggests that either non-specific treatment factors were at work or that both interventions were capable of tackling a process underlying both types of problems. No effects were found for teacher-reported behavior problems. One explanation for this unexpected result might be that children’s behavior problems are less apparent at school and that, therefore, change was less noticeable. Indeed, the data indicated that teachers in general reported less problem behavior as compared to parents. Alternatively, it is also possible that the positive treatment effects did not generalize to the school setting and that the interventions are only effective in the context where they have been implemented (i.e., at home). The finding that PMTO did not result in a greater decrease of externalizing behavior problems than CAU, is in contrast with the results of previous studies showing a superiority of PMTO over control interventions (e.g., DeGarmo & Forgatch, 2005; Forgatch & DeGarmo, 1999; Forgatch et al., 2005a). However, it is important to note that most of the earlier studies that have been conducted in the United States compared PMTO to a waiting list control condition. The families included in the control condition of our study also received a proper treatment, which turned out to be rather effective in reducing children’s externalizing problems. Our findings seem to be more in line with the results of two PMTO effectiveness studies conducted in Norway and Iceland, which also included a control group that received an alternative treatment (Ogden & AmlundHagen, 2008; Sigmarsdóttir et al., 2014). The Norwegian study demonstrated that PMTO initially resulted in a larger decrease in problem behaviors than CAU, but also found that this difference was no longer significant at one-year follow-up (Amlund-Hagen et al., 2011). In the Icelandic study, PMTO produced a better treatment effect than CAU on children’s social skills, but not on behavior problems (Sigmarsdóttir et al., 2014). It is noteworthy that the effect size of CAU in our study was generally larger than that obtained in the other studies, which indicates that the general treatment offerings for children with externalizing problems in The Netherlands appears to be of good quality. This probably is a result of the fact that many psychologists in this country are trained to apply cognitive-behavioral techniques, which seem to be an important ingredient of effective interventions for children with externalizing problems (Greene, Ablon, Goring, Fazio, & Morse, 2004). In addition, PMTO is not the only treatment for externalizing behavior problems in The Netherlands that was not more effective than CAU (e.g., Triple P; Kleefman, Jansen, Stewart, & Reijneveld, 2014; Spijkers, Jansen, & Reijneveld, 2013). 43 2 Chapter 2 | Effectiveness of PMTO Contrary to our expectations, no significant differences between PMTO and CAU were found with regard to the application of effective parenting skills. Only three significant findings on parenting skills emerged. The first one was that parents in both conditions reported a significant increase in self-reported discouragement of undesirable behaviors over time. This suggests that parents in general became more responsive to the misbehaviors of their child. Second, parents reported an increase in their overall use of self-reported effective parenting practices over time. Third, when analyzing the behavioral observation data on parenting behavior, a difference between PMTO and CAU was found for interpersonal atmosphere of the second caregiver. The second caregiver who had received PMTO demonstrated a more positive interpersonal atmosphere over time as compared to the second caregiver who had received CAU. PMTO, as derived from the SIL model, assumes that the reduction of problematic child behavior is mediated by improvements in parenting skills. In particular effective discipline is thought to be an important target mechanism involved in the elimination of child externalizing problems (Ogden & Amlund-Hagen, 2008; Patterson, 2005). Note, however, that this could only be demonstrated with the self-report measure in our study, and this may be due to several reasons. First of all, the observational tasks we used did not elicit particularly high levels of negative behaviors in the child, so parents hardly had to discipline their child during these assessments. Even at baseline, when children were expected to show clear signs of externalizing behavior, the frequency of such problems was less than one out of the seven observation tasks. A second explanation concerns the (un)reliability of the observations. It should be noted that not all parenting scales had satisfactory inter-rater reliability (e.g., discipline). Further, one could argue that the SITs were too well-structured for the oppositional-defiant behavior of the child and the accompanying parenting responses to emerge, which of course questions the ecological validity of our observation measure. Still, it eludes us why our children ‘behaved so well’ during the tasks, because we used tasks very similar to the ones used in the original studies (e.g., Forgatch & DeGarmo, 1999; Forgatch et al., 2005a). One difference is that our SITs were typically administered in a plain room with few distractors, while in the original studies toys and other distractors were available and present in the room. Similar points of critique can be raised regarding the selfreport measure of parenting skills. The internal consistency of five out of six subscales of the CWL was unsatisfactory, and there are data that seriously question the validity of this measure (Thijssen, Muris, & de Ruiter, 2015). Nevertheless, the two reliable scales of the CWL (discouraging undesirable behavior and CWL total score) did show a positive treatment effect. In both conditions, significant reductions of parenting stress and parental psychopathology within the first six months were found, with no significant differences 44 Effectiveness of PMTO | Chapter 2 observed between PMTO and CAU. These results indicate that parents generally felt better as a result of both types of treatment. Apparently, the improvements in their child’s behavior make parents feel less stressed during daily interactions with their child, which may well translate into an overall improved sense of well-being, although the direction of this effect may also be reversed: receiving treatment may boost parental self-efficacy and well-being, which in turn has a positive impact on children’s behavior (e.g., DeGarmo, Patterson, & Forgatch, 2004; Sigmarsdóttir et al., 2013). Not all children profited equally from the PMTO and CAU interventions. A detailed analysis (combining reliable change and clinical cut-off ) indicated that 17% of the children within the PMTO group recovered and 46% showed reliable improvement in externalizing behavior. In comparison, in the CAU condition 10% of the children recovered and 43% reliably improved. To determine if some children benefited more from PMTO than others, several possible moderators were examined. Only one moderator effect was found: children who improved or recovered had significantly higher parent-rated externalizing behavior problems at baseline as compared to children who did not change or worsened. Thus, especially children with serious externalizing behavior problems appeared to benefit more from PMTO. This result is probably due to the fact that there was simply more room for improvement for these children. Possibly, more moderator effects would have been found when using only the recovered and deteriorated children in the comparison. However, in the present study, these subgroups were too small to conduct such analyses. A number of limitations of the present study should be mentioned. First, although the study was originally designed as a RCT, due to practical constraints, we had to continue as a quasi-experimental investigation about halfway through the study. This also resulted in an unequal number of families in the PMTO and CAU conditions. Second, we did not have information about the actual number of treatment sessions that families in both conditions received, and therefore we were not able to control for treatment exposure. Third, as described above, the assessment of parenting practices appeared to be quite problematic, and this appeared true for both the self-report measure (CWL) and the observations (SITs). With regard to the observational index, an additional shortcoming was that coders not always remained blind to treatment condition and time-of-assessment (i.e., T0, T1, T2, T3), because of (unwanted) comments about the treatment made by parents or the assessor during the interaction tasks. In spite of these limitations, we can conclude that a PMTO intervention produced positive effects in a clinically referred sample of children with externalizing problems in the Netherlands. More precisely, this treatment was effective in reducing children’s problem behaviors (even showing a quite large effect size), increasing the use of self-reported effective parenting practices, and reducing parenting stress and psychopathological 45 2 Chapter 2 | Effectiveness of PMTO symptoms of the parents, albeit no more effective than CAU. For both conditions, the improvements were most evident during the first six months of the study and remained stable until 18 months after baseline. Although many effects of the present study were in favor of PMTO and comparable to the effects of PMTO in other European countries, CAU in our study appeared to perform better than the control conditions in most other studies. It is remarkable to note that many of the CAU interventions performed within the Dutch youth care system also include the therapeutic ingredients, such as the use of ‘time out’ for disciplining and rewarding desired behaviors, that are considered important in PMTO. In a future study, the cost-effectiveness of PMTO will be compared to CAU. Annual youth service costs have been rising steadily over the past decade in The Netherlands, and a cost-benefit analysis will provide policy makers and insurance companies with quality information to guide decision-making, in the interest of young children, families and society at large. 46 Treatment fidelity as a determinant of the effectiveness of Parent Management Training Oregon model in The Netherlands 3 Thijssen, J., Albrecht, G., Muris, P., & de Ruiter, C. (2015). Treatment fidelity as a determinant of the effectiveness of Parent Management Training – Oregon model in The Netherlands. Submitted for publication Chapter 3 | Treatment fidelity in PMTO ABSTRACT The present study examined treatment fidelity as a possible determinant of the effectiveness of Parent Management Training – Oregon model (PMTO) in The Netherlands. Clinically referred children (N = 86) aged 4 to 11 years and their parents received PMTO and were assessed at four time points: at baseline, and after 6, 12, and 18 months. Difference scores between baseline and follow-up assessments of externalizing behavior problems, parenting practices, and parental psychopathology and parents’ overall ratings of treatment satisfaction, were correlated with treatment fidelity scores measured prior to the intervention. Furthermore, differences in therapists’ fidelity scores between treatment completers and drop-outs were examined. Results showed that higher fidelity scores of PMTO therapists were associated with larger improvements in externalizing behavior, parenting practices, and parental psychopathology, especially after 18 months. In addition, parents who completed the treatment had a significantly more adherent therapist than families who dropped out. However, the correlations between treatment fidelity and treatment satisfaction was non-significant. These findings indicate that therapists’ high adherence to the PMTO treatment principles decreases the chance of treatment drop-out and positively affects the long-term effectiveness of PMTO. 48 Treatment fidelity in PMTO | Chapter 3 INTRODUCTION Several longitudinal studies have demonstrated that early childhood behavioral problems such as disobedience, fighting, lying, and stealing are predictive of antisocial conduct and delinquency in later life (Coté, Vaillancourt, LeBlanc, Nagin, & Tremblay, 2006; Simonoff et al., 2004). Moreover, the earlier these externalizing problems begin, the higher the risk of serious problems during adulthood (Moffitt, 1993; Patterson, DeBaryshe, & Ramsey, 1989; Reef, Diamantopoulou, van Meurs, Verhulst, & van der Ende, 2011). The fact that externalizing behavior problems in children may have adverse longterm consequences underlines the importance of early intervention programs. Many of these programs focus on the improvement of parenting practices and there is indeed evidence showing that the enhancement of positive and more effective parenting strategies has positive effects on children’s development of prosocial behavior (DeGarmo & Forgatch, 2005, 2007; Forgatch & DeGarmo, 1999; Martinez & Forgatch, 2001; Ogden & Amlund-Hagen, 2008). Parent Management Training - Oregon model (PMTO) is a program especially developed for the parents of children between 4 and 12 years of age with severe externalizing behavior problems such as Oppositional Defiant Disorder or Conduct Disorder. PMTO is based on the Social Interaction Learning (SIL) model, which assumes that contextual factors, such as socio-economic disadvantage and parental psychopathology, have a negative impact on child outcomes by undermining parenting quality (Snyder & Patterson, 1995). Therefore, PMTO focuses on teaching parents how to reduce negative, counterproductive parenting practices and to replace these with positive, more effective parenting practices, such as encouragement (i.e., stimulation of prosocial behaviors in the child, for example by using positive reinforcement), effective discipline (i.e., consistent use of mild sanctions), monitoring (i.e., keeping track of the child’s activities), problem solving (i.e., responding adequately to rule-breaking behaviors and settling arguments with the child), and positive involvement (i.e., giving love and warm attention to the child; Forgatch & DeGarmo, 1999; Forgatch, Patterson, & DeGarmo, 2005). In previous studies, PMTO has been shown to produce positive effects on child externalizing behavior outcomes and parenting practices, and this appears not only true in the United States where the program was developed (e.g., Forgatch & DeGarmo, 1999; Forgatch, Patterson, DeGarmo, & Beldavs, 2009), but also in other countries such as Norway, Iceland, and The Netherlands (Ogden & Amlund-Hagen, 2008; Sigmarsdóttir, Thorlacius, Guðmundsdóttir, & DeGarmo, 2014; Thijssen, Vink, Muris, & de Ruiter, 2015). A logical next step in the process of evaluating PMTO would be to gain more insight into the determinants of the effectiveness of this intervention. In the present study, we investigated to what extent therapists’ adherence to PMTO principles (i.e., treatment 49 3 Chapter 3 | Treatment fidelity in PMTO fidelity) is important for the effectiveness of PMTO. Increased knowledge of possible determinants of PMTO’s effectiveness could be helpful in the further improvement of the treatment method. Prior research has provided tentative evidence indicating that therapists’ close adherence to the treatment protocol is indeed associated with greater improvements in parenting skills and more clear-cut reductions of externalizing behavior problems, and this appeared not only true for PMTO (Forgatch & DeGarmo, 2011; Forgatch et al., 2005; Hukkelberg & Ogden, 2013) but also for other interventions aimed at externalizing behavior problems in children (e.g., Hogue et al., 2008; Huey, Henggeler, Brondino, & Pickrel, 2000). Within PMTO, treatment fidelity is measured by using the Fidelity of Implementation (FIMP) rating system (Knutson, Forgatch, Rains, & Sigmarsdóttir, 2009) which evaluates the therapist on five dimensions: PMTO knowledge, Structure, Teaching, Process skills, and Overall development. The FIMP is a standard part of the PMTO training program to evaluate the progress of PMTO therapists and it is also used for certification and recertification. The present study relies on the data collected for a study evaluating the effectiveness of PMTO in The Netherlands (Thijssen et al., 2015c). In this study, only certified PMTO therapists were involved, which implies that all therapists had at least a sufficient FIMP score. However, therapists may still show considerable variation in the extent to which they adhere to the method, with some of them being really strict at following the guidelines of the intervention and others carrying out the protocol in a looser fashion. The aim of the present study is to examine whether treatment fidelity scores as obtained before treatment (i.e., at certification) are positively associated with treatment outcome in terms of externalizing behavior problems, parenting practices, parental psychopathology, and treatment satisfaction. Furthermore, we will examine whether treatment fidelity was associated with treatment completion. We expected that higher treatment fidelity scores of the therapists would be associated with larger improvements on outcome measures and higher rates of treatment completion. The current study’s procedure was different from previous studies on treatment fidelity in PMTO (Forgatch & DeGarmo, 2011; Forgatch et al., 2005; Hukkelberg & Ogden, 2013) in a number of ways: the present study (a) included multiple outcome measures of PMTO instead of only parenting practices or externalizing behavior; (b) examined the FIMP dimensions separately instead of using a mean or a single construct; and (c) examined the association between fidelity scores and treatment outcome at different assessment points. This new approach could provide new insights into the association between PMTO treatment fidelity and treatment outcomes. 50 Treatment fidelity in PMTO | Chapter 3 METHOD Participants FIMP scores were available for 86 of the 91 families receiving PMTO. Children and their parents were recruited through five child service agencies across The Netherlands. The mean age of these children (62 were boys and 24 were girls) was 7.16 years (SD = 1.81). Mean age of the main caregivers was 38.51 years (SD = 1.27); 25% of them were single parents and 76% had Dutch nationality. Approximately 66% had a college or university degree, 28% had completed high school, and 3% had only finished elementary school. Families were recruited as part of a study evaluating the effectiveness of PMTO in The Netherlands, in which this intervention was compared to care as usual. For obvious reasons, only families who received PMTO were used in the present study. To be included, the child had to reveal a T-score of 60 or higher on the externalizing subscale of the Child Behavior Checklist (CBCL; Achenbach, 1991; Dutch version: Verhulst, van der Ende, & Koot, 1996) and the child had to be residing at home with at least one biological or adoptive parent. Exclusion criteria were severe intellectual disability or psychopathology of the parent(s) (including substance use disorders) that would interfere with participation in treatment, sexual abuse in the family, an IQ of the child lower than 70. Parents in the present study received PMTO from 25 certified therapists whose treatment fidelity scores were determined prior to the start of the treatment. All therapists had completed the full PMTO training program of approximately 24 months. During this training period, therapists had to treat at least three families with PMTO before they were allowed to take part in the official PMTO certification procedure. This procedure involved treating another family with PMTO. On the basis of videotaped sessions of this therapy, it was determined whether or not the therapists received their license to carry out PMTO in clinical practice. Following the completion of the training program, therapists were regularly monitored on their treatment fidelity, leading to annual recertification of their license. Procedure Families were included in the period between June 2009 and January 2014. As soon as families were referred to the child service agency, it was checked whether they met the criteria for being included in the study. Families who met the criteria received information about the study and its procedure and were invited to participate. When parents agreed, they were asked to give their written consent. The study protocol was approved by the Medical Ethics Committee of Maastricht University Medical Centre. 51 3 Chapter 3 | Treatment fidelity in PMTO Families were assessed at four time points: a baseline assessment (T0) and three follow-up measurements, after 6 (T1), 12 (T2), and 18 months (T3). Trained research assistants, who were not the family’s therapist, administered the interview for measuring parenting practices. A web-based system was used for the self-report assessments, which enabled parents to complete questionnaires on a computer at home. In all cases, the main caregiver (i.e., the parent who spent most time with the child) filled out the scales. All participating parents received a small financial compensation in gift vouchers for the three follow-up measurements (i.e., €10 at T1; €20 at T2; and €30 at T3). In PMTO, the therapist works with the parent(s) of one family. The children are not present during these sessions. Role play is an important mechanism in the PMTO sessions to teach and practice the effective parenting practices. In between sessions, the therapist calls the parent to ask about homework, to provide advice and support, and to answer questions. Treatment duration depends on the family’s needs and progress throughout the therapy, but typically lasts between 15 and 25 weekly sessions. Assessments Treatment fidelity Videotapes of four standard therapy sessions, obtained during the certification procedure, were assessed by means of the FIMP rating system (Knutson et al., 2009). This rating system measures the degree of therapists’ adherence to the PMTO treatment model along five dimensions: PMTO knowledge, Structure, Teaching, Process skills, and Overall development, which will be described briefly hereafter. Each dimension is scored on a 9-point scale, in which 1-3 can be considered as ‘unacceptable’ (‘needs work’), 4-6 as ‘acceptable’, and 7-9 as ‘good’. The FIMP ratings were performed by a group of certified FIMP coders (all PMTO educators), both from the US and the Netherlands. All of the FIMP raters have passed the annual reliability tests conducted by the Oregon Social Learning Center. Each session of the same PMTO therapist was rated by a different FIMP coder. In order to ‘pass’, an average rating of 6.0 or higher (as a mean score including all 5 dimensions) was needed. The FIMP dimensions are defined as follows. PMTO knowledge, which reflects the therapist’s understanding of PMTO and SIL principles. Structure refers to balancing several activities during the session, such as following an agenda, maintaining an orderly flow, leading without dominating, responsiveness to family issues, and using sensitive timing and pacing. Teaching comprises proficiency in strategies and tools that promote parents’ mastery and use of PMTO practices. The use of verbal teaching (e.g., giving information, making suggestions, providing rationale) and active teaching (engaging the family in the learning process by brainstorming, prompting, role play, asking solution-eliciting questions) should be balanced. Process skills provide support 52 Treatment fidelity in PMTO | Chapter 3 to create a safe and supportive learning context (e.g., maintaining appropriate balance, encouraging skill development, joining the family’s storyline). And finally, Overall development reflects the accomplishment of session goals and includes, for example, the likelihood that the parents will use the procedures, the parent’s apparent satisfaction, and the likelihood they will continue with the intervention. Externalizing behavior problems The Child Behavior Checklist (CBCL) is a widely used rating scales for assessing behavioral and emotional problems of children aged 6 to 18 years (Achenbach, 1991; Dutch version: Verhulst et al., 1996). Each scale consists of 120 items scored on a 3-point Likert scale (0 = not true; 1 = somewhat or sometimes true; 2 = very or often true). For the present study, only the externalizing scale was used. The CBCL has good reliability and validity (Achenbach, 1991; Verhulst et al., 1996). The Parent Daily Report (PDR) is a reliable 34-item checklist of problem behaviors (Patterson, Chamberlain, & Reid, 1982) and is administered by a research assistant other than the therapist. At first, the checklist is administered face-to-face with the main caregiver to assess whether specific problem behaviors of the child have occurred during the past six months. Next, the items to which the parent responds affirmatively are again administered via telephone on three consecutive days to examine whether these behaviors have occurred during the past 24 hours. The items involve both severe (e.g., arson, stealing) and less serious problem behavior (e.g., being rude, complaining). The scores of the three consecutive days are summed to obtain the total PDR score. Parenting practices The Dutch translation of the Caregiver Wish List (CWL; Hodges, 2005; Hodges, de Ruiter, & Thijssen, 2009) is an interview-based instrument consisting of 53 items questioning the parent about his/her parenting skills. The interviewer reads the questions to the parent, who has to indicate the most applicable response option using a 5-point Likert scale. Items are allocated to six domains of parenting skills: providing direction and following up (4 items), encouraging good behavior (5 items), discouraging undesirable behavior (24 items), monitoring activities (13 items), connecting positively with the child (3 items), and problem solving orientation (4 items). Each domain score should be regarded as a dimension with weak parenting skills on one end and strong parenting skills on the other. In the current study, only the discouraging undesirable behavior domain and the CWL total score were used, since these were the only scales showing adequate internal consistency (.80 and .76, respectively). 53 3 Chapter 3 | Treatment fidelity in PMTO Parental stress and psychopathology The Nijmeegse Ouderlijke Stress Index (NOSI; de Brock, Vermulst, Gerris, & Abidin, 1992) is an adaptation of the Parenting Stress Index (Abidin, 1983) and measures stress experienced by parents in the relationship with their child. The NOSI comprises 123 items that have to be rated on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). Ratings on all items can be summed to create a total stress score, with higher scores reflecting higher levels of perceived stress by the parent. The NOSI has adequate reliability and validity (de Brock et al., 1992). Psychopathological symptoms of the parents were measured by the Dutch version of the Symptom Checklist-90 Revised (Arrindell & Ettema, 2003). The 90 items are rated using a 5-point scale (1 = no problem to 5 = very serious) to indicate the extent to which the parent has experienced the listed symptom during the previous week. In contrast to the original version of the SCL-90-R, the Dutch version comprises eight instead of nine subscales (e.g., Anxiety, Somatic Symptoms, Depression). For the present study, the SCL90-R total score was used, for which higher scores on the SCL-90-R indicate more serious psychopathology. Parent satisfaction The short form of the Working Alliance Inventory (WAI-S; Tracey & Kokotovic, 1989) was used to assess the quality of the parent-therapist alliance. The WAI-S comprises 12 items that can be allocated to three subscales of four items each: (a) agreement between parent and therapist on the goals of the therapy; (b) the degree to which the tasks of the therapy will actually address the parent’s problems; and (c) the quality of the bond between the parent and the therapist. Normally, the items of the WAI-S are rated using a 7-point Likert scale. However, in the present study, a 5-point scale was used for practical reasons. Ten items are positively worded and two items are negatively worded. The scores on the negatively stated items are reversed, so that all scores can be summed to obtain a total score, with higher scores reflecting a better parent-therapist working alliance. Treatment completion The PMTO treatment was considered completed when the core of PMTO parenting skills was dealt with in a satisfactory way (good directions – encouragement – setting limits). Of the 86 PMTO families in the present study, 64 families completed the PMTO treatment. Missing data The proportion of missing values in this data set ranged from 0% in the demographic variables to approximately 50% on the NOSI. We applied an intention-to-treat procedure 54 Treatment fidelity in PMTO | Chapter 3 by conducting multiple imputation (Rubin, 1987) to handle the missing data through a chain of conditional regression models (fully conditional specification; Van Buuren, Boshuizen, & Knook, 1999). We used predictive mean matching (PMM; Little, 1988; Rubin, 1986) for the scale variables and polytomous regression for ordered categorical data. All computations were carried out with Mice (Van Buuren & Groothuis-Oudshoorn, 2011) in R (R Core Team, 2014), with 150 iterations for the algorithm to converge and 25 multiply imputed datasets, using available and custom imputation routines. The outcomes over the 25 datasets were combined into a single inference using Rubin’s (1987) rules. Analytic strategy To examine the association between the FIMP scores and treatment outcome, Pearson correlation coefficients were calculated between the five FIMP dimensions and outcome scores for the questionnaires measuring externalizing behavior problems, parenting skills, parenting stress, psychopathological symptoms of the parents, and treatment satisfaction. Only those measures were used for which the effectiveness study revealed significant effects of PMTO (see Thijssen et al., 2015c). Treatment effects were operationalized in terms of difference scores and calculated for each assessment instrument: scores at baseline (T0) were subtracted from scores at 6 (T1), 12 (T2), and 18 (T3) months. Negative difference scores indicated improvement for externalizing behavior problems, parenting stress, and parental psychopathology. For parenting practices, improvement was indicated by positive difference scores. RESULTS First, we analyzed whether treatment fidelity was related to treatment outcome in terms of externalizing behavior problems. We calculated Pearson correlation coefficients between the five FIMP dimensions and overall FIMP score and difference scores obtained with the CBCL and PDR at the various time points. As can be seen in Table 1, a number of significant associations emerged for the CBCL. For the T3-T0 difference score, significant negative correlations were found for three of the five FIMP dimensions: PMTO knowledge, Teaching, and Overall development. This finding indicates that the higher the score on these FIMP dimensions, the larger the decrease in CBCL externalizing behavior from baseline to 18 months follow-up. Furthermore, a significant negative correlation was found between PMTO knowledge and the CBCL T1-T0 difference score. For the T2-T0 difference score, none of the correlations attained statistical significance and this was also the case for correlations involving PDR difference scores. 55 3 Chapter 3 | Treatment fidelity in PMTO Table 1 | Pearson correlation coefficients between FIMP scores and treatment outcome as indexed by CBCL and PDR difference scores CBCL PDR T1-T0 T2-T0 T3-T0 T1-T0 T2-T0 T3-T0 PMTO knowledge -.22* -.09 -.33** .06 .16 .08 Structure -.17 -.00 -.19 -.07 .06 .01 Teaching -.12 -.03 -.23* -.05 .04 -.01 Process skills -.16 -.04 -.15 -.06 .05 -.01 Overall development -.17 -.07 -.27* .02 .12 .04 Note. FIMP = Fidelity of IMPlementation; CBCL = Child Behavior Checklist; PDR = Parent Daily Report *p < .05 (1-tailed) **p < .01 (1-tailed) Second, we examined the association between treatment fidelity and changes in parenting practices. Pearson correlation coefficients were calculated between the FIMP and the difference scores on CWL Discouraging undesirable behavior domain and the CWL total score. Table 2 shows that only one correlation reached significance for Discouraging undesirable behavior: that is, FIMP Overall development was positively associated with discouraging undesirable behavior from T0 to T3, indicating that the higher the score on Overall development, the more improvement parents reported on Discouraging undesirable behavior. For the CWL total score, for both the T1-T0 and T3-T0 difference scores, significant positive correlations were found for the FIMP dimensions referring to Structure, Teaching, and Overall development. These results show that the higher a PMTO therapist’s scores on these FIMP dimensions, the greater the improvement in parenting skills. Again, no significant correlations were found for the T2-T0 difference score. Table 2 | Pearson correlation coefficients between FIMP scores and treatment outcome as indexed by CWL discouraging undesirable behavior domain and CWL total difference scores Discouraging undesirable behavior T1-T0 T2-T0 T3-T0 PMTO knowledge -.07 -.10 Structure .05 .05 Teaching .04 Process skills .01 Overall development .07 T1-T0 T2-T0 .06 .11 -.06 .09 .14 .27* .06 .21* .08 .18 .21* .10 .24* .02 .10 .16 .03 .15 .09 .21* .24* .12 .26* Note. FIMP = Fidelity of IMPlementation; CWL = Caregiver Wish List *p < .05 (1-tailed) 56 CWL total score T3-T0 Treatment fidelity in PMTO | Chapter 3 Third, we examined the association between treatment fidelity and parental psychopathology. Difference scores of the NOSI and SCL-90-R were correlated with the FIMP dimensions. As shown in Table 3, for the NOSI, a significant negative correlation was found between the T1-T0 difference score and FIMP Process skills. Furthermore, all FIMP dimensions were significantly negatively associated with the NOSI T3-T0 difference score. For the SCL-90-R, a significant correlation was found between the T3T0 difference score and FIMP PMTO knowledge. Thus, a higher score on this dimension is related to a larger decrease in psychopathological symptoms reported by parents. No significant correlations were found for the T1-T0 and T2-T0 difference scores in parental psychopathology. Table 3 | Pearson correlation coefficients between FIMP scores and treatment outcome as indexed by NOSI and SCL-90-R difference scores NOSI SCL-90-R T1-T0 T2-T0 T3-T0 T1-T0 T2-T0 T3-T0 PMTO knowledge -.17 -.11 -.30* -.13 -.10 -.23* Structure -.23 -.11 -.30* -.11 -.13 -.21 Teaching -.18 -.12 -.28* -.07 -.10 -.15 Process skills -.26* -.13 -.30** -.17 -.12 -.18 Overall development -.21 -.16 -.31** -.08 -.13 -.20 Note. FIMP = Fidelity of IMPlementation; NOSI = Parenting Stress Index; SCL-90-R = Symptom Check List-90Revised *p < .05 (1-tailed); **p < .01 (1-tailed) Fourth, we tested whether FIMP scores were related to treatment satisfaction of the parents. Scores on the WAI-S at T1, T2, and T3 were used in the analyses. No significant correlations were found between the FIMP dimensions and the WAI-S scores, implying that FIMP scores were unrelated to treatment satisfaction of the parents. Finally, we examined whether FIMP scores were related to treatment completion. Independent samples t-tests were performed to compare completers and drop-outs on the five FIMP dimensions. Significant differences between treatment completers and drop-outs were documented for all FIMP dimensions (see Table 4). Therapists’ of families who completed the core of the PMTO treatment had mean FIMP scores in the ‘good work’ range on all dimensions, whereas therapists’ of families who dropped out had scores in the ‘acceptable’ range. This indicates that parents were more likely to complete the PMTO treatment when their therapists adhered more closely to the PMTO principles. 57 3 Chapter 3 | Treatment fidelity in PMTO Table 4 | Mean FIMP scores of treatment completers and drop-outs PMTO knowledge completers (n = 64) drop-outs (n = 22) t p 7.10 6.66 2.70 .01 Structure 7.23 6.55 4.19 .00 Teaching 7.03 6.31 4.20 .00 Process skills 7.23 6.70 3.67 .00 Overall development 7.20 6.65 4.09 .00 Note. The maximum score is 9. FIMP = Fidelity of IMPlementation DISCUSSION The aim of the present study was to gain further insight into the role of treatment fidelity as a determinant of the effectiveness of PMTO in The Netherlands. We examined whether treatment fidelity scores obtained for PMTO certification purposes prior to the intervention would be associated with larger treatment effects on various outcome variables, including child externalizing behavior problems, parenting practices, parental psychopathology, parental treatment, and treatment completion. This is the first study that explores the link between treatment fidelity and multiple outcome measures of PMTO. Previous studies on this topic mainly focused on the association between FIMP scores and parenting practices or externalizing behavior problems as the outcome variable (Forgatch & DeGarmo, 2011; Forgatch et al., 2005; Hukkelberg & Ogden, 2013). In line with earlier investigations, the results of the present study provide indications that treatment fidelity is indeed related to greater improvements in child externalizing behavior. The higher therapists’ level of PMTO knowledge the larger the change in externalizing behavior on the CBCL between baseline and 6 months. Furthermore, higher levels of PMTO knowledge, Teaching, and Overall development were related to larger improvements in externalizing behavior at the 18-months follow-up. This points out that higher treatment fidelity scores of the therapist on these dimensions were accompanied by a larger decrease in externalizing behavior on the CBCL, and that this was especially the case at the 18-months follow-up. It is noteworthy that the 18 months follow-up point lies well beyond the PMTO treatment period, which usually lasts between 6 and 9 months. Thus, our findings point to the relevance of treatment fidelity of PMTO’s longer term effectiveness. Remarkably, no significant associations were found between FIMP scores and PDR difference scores. However, the treatment effect based on the PDR (d = .32) is lower compared to the treatment effect based on the CBCL (d = .71). This could explain why no associations were found between the FIMP and PDR scores. 58 Treatment fidelity in PMTO | Chapter 3 With regard to parenting practices, it was found that higher scores on the FIMP dimension Overall development were related to larger increases in the use of discouraging undesirable behavior from baseline to 18-months follow-up. Higher CWL total scores were associated with higher levels of the FIMP dimensions Structure, Teaching, and Overall development from baseline to 6 and 18 months. Our findings imply that higher treatment fidelity is associated with larger improvements in parenting practices and with an increase in discouragement of undesirable behavior in particular. Treatment fidelity was also related to decreases in parenting stress and parental psychopathology. Parents experienced significantly less parenting stress after six months when therapists had higher levels of Process skills. After 18 months, higher scores on all FIMP dimensions were associated with decreased levels of parenting stress. Psychopathological symptoms were significantly lower after 18 months when therapists had higher levels of the FIMP dimension PMTO knowledge. Thus, parents report less parenting-related stress and psychopathological symptoms, especially at 18-months follow-up, when their therapists more strongly adhere to the PMTO program. Meanwhile, treatment fidelity was not associated with greater parental satisfaction, in other words, parents were not more satisfied with their therapist and the intervention when their therapist adhered more strictly to PMTO guidelines. Treatment satisfaction appears independent of how well the therapist delivers the PMTO treatment. Most significant associations were found from baseline to 18-months followup, whereas fewer significant associations could be documented from baseline to 6 months and from baseline to 12 months. This suggests that high treatment fidelity is particularly important for the long-term effects of PMTO. In the short term, it seems to be less important how PMTO is delivered by the therapist. However, for the endurance of the effects of PMTO, it seems to be essential that the therapist strongly adheres to the PMTO method. Note that in the present study all therapists had been certified and therefore all of them had been judged as sufficiently adherent. To achieve optimal longterm treatment effects, it seems to be necessary for PMTO therapists to perform better than merely sufficient. The FIMP dimensions Structure and Overall development were most frequently related to treatment outcome measures and, thus, seem to be the most important dimensions of treatment fidelity. Hence, the therapist should be especially capable of balancing several activities during the session, to increase the likelihood that the parents apply the PMTO parenting strategies, and enabling them to manage unique or difficult situations. PMTO knowledge was mainly related to a decrease in externalizing behavior problems, suggesting that the therapist’s understanding of PMTO and the SIL principles are important for improvement in externalizing behavior problems, while creating balance between several activities during the session seems to be important 59 3 Chapter 3 | Treatment fidelity in PMTO for improvements in parenting practices. The FIMP Process skills dimension was only related to a decrease in parenting stress. Therefore, creating a safe and supportive learning context seems to be important in reducing stress of parents, which might enable them to benefit more from the treatment. Finally, we examined whether FIMP scores differed between therapists of families who completed the treatment and families who dropped out. Parents who completed PMTO more often had a therapist who adhered more strongly to the PMTO principles than parents who dropped out. This implies that PMTO therapists scoring high on treatment fidelity are better able to keep parents in treatment than therapists showing lower treatment fidelity. Previous studies into PMTO treatment fidelity showed that competent adherence to the PMTO principles predicted the degree of change in parenting (i.e., Forgatch & DeGarmo, 2011; Forgatch et al., 2005; Hukkelberg & Ogden, 2013). These previous studies used a different analytical approach, which makes direct comparison with our results difficult. However, two studies reported descriptive statistics, including correlations. In the study of Hukkelberg and Ogden (2013), none of the correlations between the FIMP total score and child problem behavior outcomes were significant. Forgatch and DeGarmo (2011) only found a significant positive correlation between the mean FIMP total score and father’s post-treatment parenting (r = .27). FIMP scores were not significantly related to change in mothers’ and fathers’ parenting. The differences in correlations between the present study and previous studies could be explained by the different approach of the studies. The present study examined more aspects of treatment fidelity than previous studies evaluating treatment fidelity in PMTO. None of the previous studies examined the separate FIMP dimensions. Our study showed that some dimensions are more strongly related to therapy outcome variables than others, which could be useful information for improving the treatment method. Further, in contrast to the previous studies on treatment fidelity in PMTO, the present study investigated the relations between fidelity and effectiveness at different assessment points and showed that treatment fidelity was especially related to longer-term outcomes. Some limitations of the present study should be mentioned. First, the present study was correlational in nature. Obviously, this type of research indicates that there is a relationship between two variables, but cannot prove that one variable causes a change in the other. Second, therapists’ FIMP scores obtained during certification were used; these scores pertained to other families than the families treated in the present study. It is possible that some therapists deteriorated in terms of fidelity while others improved since the time of their certification. As a result, PMTO therapists might have performed differently during certification than during treatment of the PMTO families in the present study. However, correlations between the FIMP scores and outcome 60 Treatment fidelity in PMTO | Chapter 3 measures found in our study are not lower than those reported in previous studies in which FIMP scores were obtained from the same families as the outcome variables (see Forgatch & DeGarmo, 2011; Hukkelberg & Ogden, 2013). Furthermore, Hukkelberg and Ogden (2013) found a high stability in PMTO therapists’ fidelity scores over a period of approximately 9 months (correlations ≥ .30). Therefore, FIMP scores obtained prior to treatment seem to be appropriate as measure of subsequent treatment fidelity. In conclusion, the current study found support for the notion that treatment fidelity in PMTO is related to treatment outcome. The higher the therapist’s fidelity scores, the larger the improvements in externalizing behavior, parenting practices, parenting stress, and parental psychopathological symptoms. This was not translated into higher levels of treatment satisfaction; parents were not more satisfied with their therapist and the intervention when the therapist was more adherent to the protocol. The finding that treatment fidelity was most consistently associated with outcome at the 18-months follow-up assessment indicates that for the long-term effects of PMTO, it is important that PMTO is delivered as intended by the developers and according to the treatment principles in more than merely an acceptable fashion. Furthermore, it was found that parents were more likely to complete the PMTO treatment when their therapist strongly adheres to the PMTO principles. Based on these results, it is advisable to raise the bar for certification, because performing acceptably seems to be insufficient for attaining positive long-term treatment effects with PMTO. 61 3 Initial validation of the Dutch translation of the Caregiver Wish List, an interview-based scale for measuring parenting practices 4 Thijssen, J., Broers, N. J., Muris, P., & de Ruiter, C. (2015). Initial validation of the Dutch translation of the Caregiver Wish List, an interview-based scale for measuring parenting practices. Submitted for publication Chapter 4 | Initial validiation of the Caregiver Wish List ABSTRACT Research has demonstrated that adequate parenting is an important determinant of a healthy social-emotional development in children. There is a great need for valid assessment tools for measuring the quality of parenting, particularly in clinical settings. The Caregiver Wish List (CWL) is a new 53-item interview-based scale for assessing parenting practices. We examined the factor structure of the CWL in a sample of 348 parents of children aged between 4 and 11 years, of which 220 were drawn from the general population and 128 from a clinical setting. Exploratory factor analysis revealed five factors, which did not fully correspond with the hypothesized, original factor structure. Nonetheless, the extracted factors were meaningful and could be labeled as: Adequate discipline, Controlled responding, Focus on positive behavior, Consistency, and Monitoring. The factors demonstrated adequate internal consistency. The factor structures in the community and clinical samples were by and large comparable, which supports the generalizability of the factor structure. Furthermore, the factors differentiated between the community and clinical sample, with better parenting skills observed in the community sample. Finally, all factors were significantly and negatively related to child psychopathology, with stronger correlations demonstrated for externalizing than for internalizing problems. Directions for future research with the CWL are discussed. 64 Initial validiation of the Caregiver Wish List | Chapter 4 INTRODUCTION Ineffective parenting is the most widely studied and empirically established risk factor for externalizing behavior problems in children (DeBaryshe, Patterson, & Capaldi, 1993; Forgatch, Bullock, & Patterson, 2004; McCoy, Frick, Loney, & Ellis, 1999; Nix et al., 1999; Oliver, Guerin, & Coffman, 2009; Patterson, Forgatch, Yoerger, & Stoolmiller, 1998; Sharma & Sandhu, 2006; Williams et al., 2009). In particular, harsh and inconsistent discipline, poor monitoring and supervision, and low positive involvement have all been demonstrated to be significantly associated with young people’s oppositional-defiant and conduct problems (Bierman & Smoot, 1991; Cunningham & Boyle, 2002; Ehrensaft et al., 2003; Frick et al., 1992; Nicholson, Fox, & Johnson, 2005). Many interventions for childhood behavior problems focus on increasing effective parenting and decreasing the use of ineffective rearing practices. Research generally supports the theoretical notion that improving parenting practices is the most important mechanism contributing to a decrease in children’s behavior problems (DeGarmo & Forgatch, 2005, 2007; Forgatch & DeGarmo, 1999; Kazdin, 2007; Martinez & Forgatch, 2001; Ogden & Amlund-Hagen, 2008). Improvements in the behavior problems of the children also tend to reduce stress and depression in parents, which in turn will increase the likelihood that the effective parenting practices will be maintained, thereby promoting the long-term effects of therapy (DeGarmo, Patterson, & Forgatch, 2004). Many parent training programs are based on the Social Interaction Learning (SIL) model (Forgatch et al., 2004). The SIL model assumes that the psychosocial development of the child is directly influenced by the parents and their parenting strategies. Contextual factors, however, can have a negative impact on the parenting quality, and thus, indirectly on the child. For example, when parents are going through a divorce, this puts pressure on the parenting quality (Conger, Patterson, & Ge, 1995; Hetherington, Bridges, & Insabella, 1998). According to the SIL model, there are five core effective parenting practices: skill encouragement, discipline, monitoring, problem solving, and positive involvement. Skill encouragement refers to the enhancement of a prosocial development by using scaffolding techniques (e.g., breaking behavior into small steps, prompting appropriate behavior) and the provision of positive reinforcement (e.g., giving compliments when the child behaves well; Forgatch & DeGarmo, 1999). Discipline decreases problem behaviors by the appropriate and consistent use of mild sanctions, like giving a time out or taking away privileges (Patterson, 1986). Monitoring involves knowing the child’s friends and keeping track of the child’s activities to protect him/her against the negative influence of deviant peers (Snyder, 2002). Problem solving helps families to negotiate during arguments, to clearly determine rules and consequences when the rules are violated (Forgatch, Patterson, & DeGarmo, 2005). And finally, positive involvement concerns the many ways in which parents give their child loving attention 65 4 Chapter 4 | Initial validiation of the Caregiver Wish List and engagement in joint fun activities (Forgatch & DeGarmo, 1999; Forgatch et al., 2005). Many studies rely on observations of parent-child interactions as a method to assess parenting skills (Martinez & Forgatch, 2001; Ogden & Amlund-Hagen, 2008; Patterson, DeGarmo, & Forgatch, 2004). For instance, parents and child can be observed during a structured playroom session or in a natural situation at home or in school. Self-report measures can be used as an alternative assessment method which complements the observation measures of parenting. In fact, there are indications that a multimethod assessment of parenting possesses incremental validity over using only one measurement method (Harvey, Danforth, Ulaszek, & Eberhardt, 2001; Meyer et al., 2001). There are many self-report measures that can be used for the assessment of parenting, although there is a wide variety in terms of definition and theoretical framework. The Caregiver Wish List (CWL; Hodges, 2002) is a structured interview-based instrument that enables parents and other caregivers to reflect on their own parenting practices and talk about their experiences with parenting. Interestingly, the CWL was designed to measure the core parenting practices of the SIL model and seems to be particularly useful in clinical practice for a number of reasons. First, the CWL may help clinicians to gather information on the strengths and weaknesses in parenting practices as perceived by parents themselves. Furthermore, the CWL can be expected to enhance the therapeutic alliance and the positive engagement between parents and the clinician. Finally, the CWL may assist in identifying and clarifying the goals of treatment. The goal of the CWL is to empower parents by encouraging them to see themselves as the main agent of change for their child (Hodges, 2005). The CWL contains six a priori domains of parenting skills: providing direction and following up, encouraging good behavior, discouraging undesirable behavior, monitoring activities, connecting positively with the child, and problem solving orientation. In the Netherlands, the CWL is being used as an intake assessment for Parent Management Training - Oregon model (PMTO), a parent training program that teaches parents effective parenting skills according to the SIL model with the aim of reducing externalizing behavior problems in children aged 4 to 12 years. This makes sense as the CWL directly measures the parenting skills that are the main focus of change in PMTO. A Dutch translation of the CWL was also employed in a large-scale study evaluating the effectiveness of PMTO in The Netherlands (Thijssen, Vink, Muris, & de Ruiter, 2015). However, in that study, it was found that the reliability of the CWL was rather poor, with only one of the six original parenting domains showing adequate internal consistency. In fact, to our knowledge, no study can be found that examined the psychometric properties of the CWL. The aim of the present study was a first attempt to fill this gap. The CWL was administered in a large sample of parents of children from the community as well as a clinical setting visited by parents of whom children displayed disruptive 66 Initial validiation of the Caregiver Wish List | Chapter 4 behavior problems. In this way, it became possible to explore the factor structure of the instrument as well as its reliability (internal consistency and correlations between factors) and validity (relations between parenting scores and children’s psychopathological symptoms). METHOD Participants The total sample consisted of 348 parents (92.1% mothers) of 4- to 11-year-old children (mean age = 8.07 years, SD = 1.61). Of these children, 185 (53.2%) were boys and 163 (46.8%) were girls. The mean age of the main caregiver was 39.06 years (SD = 5.15). The majority of the main caregivers was the biological parent of the child (92.8%), had the Dutch nationality (89.1%), and was employed (72.4%). Of the 348 main caregivers, 222 (63.8%) were living with a partner. Seventy-three percent of the main caregivers had a college or university degree, 17.9% had completed high school, while 1.1% had only finished elementary school. In Table 1, the demographic characteristics are reported for the community and clinical sample separately. In comparison to the clinical sample, children and parents in the community sample were significantly older [t(346) = 7.94, p < .01 and t(327) = 4.95, p < .01, respectively]. Moreover, the community sample contained more girls [χ2(1) = 16.00, p < .01] and more two-parent households [χ2(1) = 8.57, p < .01] than the clinical sample. Table 1 | Demographic data and significant differences for the community and clinical sample Community (n = 220) M (SD) or % Clinical (n = 128) M (SD) or % p Child age 8.58 (1.27) 7.18 (1.75) .00 Parent age 40.11 (4.42) 37.00 (5.83) .00 Child is a boy 45.0 67.2 .00 Main caregiver is female 93.6 89.1 .21 Biological parent 96.8 98.2 .64 Two parent household 89.1 77.0 .00 Dutch nationality 92.7 94.6 .74 Employment 77.7 76.4 .79 67 4 Chapter 4 | Initial validiation of the Caregiver Wish List Procedure The parents of the community sample were recruited in elementary schools where information letters explaining the nature of the study and consent forms were distributed, together with a brief survey to obtain background information on the parents (e.g., ethnicity, educational level, employment status). When parents agreed to participate, they signed the consent form, filled out the demographic characteristics survey, and returned these materials to the researchers via the child’s teacher at school. The parent who spent most time with the child (main caregiver) was contacted to make an appointment for the CWL interview. The CWL interview was administered at the child’s school and lasted for about 30 minutes. As a reward, parents received a €10 gift voucher. This study was approved by the Ethical Committee of the Faculty of Psychology and Neuroscience at Maastricht University. The parents of the clinical sample were recruited as part of a study on the effectiveness of PMTO in children with disruptive behavior problems. As soon as families were referred to the child service agency, it was checked whether they met the inclusion criteria for the study (for details, see Thijssen et al., 2015c). Families who met the criteria received information about the study and its procedure and were invited to participate. When parents agreed, they were asked to give their written consent. The CWL interview was completed by the main caregiver and mostly took place at the child service agency. For the present study, only the CWL data that were obtained prior to the treatment were employed. Instruments As noted in the introduction, the Caregiver Wish List (CWL; Hodges, 2005; Hodges, de Ruiter, & Thijssen, 2009) is an interview-based instrument consisting of 53 items questioning the parent about his/her parenting skills. The interviewer reads the questions to the parent, who has to indicate the most applicable response option using a 5-point Likert scale. The response options are specific to the question, although they mostly refer to the frequency of behaviors. The parent also has the option to respond with ‘not applicable’ when the described situation does not apply to their family. Twenty-nine items are positively formulated, whereas 24 items are negatively phrased. The latter items are reversed so that higher CWL scores reflect higher levels of effective parenting. On the basis of conceptual work by Hodges (2002), items can be allocated to six a priori domains of parenting skills: providing direction and following up (4 items), encouraging good behavior (5 items), discouraging undesirable behavior (24 items), monitoring activities (13 items), connecting positively with child (3 items), and problem solving orientation (4 items). Each domain score can be regarded as a dimension with weak parenting skills on one end and strong parenting skills on the other. 68 Initial validiation of the Caregiver Wish List | Chapter 4 The Child Behavior Checklist (CBCL) is a widely used rating scales for assessing behavioral and emotional problems of children aged 6 to 18 years (Achenbach, 1991; Dutch version: Verhulst, van der Ende, & Koot, 1996). Each scale consists of 120 items scored on a 3-point Likert scale (0 = not true; 1 = somewhat or sometimes true; 2 = very or often true). For the present study, only the internalizing and externalizing scales were used. The CBCL has good reliability and validity (Achenbach, 1991; Verhulst et al., 1996). Data analyses Since no research on the factor structure of the CWL has been published, a confirmatory factor analysis was considered as less appropriate. Instead, we performed an exploratory factor analysis (EFA), using the pre-imposed conceptual structure as a guide in the process of exploration. Exploratory factor analyses were conducted with Mplus (version 5.21), using robust maximum likelihood estimation (MLR) to correct for the non-normality in the distribution of the items. An oblique rotation method (geomin) was applied as it seemed plausible that the extracted factors of positive parenting would be correlated. Unlike principal components analysis, EFA does not permit the quantification of the proportion of common variance explained. For an alternative indication of the quality of our factor solution, we decided to examine a few model fit indices that are usually reported in the context of confirmatory factor analysis: the Tucker Lewis Index (TLI), the Root Mean Square Error of Approximation (RMSEA) and the Standardized Root Means Square Residual (SRMR). These fit indices give complementary information. Although they should be considered with caution when conducting EFA, they provide a useful impression on the quality of the factor solution. Fit values that would indicate an accurate description of the original correlation matrix would be a TLI value larger than 0.95, a RMSEA value smaller than 0.05, and an SRMR value smaller than 0.08 (Hu & Bentler, 1999). Even though the separate samples were relatively small in comparison to the number of items to be processed in the factor analysis, we decided to initially explore the factor structure for each of the samples separately. Using smaller samples in the factor analysis will increase the instability of the solution, but prevents possible confounding because of using very heterogeneous populations. Only in case of sufficient similarity in the factor solutions for the separate samples, it seemed justified to interpret the analysis for both samples combined. To examine whether the CWL factors were significantly able to discriminate between the community and clinical sample, independent samples t-tests were performed. The internal consistency of the extracted factors was examined using Cronbach’s alpha. Convergent validity was assessed by means of correlational tests between the CWL factors and CBCL. An adapted version of Steiger’s (1980) formula was used to conduct 69 4 Chapter 4 | Initial validiation of the Caregiver Wish List tests for comparing correlation coefficients (Lee & Preacher, 2013), in order to examine whether there were significant differences in the correlations between CWL factors and internalizing and externalizing problems. RESULTS Preliminary considerations Before conducting the factor analysis, some preliminary decisions had to be made. First, the issue of ‘not applicable’ responses in the CWL had to be resolved. As noted earlier, parents had the option to select this response in case a given situation did not occur in daily life. We decided to treat these ‘not applicable’ responses as missing values. As we used robust maximum likelihood estimation for our explorative factor analysis, all cases with missing values on some of the items could be retained in the analysis. Meanwhile, items that generated a relatively high (> 10%) percentage of ‘not applicable’ responses were removed from the analysis. The threshold value of 10% was chosen because until this percentage is reached, most of the traditional methods for dealing with missing values produce fairly similar results (e.g., see Barzi & Woodward, 2004). Based on this criterion, three items were removed. Second, we screened all item distributions in order to identify items that were excessively skewed. Items for which a score of 4 or 5 was given in 95% of the cases in one sample and at least 90% of the cases in the other, were excluded from the analyses, because it can be assumed that these items had little discriminatory power. There were six items showing such extreme distributions in both samples, which were therefore excluded. Further, we noticed a number of items that had conspicuously low mean scores, indicating that the pertinent behaviors were hardly practiced by the parents. One would expect that the scores on parenting skills in the community sample would at least be ‘neutral’ (i.e., somewhere around the rating of 3 or higher). To be conservative, we excluded all items with a mean value lower than 2.50 in the community sample. Obviously, a relatively low item mean in the community sample does not necessarily invalidate the item as a diagnostic marker when the corresponding item mean in the clinical sample is still significantly lower. However, the items that had a mean value lower than 2.50 in the community sample all had higher means in the clinical sample, with the exception of one item. This item had a lower mean in the clinical sample, but did not differ significantly from the mean in the community sample. We therefore decided to exclude four items, because they seemed to represent behaviors that were hardly practiced by parents in The Netherlands. In total, 13 items were removed before conducting the EFA. 70 Initial validiation of the Caregiver Wish List | Chapter 4 Exploratory Factor Analysis The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy showed acceptable values for both the community sample (KMO = 0.67) and the clinical sample (KMO = 0.64). The initial attempt to conduct EFA for the community sample failed to reach convergence in the estimation of parameter values. Inspection of the output provided, showed conspicuous estimates of means and covariances involving four items. Although failure to reach a convergent solution may have several causes, the most likely explanation seemed related to low initial communality values for these four items. We therefore decided to remove these items from the analysis, after which no further problems in convergence occurred. The EFA in the community sample was then rerun with the remaining 36 items and also performed for the clinical sample. Based on the Kaiser criterion (eigenvalues > 1), 12 factors were identified for both the community and clinical sample. Obviously, this was not a useful criterion. Inspection of the scree plot suggested a five-, six-, or a seven-factor solution. Inspection of the fit indices for both the community and clinical sample pointed in the direction of either a five- or a six-factor solution for both samples. The sixth factor was difficult to interpret and, therefore, a five-factor solution seemed most appropriate. This resulted in wellinterpretable factors with acceptable RMSEA and SRMR values, but rather low TLI values (see Table 2). Since the pattern of factor loadings was quite comparable in both samples, we decided to rerun the analysis for both samples combined. The resulting five-factor solution produced good fit values for the RMSEA and the SRMR measures, while the TLI improved but was still below the acceptable limit. Table 2 | Fit indices of the five-factor solution for the community, clinical, total sample, and final factor solution based on the included items of the total sample Fit index TLI RMSEA SRMR Community sample .50 .069 .052 Clinical sample .54 .074 .061 Total sample .78 .049 .042 Final factor solution total sample .87 .050 .035 Note. TLI = Tucker Lewis Index; RMSEA = Root Mean Square Error of Approximation; SRMR = Standardized Root Means Square Residual 71 4 Chapter 4 | Initial validiation of the Caregiver Wish List Table 3 | Factor loading for items on the Caregiver Wish List for the community, clinical, and total sample Items Community Clinical sample sample Total sample Factor 1: Adequate discipline .434 .418 .487 .580 .805 .668 .696 .330 .597 .749 .444 .630 .408 .632 .528 When you tell your child what to do, how often are you in an unhappy, mad or frustrated mood? 25 When your child misbehaves, how often do you get angry? .450 .396 .422 .740 .775 .849 26 When your child misbehaves, how often do you yell? .678 .746 .708 29 When you correct your child/give a consequence — and your child reacts badly (yelling back, refusing) — how often do you get angry and let it show? .375 .581 .557 .457 .231 .514 .490 .700 .444 .459 .405 .513 .529 .410 .437 .406 .698 .518 .070 .479 .374 .499 .673 .643 .550 .787 .689 .394 .431 .354 10 When your child misbehaves, how often do you choose to give no consequences because you think that your child is too stressed? 11 When your child misbehaves, how often do you choose to give no consequences because you think your child may have a temper tantrum? 13 When your child misbehaves, how often do you choose to give no consequences because you think your child may feel less loved? 14 When your child misbehaves, how often do you choose to give no consequences because you don’t want to lose your temper? 15 When your child misbehaves, how often do you choose to give no consequences because it would cause a problem for you (for example, you would have to stay home)? Factor 2: Controlled responding 3 Factor 3: Focus on positive behavior 2 Which do you do more—telling your child what NOT to do (STOP doing a bad behavior) versus telling your child what TO DO (start doing a good behavior)? 5 When your child behaves well, how often do you praise or compliment your child? 17 What do you do most—praise your child for good behavior or correct your child for bad behavior? 49 How often are you able to express positive feelings toward your child? Factor 4: Consistency 16 When your child misbehaves, how often do you choose to give NO consequences because it is hard for you to be strict? 18 When your child misbehaves, how often do you do nothing, because nothing seems to work? 20 When your child misbehaves, how often do you say you will give consequences, but don’t follow through? 30 When you correct your child/give a consequence — and your child reacts badly (yelling back, refusing) — how often do you give your child what he/she wants? 31 When you correct your child/give a consequence — and your child reacts badly (yelling back, refusing) — how often do you give up what you wanted (for example, a cleaned-up room)? 72 Initial validiation of the Caregiver Wish List | Chapter 4 Table 3 | Factor loading for items on the Caregiver Wish List for the community, clinical, and total sample Items Community Clinical sample sample Total sample Factor 4: Consistency 32 When you correct your child/give a consequence — and your child reacts badly (yelling back, refusing) — how often do you give additional consequences, but NOT follow through? .382 .458 .457 .778 .597 .668 .765 .905 .907 .601 .634 .583 .351 .411 .347 Factor 5: Monitoring 41 How much do you know about your child’s playmates, friends, or kids he/she hangs out with? 42 How much do you know about the families of these kids — like where they live, their phone number, and what their parents do? 44 When your child goes out, how much do you know about who he/ she will be with? 45 When your child goes out, how much do you know about what he/ she is doing and where he/she is going? 4 Note. Boldfaced factor loadings are significant at p < .05 In the five-factor solution for the total sample, 13 items had non-significant factor loadings or factor loadings less than .40 on all factors and were, therefore, not included in the final solution. The final factor loadings for each included item in the community, clinical, and total sample are shown in Table 3. Factor 1 consisted of five items and was named Adequate discipline because all items reflect disciplining behaviors of parents after misbehavior of their child. All five items were from the original ‘discouraging undesirable behavior’domain. Factor 2 encompassed four items that referred to Controlled responding when the child misbehaves (e.g., calm, neutral tone of voice). Again, all items belonged to the original ‘discouraging undesirable behavior’ scale, except for item 3 which came from the original scale ‘providing direction and following up’. Factor 3 contained four items and was labelled Focus on positive behavior. These items were all from different original domains, but all appeared to tap parents’ responsiveness to positive, prosocial behavior of the child. Factor 4 was composed of six items reflecting parents following through after disciplining their child. Therefore, this factor was labeled Consistency. All items were from the original ‘discouraging undesirable behavior’ domain. The final and fifth factor was named Monitoring and consisted of four items which were all part of the original ‘monitoring activities’ scale. Items 17 and 45 had substantial secondary loadings, but based on the highest loadings they were assigned to Factor 3 and Factor 5, respectively. 73 Chapter 4 | Initial validiation of the Caregiver Wish List Reliability The internal consistency coefficients of the five factors were calculated for the community, clinical, and total sample. As can be seen in Table 4, all Cronbach’s alphas were in the .61 to .82 range. When adopting an alpha of .70 as cut-off point, it can be concluded that in the clinical and total sample four of the five factors displayed good internal consistency. The CWL total score had a good internal consistency in both the community and clinical sample, as well as in the total sample. In the community sample, Cronbach’s alphas were in general somewhat lower: only the factors Adequate discipline and Monitoring had adequate Cronbach’s alphas. Note, however, that all CWL factors consisted of a small set of items, and that alpha values above .60 for a limited item set could still be regarded as acceptable. Table 4 | Internal consistencies (Cronbach’s α) of the extracted factors of the community, clinical, and total sample Community sample Clinical sample Total sample Adequate discipline .70 .77 .76 Controlled responding .63 .71 .71 Focus on positive behavior .64 .65 .66 Consistency .61 .72 .73 Monitoring .75 .73 .76 Total score .71 .78 .82 Pearson product moment correlations among the five factors were calculated for the two samples separately and for the total sample (Table 5). Overall, the correlations were quite low. Two correlations were significant in all three samples, namely the correlation between Controlled responding and Focus on positive behavior and between Adequate discipline and Consistency. These correlations were positive, which indicates that parents who were well able to control their emotions during limit setting were also good in focusing on positive behavior of the child, and vice versa. Furthermore, parents who did not allow other factors to influence their limit setting were also more consistent, and vice versa. It was expected that especially the three factors derived from the original ‘discouraging undesirable behavior’ domain (i.e., Adequate discipline, Controlled responding, and Consistency) would be correlated, but this was only true for the total sample. 74 Initial validiation of the Caregiver Wish List | Chapter 4 Table 5 | Pearson correlations among the five extracted factors for the community, clinical, and total sample 1 2 3 4 5 Community sample 1. Adequate discipline 1.00 2. Controlled responding .03 3. Focus on positive behavior .03 .26** 1.00 4. Consistency .41** .13 .11 1.00 5. Monitoring .14* -.06 .17* .01 1.00 4 1.00 Clinical sample 1. Adequate discipline 1.00 2. Controlled responding .09 1.00 3. Focus on positive behavior -.09 .47** 1.00 4. Consistency .39** .27** .14 1.00 5. Monitoring .05 .09 .04 .14 1.00 Total sample 1. Adequate discipline 1.00 2. Controlled responding .16** 1.00 3. Focus on positive behavior .09 .44** 1.00 4. Consistency .47** .32** .26** 5. Monitoring .15 .10 .18** 1.00 .17** 1.00 * p < .05; ** p < .01 External validity As stated before, the pattern of factor ladings was roughly similar across the two samples, motivating us to re-analyze the data on the basis of the total sample, thus reducing instability in the parameter estimates. Table 6 shows the mean scores and standard deviations on the factors for both the community and clinical sample. The difference between both samples was significant for all factors and the CWL total score (with 23 items included in the factors). On all factors, parents from the community sample had a significantly higher mean than parents from the clinical sample, indicating that the parenting skills of parents in the community sample were better than those of parents in the clinical sample. 75 Chapter 4 | Initial validiation of the Caregiver Wish List Table 6 | Means (standard deviations) of the community and clinical sample on the extracted factors and their significant differences Community sample M (SD) Clinical sample M (SD) t p Adequate discipline 22.77 (2.98) 20.33 (4.79) 5.19 .00 Controlled responding 13.63 (2.90) 11.09 (2.85) 7.89 .00 Focus on positive behavior 15.78 (2.56) 13.62 (2.57) 7.58 .00 Consistency 27.23 (2.80) 23.83 (4.95) 7.11 .00 Monitoring 17.22 (2.07) 16.01 (3.15) 3.90 .00 Total score 96.62 (7.38) 84.85 (10.86) 10.82 .00 Pearson product moment correlations between the five CWL factors and CBCL internalizing and externalizing subscales were calculated. As can be seen in Table 7, all CWL factors and the total score were significantly and negatively correlated with both CBCL internalizing and externalizing scales. Relatively low, but still significant, negative correlations were found for the CWL factor Monitoring. Compared to the correlations for the CBCL internalizing scale, the correlations for externalizing were higher, even significantly higher for three of the five CWL factors, namely Controlled responding, Focus on positive behavior, and Consistency, as well as for the CWL total score (see Table 7). This result was as expected, since the assessed parenting skills are considered to be important for improving externalizing behavior problems of the child. Table 7 | Pearson correlations between the five extracted CWL factors and CBCL internalizing and externalizing scores and the results of tests comparing the strength of these correlations Adequate discipline CBCL Internalizing CBCL Externalizing p -.24 -.30 .13 Controlled responding -.26 -.41 .00 Focus on positive behavior -.29 -.40 .02 Consistency -.23 -.34 .02 Monitoring -.19 -.18 .43 Total score -.38 -.52 .00 Note. All correlations were significant at p < .01. CWL = Caregiver Wish List; CBCL = Child Behavior Checklist DISCUSSION This study was a first attempt to investigate the psychometric qualities of the CWL, an interview-based scale for measuring parenting behaviors as defined in the SIL model (Forgatch et al., 2004). EFA was performed to examine the factor structure of the CWL using the data of parents from both a community and a clinical sample of elementary 76 Initial validiation of the Caregiver Wish List | Chapter 4 school age children. Furthermore, the internal reliability of the extracted factors was investigated and we examined whether the CWL factors showed theoretically meaningful associations with behavior problems of the parents’ children and whether they were able to differentiate between the community and clinical sample. The EFA revealed five factors, which were named Adequate discipline, Controlled responding, Focus on positive behavior, Consistency, and Monitoring. The five-factor structure provided a reasonable description of the original correlation matrix, according to two of the conventional fit indices. However, on one fit index, the fit was quite poor. The factor Adequate discipline measures disciplining behavior of the parents when their child misbehaves. Controlled responding assesses parents’ responding to misbehavior of their child. For example, whether the parent stays calm and responds in a neutral tone of voice. The factor Focus on positive behavior relates to parents’ ability to note and reinforce desirable behaviors of their child. Consistency reflects the tendency of parents to follow-through on their expressed consequences, despite aversive responses of their child to the sanction. Finally, the Monitoring factor assesses parents’ knowledge of their child’s whereabouts and friendships. The extracted factors are only partially in line with the domains proposed by the developer of the CWL. Hodges (2002) had in mind that the CWL would be composed of six domains of parenting, whereas in the present study only five factors were found. Three of these five factors, viz. Adequate discipline, Controlled responding, and Consistency, were entirely composed of items from the original ‘discouraging undesirable behavior’ domain. This finding was hardly surprising as the original ‘discouraging undesirable behavior’ domain contained about half of all original CWL items which included a rather heterogeneous set of parenting behaviors all targeting unwanted behaviors of the child. The factor Monitoring was in keeping with the original ‘monitoring activities’ domain, but contained only four of the original 13 items. The factor Focus on positive behavior contained items from several CWL domains. The only domain that did not emerge in the present factor structure was Problem solving. The four original Problem solving items were discarded during the EFA procedure, indicating that problems solving did not emerge as a separate and homogeneous parenting factor. Altogether, the five factors found for the CWL in the present study correspond to four of the five core parenting practices of the SIL model (i.e., effective discipline, encouragement, positive involvement, monitoring). Only 23 of the 53 original CWL items loaded on at least one of the five factors, indicating that the original CWL can probably be reduced in length without losing critical content. Despite the limited number of items included in the five factors, the factors showed adequate internal consistencies, with Cronbach’s alphas ranging between .61 and .77. The CWL total score, based on the items included in the factors, 77 4 Chapter 4 | Initial validiation of the Caregiver Wish List had good internal consistency coefficients in the three samples (.71 - .82). Only two correlations among CWL factors were consistently found in the community, clinical, and total sample, namely those between Controlled responding and Focus on positive behavior and between Adequate discipline and Consistency. The correlations between the other factors were quite low. Especially the correlation between Adequate discipline and Controlled responding was unexpectedly low, given the fact that these two factors were derived from the same original CWL domain (i.e., discouraging undesirable behavior). The factor structure of the community and clinical samples was highly similar, and thus the extracted CWL factor structure seemed to be applicable to parents in both populations. Meanwhile, it was found that the two samples differed in terms of absolute scores on the five extracted CWL factors and the total score in the expected direction: parents in the community sample exhibited higher scores and hence better parenting skills than parents in the clinical sample. Finally, all factors and the CWL total score were significantly negatively related to internalizing and externalizing behavior of the child: the better the parenting practices of the parents, the lower the behavior problems. The factors Controlled responding, Focus on positive behavior, and Consistency were significantly stronger associated with externalizing problems than with internalizing symptoms. Since the CWL measures parenting practices that are assumed to be especially relevant within the context of externalizing behavior problems, these results can be taken as support for the external validity of this parenting measure. A few limitations of our study have to be noted. First, the majority of the families consisted of two biological Dutch parents with steady employment, so it is unknown to what extent our findings can be generalized to, for example, single parents, stepfamilies or parents from other ethnic and lower socioeconomic backgrounds. Second, the sample size was quite small for factor analysis when considering the number of items included in the original CWL. The criterion for factor analysis is that there should be at least five cases per item (Hair, Black, Babin, & Anderson, 2009). The CWL contains 53 items, which means that at least 246 participants were needed to meet this criterion. However, before the final factor analysis was run, 17 items needed to be excluded based on various criteria, which means that only 36 items were used in the analysis. Thus, a sample of 180 participants was sufficient. The community sample and the total sample met this criterion, but the clinical sample (n = 128) did not. Third, there was a difference in the frequency of endorsement of ‘not applicable’-responses between the community and the clinical sample. The clinical sample showed more not applicable-responses in comparison to the community sample. Two possible explanations could be given for this. It could be that certain parenting situations are less common in a clinical sample than in a community sample (for example, giving allowance) or it could be related to the person 78 Initial validiation of the Caregiver Wish List | Chapter 4 who interviewed the parents. Parents from the community sample were interviewed by trained students who were specifically instructed to avoid not-applicable responses and to stimulate parents to give the best suitable answer. Parents from the clinical sample were interviewed by trained clinicians who received the same instructions. However, these parents were recruited and interviewed over a longer period of time, which might have caused drift in the application of these instructions. Although the CWL was not specifically developed as an outcome measure for research on treatment effectiveness, this instrument could be used to examine change in the specific parenting practices of the SIL model. However, the factor structure found in the present study should be replicated before more definitive conclusions about the instrument’s structural validity can be drawn. Furthermore, it needs to be examined if the extracted factors of the CWL are sensitive to change. In our study evaluating the effectiveness of Parent Management Training – Oregon model, in which the CWL was used as an outcome measure, only the original ‘discouraging undesirable behavior’ domain and the CWL total score had adequate internal consistencies and these scales showed significant change in response to parent training (Thijssen et al., 2015c). Additionally, future research should explore whether the parenting practices assessed by the CWL predict the development or maintenance of externalizing behavior problems in children. The association between CWL self-reported parenting practices and parenting assessed on the basis of actual parent-child interactions studied in real life or in the lab should also be examined. In conclusion, findings from our study suggest that the CWL in its current form needs to be amended to render it more suitable for research purposes. The a priori factor structure could not be replicated and several of the extracted factors need additional items to increase the CWL’s psychometric properties. Furthermore, although the CWL was designed to assess parenting skills defined by Social Interaction Learning theory, it failed to assess Problem solving as one of the main SIL skills. Notwithstanding, the CWL’s interview format provides opportunities for rapport building between parent and interviewer, and for reflection on parenting skills in terms of strengths and vulnerabilities. 79 4 Callousunemotional t r a i t s PART II Emotional memory for central and peripheral details in children with callous- unemotional traits 5 Thijssen, J., Otgaar, H., Meijer, E.H., Smeets, T., & de Ruiter, C. (2012). Emotional memory for central and peripheral details in children with callous-unemotional traits. Behavioral Sciences and the Law, 30, 506-515. doi: 10.1002/bsl.2021 Chapter 5 | Emotional memory in children with CU traits ABSTRACT A limited number of studies have shown that adults and adolescents with psychopathic traits suffer from emotional memory impairment. The present study examined whether this finding could be replicated in a sample of children between 8 and 12 years of age with callous-unemotional (CU) traits. Children with high CU traits (n=24) were compared with children with low CU traits (n=18) with regard to performance on a peripheral memory recognition test that examined memory for central and peripheral components of neutral and negative pictures. Results showed that overall recognition rates did not differ between the high- and low-CU groups. For negative pictures, both groups demonstrated better recognition of the central component at the expense of the peripheral component, while for neutral pictures, the peripheral component was better recognized than the central component. This study is the first to demonstrate that children with high CU traits do not suffer from an impaired emotional memory. 84 Emotional memory in children with CU traits | Chapter 5 INTRODUCTION According to the DSM-IV-TR (American Psychiatric Association, 2000), conduct disorder (CD) refers to children’s and adolescents’ recurrent and persistent rule-breaking behavior in which the basic rights of others or major age-appropriate societal norms or rules are violated. Children with CD, for example, tend to frequently steal, vandalize, lie, and cause physical harm to other people or animals. A subset of children with CD is characterized by so-called callous-unemotional (CU) personality traits, which closely resemble the emotional detachment component of psychopathy in adult forensic samples (Frick, 2006; Frick, Cornell, Barry, Bodin, & Dane, 2003). Although not all children with CU traits become psychopaths, they are at greater risk for developing psychopathy (Fontaine, McCrory, Boivin, Moffitt, & Viding, 2011). Children with CU traits are characterized by low fear in combination with high impulsivity, are not truly concerned with other people’s feelings, and typically do not feel bad or guilty when showing rule-breaking behavior. Furthermore, children with CU traits have more difficulty in identifying fearful and sad facial expressions (Blair, Colledge, Murray, & Mitchell, 2001; Stevens, Charman, & Blair, 2001; Woodworth & Waschbusch, 2008) and show reduced psychophysiological responding to distressing and threatening pictures, films and words, which indicates reduced affective arousal (Anastassiou-Hadjicharalambous & Warden, 2008; Blair, 1999; Kimonis, Frick, Fazekas, & Loney, 2006; Loney, Frick, Clements, Ellis, & Kerlin, 2003). Collectively, these studies show that children with CU traits have a specific deficit in processing negative emotional stimuli (Frick, 2006). To date, however, less is known whether this deficit in processing negative stimuli also impacts memory performance. Generally, emotional events are remembered better than neutral ones (Buchanan, 2007; LaBar & Cabeza, 2006). Moreover, research has demonstrated that high levels of arousal enhance memory for negative stimuli in particular (Cahill & McGaugh, 1995; Steinmetz, Addis, & Kensinger, 2010). Specifically, central details of an emotional event tends to be well remembered over time at the expense of memory for peripheral details (i.e., less contextually relevant or spatially peripheral to the attended event; see e.g. Burke, Heuer, & Reisberg, 1992; Christianson, 1992; Christianson, Loftus, Hoffman, & Loftus, 1991; Heuer & Reisberg, 1990; Levine & Edelstein, 2009; Otani, Libkuman, Widner, & Graves, 2007). However, in contrast to normally functioning healthy individuals, people with psychopathic traits do not seem to exhibit enhanced memory for (negative) emotional material (Christianson et al., 1996; Dolan & Fullam, 2005, 2010). For example, Christianson and colleagues (1996) showed that relative to non-psychopathic offenders, psychopathic offenders did not display enhanced memory for central emotionally negative details over peripheral details. In a similar vein, Dolan and Fullam (2005) found that psychopathic offenders were in fact worse at recalling emotional slides than healthy 85 5 Chapter 5 | Emotional memory in children with CU traits controls (for a replication study among conduct disordered adolescents, see Dolan & Fullam, 2010). To the best of our knowledge, however, there are no studies investigating whether children with CU traits also display diminished memory for emotional stimuli. If one would indeed find impairments in emotional memory in children with CU traits, this would provide valuable insights into why children with CU traits learn less from negative emotional experiences, such as being punished for misbehavior (see Dadds & Salmon, 2003). Research has shown that children with CU traits are less responsive to effective parenting, because they respond with insufficient anxiety when confronted with their misbehavior (Hipwell et al., 2007; Oxford, Cavell, & Hughes, 2003; Viding, Fontaine, Oliver, & Plomin, 2009; Vitacco, Neumann, Ramos, & Roberts, 2003; Wootton, Frick, Shelton, & Silverthorn, 1997). The assumption is that they are less likely to remember the experience of being punished, which makes them less likely to inhibit their behavior in future situations in order to avoid the punishment. Thus, the primary aim of the present study was to examine whether memory for central and peripheral components of neutral and emotionally negative pictures in children with high CU traits is impaired relative to children with low CU traits. Based on previous research, it was expected that memory for central details would be enhanced at the expense of memory for peripheral details overall (i.e., independent of CU traits or picture valence). Furthermore, for neutral pictures we expected no differences between the high and low CU groups with respect to the memory enhancing effect for central details relative to peripheral ones. However, for negative pictures it was expected that this central versus peripheral memory ratio would be larger in the low CU group relative to the high CU group. METHOD Participants The current sample of children was recruited through elementary schools. Information and consent forms that explained the nature of the study, together with the Antisocial Process Screening Device (APSD) to assess psychopathic traits in children, were distributed at elementary schools. Parents were asked to sign the consent form, fill out the APSD and to return them to the child’s school where they were collected. Children were allowed to participate only when they assented to the procedure and parental consent was obtained. This study was approved by the standing Ethical Committee of the Faculty of Psychology and Neuroscience, Maastricht University. In total, parents of 77 children between 8 and 12 years of age (mean age = 9.9) filled out the APSD and signed the consent form. To obtain extreme groups, we selected only the children with raw scores of 6 or higher on the CU subscale of the APSD as the high 86 Emotional memory in children with CU traits | Chapter 5 CU group (n = 24; 13 male) and the children with CU-subscale scores of 2 or lower as the low CU group (n = 18; 11 male). These two groups did not differ with respect to mean age (high CU group: 10.3; low CU group: 9.8; t (40) = -1.15, p = .26) and gender distribution (χ2 (2) = .68, p = .65). In Table 1, the mean scores of both groups on the APSD subscales are presented. Table 1 | Mean scores, standard deviations, and significant differences between groups on the APSD subscales Low CU group High CU group Narcissism 1.72 (1.49) 3.25 (3.22)* Impulsivity 2.89 (2.17) 4.29 (2.35) Callous-unemotional 1.33 (.84) 6.83 (1.20)* APSD total 6.17 (3.92) 17.88 (6.26)* 5 * p < .05 Materials Antisocial Process Screening Device The Antisocial Process Screening Device (APSD; Frick & Hare, 2001) is a 20-item questionnaire aimed to assess psychopathic traits in children and adolescents and is based on the Psychopathy Checklist-Revised (PCL-R; Hare, 1991). The APSD has to be completed by the child’s parents or teachers. It consists of three dimensions: callousunemotional traits (6 items), impulsivity (5 items) and narcissism (7 items). All items are answered with 0 (not at all true), 1 (sometimes true), or 2 (definitely true). Examined in both community and clinical samples, the internal consistency of the three subscales ranged from .65 to .85 (Frick, Bodin, & Barry, 2000). The APSD has been found to have good convergent and construct validity (Vitacco, Rogers, & Neumann, 2003). The Dutch translation that was used in the present study has also been well-validated (Bijttebier & Decoene, 2009). Peripheral Memory Test To test the memory for central and peripheral components of neutral and negative pictures, we developed a peripheral memory test (PMT). Stimuli consisted of pictures of a single object (central component) surrounded by a grey frame in which a red symbol was presented in each corner (peripheral component). The original stimuli consisted of 3 pictures with a width by height of 259 by 416 pixels, 12 pictures of 416 by 259 pixels, 4 pictures of 529 x 340 pixels, and 1 picture of 416 by 416 pixels. These stimuli were presented at 50% of their original size on a 15-inch screen with a resolution of 1440 by 900 pixels. 87 Chapter 5 | Emotional memory in children with CU traits A total of 20 pictures served as the central component; 10 neutral (pictures of a boot, lighthouse, locomotive, green traffic light, tree, bus, deer, fence, electric power pylon, and lamp), and 10 negative (aggressive dog, premature baby, hand with stitches, shark, weapon, black eye, crashed car, bloody knife, scary mask, and needle in an arm). Ten different symbols were used in the frame, with each symbol occurring once in a neutral and once in a negative picture, with varying orientation (see Figure 1 for an example). Figure 1 | Example of a negative picture with symbol used in the PMT. The PMT consisted of 3 practice trials and 20 test trials (10 with a neutral and 10 with a negative picture). Children were instructed to take a good look at the whole picture, including the frame, and were asked to try to remember everything. They were also warned that the picture would be presented only for a short moment. As can be seen in Figure 2, a fixation cross was presented first for 500 ms. Next, the target picture plus frame with symbols was presented for 1 sec followed by a black screen for 10 sec. Afterwards, the original target picture was presented together with 3 similar 88 Emotional memory in children with CU traits | Chapter 5 distractor pictures. The child had to identify the correct picture by entering the number corresponding to the picture (1, 2, 3, 4). Next, the same procedure followed for the symbol in which the correct orientation of the symbol (left, right, up, down) had to be chosen. Within each trial, the order of the forced choice recognition of the central and peripheral components was random, implying that sometimes the central components were presented first and at other times the peripheral components were presented first. The intertrial interval was 2 sec. The order of the test trials was also random. The PMT was run using the Inquisit 3 program. 5 ⬀ 㔀 洀猀 洀猀 洀猀 Figure 2 | Timeline of a PMT trial. Pilot data To check if the negative pictures were actually experienced as negative and the neutral ones as neutral, a group of children (n = 21) of 7 to 12 years of age rated the valence of the pictures on five-point Likert scales using smileys (1= very negative, 2= negative, 3= neutral, 4= positive, 5= very positive). The mean of the neutral pictures and negative pictures was 3.43 (SD = .35) and 1.89 (SD= .30), respectively (t (20) = 20.68, p <. 01, Cohen’s d = 4.76). Furthermore, we checked if either the neutral or the negative target pictures were easier to recognize than the other. Children had to rate the similarity of the target picture between the distractor pictures on a five-point Likert scale with higher scores indicating a higher degree of similarity between the pictures. There was a significant difference in similarity between the neutral and negative pictures with a higher mean for the neutral pictures (t (20) = 3.19, p < .01, Cohen’s d = 0.36). However, 89 Chapter 5 | Emotional memory in children with CU traits when piloting the PMT, children did not show longer reaction times for the recognition of neutral pictures. The mean reaction time was 5.89 seconds (SD = 1.94) for the neutral pictures and 6.09 seconds (SD = 2.00) for the negative pictures. This difference was not significant (t (9) = -.44, p = .67). Design and Procedure The present experiment was a 2 (Valence: neutral vs. negative) × 2 (Detail: central vs. peripheral) × 2 (Group: low CU vs. high CU) split plot design, with the last factor as the between subjects factor. The selected children were seated comfortably in front of a computer in a quiet room at the child’s school. The PMT was always administered in the afternoon to control for potential time of day effects on, for example, fatigue, attention and concentration. The experimenter explained the test by reading the instructions presented on the computer together with the child. After the three practice trials, the experimenter asked the child whether the procedure of the test was clear before starting with the test trials. The PMT took between 10 and 15 minutes. Afterwards, children were given a small present in return for their participation. RESULTS Figure 3 shows the mean proportion of correctly recognized central and peripheral components for the neutral and negative pictures for the two groups. To examine these differences in correct recognition, a repeated measures Analysis of Variance (ANOVA) was conducted with Group (high CU group vs. low CU group) as the between subjects factor and Valence (neutral vs. negative) and Detail (central vs. peripheral) as within subjects factors. The Group x Valence x Detail interaction was not statistically significant (F (1, 40) = .01, p = .94). However, a significant interaction between Valence and Detail was found (F (1, 40) = 34.87, p < .01, ηp2 = .47). To identify the differences between levels of valence and detail, simple effects analyses were performed. For the neutral pictures, the peripheral component was significantly better recognized than the central component (t (41) = -3.03, p < .01, Cohen’s d = -.68). However, for the negative pictures, the central component was significantly better recognized than the peripheral component (t (41) = 4.15, p < .01, Cohen’s d = .93). Furthermore, there was a significant difference between the central components for the neutral and negative pictures. The central component was better recognized for negative pictures than for neutral pictures (t (41) = -5.32, p < .01, Cohen’s d = -1.03). The peripheral component was significantly better recognized for the neutral pictures than for the negative pictures (t (41) = 3.19, p < .01, Cohen’s d = .64). None of the other interactions, nor main effects reached significance. 90 Emotional memory in children with CU traits | Chapter 5 Differences in latency were also examined, using a repeated measures ANOVA with Group (high CU group vs. low CU group) as the between subjects factor and Valence (neutral vs. negative) and Detail (central vs. peripheral) as within subjects factors. Again, the Group x Valence x Detail interaction was not significant (F (1, 40) = 2.34, p = .13). Only a significant main effect was found for detail (F (1, 40) = 34.35, p < .01, ηp2 = .49) with longer reaction times for the central components (M = 5783 ms) compared to the peripheral components (M = 4398 ms). No other main effects or interactions were significant. 5 1 0,9 mean proportion correct 0,8 0,7 0,6 low CU group 0,5 high CU group 0,4 0,3 0,2 0,1 0 central peripheral neutral central peripheral negative Figure 3 | Mean proportion correctly recognized central and peripheral components for the neutral and Figure3.Meanproportioncorrectlyrecognizedcentralandperipheralcomponentsfortheneutral negative pictures. andnegativepictures. DISCUSSION Previous studies found that adults and adolescents with psychopathic traits have a memory impairment for emotional stimuli (Christianson et al., 1996; Dolan & Fullam, 2005, 2010). The present study examined whether this finding could be replicated in a non-clinical sample of children high and low on CU traits. Interestingly, the results of this study showed no difference in emotional memory between the high and low CU groups. Our PMT was successful in replicating the peripheral memory effect found in previous studies (e.g., Burke et al., 1992; Christianson, 1992; Christianson et al., 1991; Heuer & Reisberg, 1990; Levine & Edelstein, 2009; Otani et al., 2007). For the neutral pictures, the peripheral component was better recognized than the central component, 91 Chapter 5 | Emotional memory in children with CU traits while the central component was better recognized than the peripheral component for the negative pictures. Interestingly, the peripheral component was better recognized for the neutral pictures than for the negative pictures. The ten symbols used as peripheral components were the same for both the neutral and negative pictures. Only the orientation of the symbol varied. Therefore, the difference in recognition of the peripheral components could only be explained by the valence of the central picture. Attention is being drawn to the negative picture, at the expense of the attention paid to the symbol. Generally, central details receive more attention than peripheral details, which would lead to a better memory for the central details than for peripheral details (Riggs, McQuiggan, Farb, Anderson, & Ryan, 2011). In contrast to our expectations, it was found that the peripheral component was better recognized than the central component of neutral pictures. In the present study, children were explicitly instructed to look at both the picture (central component) and the frame with the symbols (peripheral component). When both components receive the same amount of attention, it seems logical that the peripheral component was better recognized than the central component of neutral pictures since the central picture is more complex than a simple symbol. For emotionally negative pictures, attention is drawn to the central component despite the instruction to look at both components. Therefore, the central component receives more attention than the peripheral component which leads to a better memory for the central component at the expense of memory for the peripheral component (Glass & Newman, 2009). Although attention is not the only condition contributing to better memory for emotional details, it is a necessary one (Christianson et al., 1991; Kramer, Buckhout, & Eugenio, 1990). A possible explanation for not finding a difference in emotional memory between the high and low CU groups could be that the impairment in emotional memory has not developed yet in children of this age. Dolan and Fullam (2010) suggest that the deficit in emotional memory in people with psychopathic traits is reasonably stable across the lifespan. However, this does not seem to be the case for children with CU traits. In a recent study from our lab, we also found that children with low and high CU traits did not differ on true recall for both neutral and negative word lists (Thijssen, Otgaar, Howe, & de Ruiter, 2013). Salekin, Debus, and Jackson (2008) noted that with regard to performance tasks, the observed association between emotional processing deficits and psychopathy in adults is not as robust in children. Therefore, the disorder might not fully develop until adulthood and the impaired emotional processing in children with high CU traits may thus have not affected memory for emotional material yet. Previous studies focusing on central and peripheral memory have mainly used pictures from the International Affective Picture System (IAPS; Lang, Bradley, & Cuthbert, 92 Emotional memory in children with CU traits | Chapter 5 1998). However, the problem with such studies is that it is difficult to ascertain what the central and peripheral details exactly are. In our PMT, this distinction was clear and it was easy to manipulate the peripheral details. A limitation of the PMT could be the central pictures which were used. Since different pictures were used in the neutral and negative condition, it cannot be ruled out that the negative pictures indeed might be easier to recognize than the neutral pictures. In the pilot study, it was found that the distractor pictures for the negative pictures were less similar than the distractor pictures for the neutral pictures, which indicates that the negative pictures are easier to distinguish than the neutral pictures. Several possible explanations could be given for why the present study was not able to replicate previous findings regarding emotional memory of adolescents and adults with psychopathic traits (Christianson et al., 1996; Dolan & Fullam, 2005, 2010). First, this might have to do with that the present study used a non-clinical sample. Specifically, while the previous studies used clinical samples, the low and high CU groups in the current study were selected from the general population. Thus, even though our groups differed significantly in CU traits, the high CU group may nonetheless be difficult to compare with a high CU group selected from a clinical population in which CU traits are expected to be more extreme. Second, previous studies used a different memory task in that they used a slide show together with a short narrative to measure memory for central and peripheral details, while in the present study unrelated pictures were used. Although speculative, it could be that in the previous studies, the memory test with narratives evoked more relational processing (Hunt & Einstein, 1981), while our memory test in the current study likely elicited more item-specific processing. Since persons with psychopathic traits seem to have deficits in their relational processing of emotional memories, one might argue that our memory test was not sensitive enough to detect these relational processing deficits. Finally, our sample size was fairly small, which may have led to insufficient power to detect small between group differences. In sum, the present study did not find a difference in emotional memory between children with high and low CU traits. However, we did show that both groups had a better memory for the central component of negative pictures at the expense of the peripheral component, while their memory for neutral pictures was better for the peripheral component than for the central component. This finding suggests that the PMT is a useful tool to examine the memory narrowing effect found in previous studies. Moreover, based on the results of the present study, children with CU traits do not seem to have a deficit in emotional memory. This suggests that impairments in emotional memory may not be an explanation for why children with CU traits are less responsive to effective parental correction. 93 5 Emotional true and false memories in children with callous - unemotional traits 6 Thijssen, J., Otgaar, H., Howe, M.L., & de Ruiter, C. (2013). Emotional true and false memories in children with callous-unemotional traits. Cognition & Emotion, 27, 761-768. doi: 10.1080/02699931.2012.744300 Chapter 6 | True and false memories in children with CU traits ABSTRACT Several studies have found that children with callous-unemotional (CU) traits have a deficit in processing emotionally negative material. The present study examined whether this deficit also affects emotional memory. Twenty-two children with low CU traits and 24 children with high CU traits between 8 and 12 years of age were selected from a community sample and presented with neutral and negative emotional words, using the Deese-Roediger-McDermott paradigm. On true recall, there was no difference between the groups. Both groups had higher true recall rates for the neutral word lists than for the negative lists. However, on false recall, although there were no group differences for neutral word lists, the high CU group recalled significantly fewer critical lures on the negative word lists than the low CU group. Furthermore, the high CU group had significantly less false recall on the negative word lists compared to the neutral word lists, while the low CU group showed no difference in false recall between the word lists. These results indicate that children with high CU traits have no deficiencies in true memory performance, yet are less susceptible to developing false memories concerning emotionally negative material. 96 True and false memories in children with CU traits | Chapter 6 INTRODUCTION Some children with conduct disorder are characterized by so-called callousunemotional (CU) traits, which closely resemble the emotional detachment component of psychopathy in adult forensic samples (Frick, 2006). Children with CU traits display low fear in combination with high impulsivity, are not truly concerned with others’ feelings, and typically do not feel bad or guilty when showing rule-breaking behavior. Like adults with psychopathic traits, children with CU traits have deficits in emotional processing. Children with CU traits seem to have more difficulty in identifying fearful and sad facial expressions compared to expressions of surprise, happy, disgust, and anger (Blair, Colledge, Murray, & Mitchell, 2001) and may be less physiologically responsive to distressing and threatening stimuli (Blair, 1999). This indicates that the deficit in emotional processing in children with CU traits is specific to negative stimuli (Frick, 2006). There is evidence implying that this deficit can lead to an impaired memory for negative emotional material, while emotional events generally are remembered better than neutral ones (LaBar & Cabeza, 2006). For example, adults and adolescents with psychopathic traits appeared to be worse at recalling negative emotional slides than healthy controls (Dolan & Fullam, 2005, 2010). If these impairments in emotional memory could also be found in children with CU traits, it could explain why these children do not learn from negative emotional experiences. For example, research has shown that children with CU traits are less responsive to parental limit setting, because they respond with insufficient anxiety when confronted with their misbehavior (Oxford, Cavell, & Hughes, 2003; Wootton, Frick, Shelton, & Silverthorn, 1997). Therefore, they are less likely to remember the experience of being punished after their misbehavior, which makes them less likely to inhibit their behavior in future situations in order to avoid punishment (see Dadds & Salmon, 2003). To our knowledge, no studies have examined emotional memory in children with CU traits. Interestingly, it is likely that the deficiency in emotional processing found in individuals with psychopathic traits will not only affect accurate remembering, but also incorrect remembering (i.e., false memories). A robust paradigm to examine both true and false memory is the Deese-Roediger-McDermott (DRM) paradigm (Deese, 1959; Roediger & McDermott, 1995). In this paradigm, participants are presented with a list of words which are all semantically related to a non-presented theme word that is called the critical lure. A false memory occurs when the critical lure is incorrectly recalled or recognized as being presented in the word list (see for an overview, Brainerd, Reyna, & Ceci, 2008a). The development of false memories elicited by the DRM paradigm could be explained by the associative activation theory (AAT; Howe, Wimmer, Gagnon, & Plumpton, 2009) 97 6 Chapter 6 | True and false memories in children with CU traits which is partly founded on activation monitoring theory (Roediger, Watson, McDermott, & Gallo, 2001). According to the AAT, the processing of one word leads to the spreading activation of corresponding nodes or concepts in our mental lexicon (i.e., knowledge base). This process can also lead to the activation of the critical lure. AAT assumes that the development of false memories is predominantly the result of increases in the amount and strength of associative relations as well as the speed and automaticity with which these associations are accessed and activated (Wimmer & Howe, 2009). Because associated relatedness is higher among emotionally negative material than among neutral material (Howe, Candel, Otgaar, Malone, & Wimmer, 2010; Howe et al., 2009; Talmi & Moscovitch, 2004), AAT assumes that spreading activation is more automatic (faster) when negative material is encountered than when neutral material is experienced. This would lead to an increased risk of false memories for negative material. Indeed, a number of studies have found higher false recognition for negative information than for neutral information (e.g., Brainerd, Stein, Silveira, Rohenkohl, & Reyna, 2008b) although some studies showed that false recall is also higher for neutral than negative emotional word lists (e.g., Howe et al., 2010). The aim of the present study was to examine whether emotional memory in children with high CU traits differs from children with low CU traits. The DRM paradigm was used to test true and false recall for neutral versus negative word lists. Since emotional events are generally remembered better than neutral ones and AAT hypothesizes more spreading activation for negative material, it was predicted that children with low CU traits would have better true recall for negative rather than for neutral word lists. We did not expect a difference between low and high CU children for neutral word lists. Children with high CU traits are less likely to benefit from enhanced memory for negative material than children with low CU traits, because of their deficit in emotional processing (Frick, 2006). Therefore, we hypothesized that the high CU group would recall fewer negative words than the low CU group. This is also in accordance with the expectations of the AAT, which implies that it is likely that children with high CU traits would process negative information less automatically because they are trying to inhibit access to that material more than children with low CU traits (Howe, Toth, & Cicchetti, 2011). Based on AAT, one would expect that children with low CU traits would show increased false memories for the negative word lists, because of the heightened spreading activation. Again, no difference between the high and low CU groups was expected for the neutral word lists. For the negative word lists, children with high CU traits would be expected to have fewer false memories, because the automatic activation spreads slower than in children with low CU traits. 98 True and false memories in children with CU traits | Chapter 6 METHOD Participants The current sample of children between 8 and 12 years of age was recruited through elementary schools as part of a larger research project (see also Thijssen, Otgaar, Meijer, Smeets, & de Ruiter, 2012). Children were allowed to participate only when they assented to the procedure and parental consent was obtained. Information and consent forms which explained the nature of the study, together with the Antisocial Process Screening Device (APSD) to assess psychopathic traits in children, were distributed at the elementary schools. Parents were asked to sign the consent form, fill out the APSD and to return them to the child’s school where they were then collected. In total, data from 111 children were obtained. This study was approved by the standing Ethical Committee of the Faculty of Psychology and Neuroscience, Maastricht University. To obtain extreme groups, we selected only the children with scores of 6 or higher on the CU subscale of the APSD as the high CU group (n = 24; 13 male) and the children with CU subscale scores of 2 or lower as the low CU group (n = 22; 13 male). These scores were selected based on the means found for clinical and community samples in previous studies. For clinical samples, means of around 6 on the CU-subscale are found (e.g., Fite, Greening, Stoppelbein, & Fabiano, 2009). For community samples, means of around 2 on the CU-subscale are found (e.g., Dadds, Fraser, Frost, & Hawes, 2005). The two groups did not differ with respect to mean age (high CU group: 10.3; low CU group: 9.7; t (44) = -1.55, p = .13) and gender distribution (χ2 (2) = .97, p = .62). Materials Antisocial Process Screening Device The Antisocial Process Screening Device (APSD; Frick & Hare, 2001) is a 20-item questionnaire to assess traits of psychopathy in children (and adolescents). The APSD has to be completed by the children’s parents or teachers. It consists of three dimensions: callous-unemotional (6 items), impulsivity (5 items) and narcissism (7 items) with all items answered with 0 (not at all true), 1 (sometimes true), or 2 (definitely true). Examined in both community and clinic samples, the internal consistency of the three subscales ranged from .65 to .85 (Frick, Bodin, & Barry, 2000). The APSD has been found to have good convergent and construct validity (Vitacco, Rogers, & Neumann, 2003). The Dutch translation that was used in the present study has also been validated (Bijttebier & Decoene, 2009). 99 6 Chapter 6 | True and false memories in children with CU traits Deese-Roediger-McDermott paradigm Five neutral and five emotionally negative word lists were used. The critical lures of the neutral lists were bread, window, sweet, smoke, and foot. The critical lures of the negative word lists were murder, pain, punishment, death, and cry. Each list was 10 items long. These word lists have been used in previous research (see Howe et al., 2010; Otgaar, Peters, & Howe, 2012). The 10 word lists were orally presented through a computer at a 3-second rate. Half of the children received the neutral word list first followed by the negative lists, while the other half first received the negative word list. After a word list had been presented, the child had to do a distractor task (circling the letter X in a string of letters). Then the child had to recall as many words he/she could remember from the word list. This procedure was repeated until all 10 word lists had been presented. Design and Procedure The present experiment was a 2 (valence: neutral vs. negative) × 2 (group: low CU vs. high CU) split plot design, with the latter factor as the between subjects factor. The selected children were tested individually in a quiet room at their school. They were told they would hear a number of words through the laptop which they should try to remember. The DRM took about 15 minutes and was always administered in the morning to control for time of day effects on attention and concentration. Afterwards, children were given a small present in return for their participation. RESULTS Separate repeated measures analyses of variance (ANOVAs) were conducted for proportion true recall, recall of critical lures, recall of unrelated lures, and net accuracy with VALENCE (neutral vs. negative) as within subjects factor and GROUP (high CU group vs. low CU group) as between subjects factor. For both groups, Table 1 shows the mean proportion of true recall, false recall, recall of unrelated lures, and net accuracy on the neutral and negative word lists. For true recall, no significant interaction between valence and group was found (F (1, 44) = 1.56, p = .22). However, there was a significant main effect of valence (F (1, 44) = 13.84, p < .05, ηp2 = .24). True recall was higher for the neutral word lists than for the negative word lists (see Table 1). For false recall, there was a significant interaction between valence and group (F (1, 44) = 5.79, p < .05, ηp2 = .12). Simple effects analyses revealed that there was no significant difference in critical lures for the low CU group between the neutral and negative word lists (F (1, 21) = 2.01, p = .17). However, for the high CU group, there was a significant difference between the neutral and negative word lists with fewer critical lures for the negative word lists than the neutral word lists (F (1, 23) = 4.31, p < .05, ηp2 = .16). 100 True and false memories in children with CU traits | Chapter 6 Table 1 | Mean proportions and standard deviations per group for true recall, false recall, recall of unrelated lures, and net accuracy Low CU group High CU group Neutral Negative Neutral Negative True recall .51 (.11) .47 (.11) .56 (.11) .49 (.10) False recall .26 (.22) .35 (.25) .30 (.21) .20 (.19) Unrelated lures .07 (.07) .10 (.09) .08 (.07) .08 (.05) Net accuracy .70 (.20) .64 (.22) .68 (.18) .75 (.21) Note. CU = callous-unemotional Furthermore, independent samples t-tests showed no significant difference in critical lures between the groups on the neutral word lists (t (44) = -.57, p = .57). However, the difference between the groups on the negative word lists was significant with the high CU group recalling significantly less critical lures than the low CU group (t (44) = 2.37, p < .05). Figure 1 shows the proportion critical lures recalled for the neutral and negative word lists per group. 0,45 0,4 Mean proportion false recall 0,35 0,3 0,25 Neutral Negative 0,2 0,15 0,1 0,05 0 Low CU group High CU group Figure 1 | Mean proportion false recall on neutral and negative word lists in children with high and low CU Figure1.Meanproportionfalserecallonneutralandnegativewordlistsinchildrenwithhighand traits. lowCUtraits. the unrelated lures, which is the proportion of all recalled words that were For unrelated words (i.e., not in the word list and not the critical lure), no significant interaction (F (1, 44) = 2.95, p = .09) or main effect was found (F (1, 44) = 1.51, p = .23). 101 6 Chapter 6 | True and false memories in children with CU traits Furthermore, we examined any group differences in net accuracy (true recall/true recall + false recall; see Otgaar et al., 2012). Again, no main effect (F (1, 44) = .02, p = .88) or significant interaction was found (F (1, 44) = 2.87, p = .10). Since previous studies have found a difference in performance on the DRM paradigm between younger and older children (e.g., Brainerd, Holliday, Reyna, Yang, & Toglia, 2010; Brainerd, Reyna, & Zember, 2011), an exploratory repeated measures ANOVA was performed with VALENCE (neutral vs. negative) as within subjects factor and GROUP (high CU group vs. low CU group) and AGE (8/9 years vs. 11/12 years) as between subjects factors. For this analysis, there were 8 8/9-year old children with low CU traits, 6 8/9 year old children with high CU traits, 6 11/12 year old children with low CU traits, and 12 11/12 year old children with high CU traits. No significant interactions were found for true recall and the recall of unrelated lures. However, there was a significant threeway interaction for false recall (F (1, 28) = 6.87, p < .05, ηp2 = .20). Simple effects analyses showed only one significant difference between neutral and negative word lists, namely for the 11/12 year old children with low CU traits. These children recalled significantly more negative than neutral critical lures (F (1, 5) = 45, p < .05, ηp2 = .90). DISCUSSION Using the DRM paradigm, the present study examined whether there is a difference in true and false memories for neutral and negative word lists between children with high and low CU traits. Our findings can be summarized as follows. First, the two groups did not perform differently on true recall. For both groups, true recall was higher for the neutral word lists than for the negative word lists. This finding is not as predicted, but consistent with previous studies (Howe et al., 2010; Otgaar et al., 2012). Interestingly, we did not find that children with high CU traits differed in their true recall on the negative word lists compared with children with low CU traits. Therefore, the present results do not suggest that children with high CU traits have an impaired emotional memory. This is in contrast to previous studies on adolescents and adults with psychopathic traits (Dolan & Fullam, 2005, 2010), but consistent with findings from a recent study from our lab. In this study, children high and low on CU traits performed similarly on a task where they had to recognize central and peripheral components in neutral and emotionally negative pictures (Thijssen et al., 2012). For false recall, there was no difference between the high and low CU groups on the neutral word lists. However, there was a difference between the groups on the negative word lists: the high CU group recalled less critical lures than the low CU group. Moreover, the low CU group showed no difference in false recall between the neutral and negative word lists. This finding is not consistent with previous research in which 102 True and false memories in children with CU traits | Chapter 6 a difference between the word lists was found (e.g., Brainerd et al., 2008b; Howe et al., 2010). This could be due to the small sample size, which may have led to insufficient power to detect a significant difference. However, it should be mentioned that mixed findings concerning the effect of valence on false memories have been reported in previous studies (see Brainerd & Reyna, 2012). The high CU group showed a significant difference in false recall between the word lists. These children had fewer false memories on the negative word lists compared to the neutral word lists. This result on false recall is in accordance with the predictions of AAT. Children with high CU traits have difficulties in automatically processing negative emotional material. Furthermore, as children with high CU traits are found to process negative material differently than children with low CU traits (e.g., Blair, 1999; Blair et al., 2001), it is equally likely that the associative networks of children with high CU traits related to negative material are not as well-integrated and dense relative to children low on CU traits. As a consequence, the flow of information in such a network in high CU traits children will be less automatic and slower. Indeed, our study demonstrated lower negative false memory rates in children with high CU traits compared to children with low CU traits. Since previous research has found an age effect in DRM-performance (e.g., Brainerd et al., 2010; Brainerd et al., 2011), the low and high CU children in the present study were divided by age. Results showed that 11/12-year old children with low CU traits recalled significantly more critical lures for the negative word lists compared to the neutral word lists. However, these results should be interpreted with caution, because the number of children per group was very small. The following limitations deserve some comment. First, we used a community sample for our study. Children of both the low and high CU groups were selected from the general population. In the present study, no differences on true recall were found between the low and high CU group. It is possible that there would be a difference in true recall when the high CU group is selected from the clinical population. However, our findings concerning false recall could be even stronger for the high CU group when selected from the clinical population. Second, only one source of information (i.e., the parent) was used to measure CU traits in the children. It would have been better if additional measures of CU traits were used. Third, only five word lists per valence were used. Especially in our relatively small sample, this could create little variability in performance and limit the generalizability. To enhance the generalizability of the findings, it would be advisable to repeat this study with a larger set of (clinical) children including more word lists or other types of false memory measures (e.g., misinformation paradigm; Loftus, 2005). Fourth, the present study did not control for arousal. There is some evidence suggesting that false memories are related to arousal level (e.g., Corson & 103 6 Chapter 6 | True and false memories in children with CU traits Verrier, 2007). However, in studies with children, valence tends to play a more important role than arousal (i.e., most memory effects are driven by changes in valence not arousal) when these factors are varied orthogonally using the DRM paradigm (e.g., see Brainerd et al., 2010). Finally, one may wonder whether our findings can be translated to forensic contexts. That is, can results obtained from semantically-related word lists provide us with critical information about events that people experience in daily life? Although it has been debated whether the DRM paradigm might be useful in the legal domain (e.g., Brainerd et al., 2011), research shows that the DRM paradigm is a robust tool to examine the mechanism behind memory illusions. Even more important, studies show that the DRM illusion is related to autobiographical memory (e.g., Gallo, 2010). In sum, the present study found that children with high CU traits do not differ from children with low CU traits concerning true recall, which implies that the deficit in processing emotionally negative material does not affect correct emotional memory in children with CU traits. However, for false recall, children with high CU traits recalled fewer not presented words at the negative word lists than children with low CU traits. This indicates that children with high CU traits are better at differentiating between true and false memories concerning negative material than children with low CU traits. Children with CU traits are more likely to encounter negative emotional situations because of their aggressive/antisocial problem behavior. Our study shows that they are less likely to spontaneously, falsely report elements related to these negative situations. Conclusively, it seems that having high CU traits lowers the risk for inaccurate memories, while leaving true emotional memory untouched. 104 General discussion 7 General discussion | Chapter 7 Aims of the present dissertation This dissertation examined the role of parenting strategies in children with externalizing behavior problems. Research has demonstrated that effective parenting practices are an important mechanism contributing to the development of prosocial behavior in the child (DeGarmo & Forgatch, 2005, 2007; Forgatch & DeGarmo, 1999; Martinez & Forgatch, 2001; Ogden & Amlund-Hagen, 2008). Therefore, many interventions for childhood behavior problems are parent-based interventions, focusing on the promotion of effective parenting skills and a decrease of ineffective child rearing behaviors. In part I of this thesis, the effectiveness of Parent Management Training – Oregon model (PMTO) was evaluated at five child service agencies in The Netherlands. The Dutch government requires that the effectiveness of interventions is also evaluated in The Netherlands, before they will be entered into the Dutch youth institute (NJi) database of effective interventions. In Chapter 2, the effects of PMTO were compared to Care As Usual (CAU). Chapter 3 examined the association between treatment integrity of the PMTO therapists and treatment outcome of PMTO. In Chapter 4, the factor structure of one of the key measures of parenting practices used in the effectiveness study, the Caregiver Wish List (CWL), was examined. The aim of Part II was to examine emotional memory in children with callous-unemotional (CU) traits. These traits are characterized by a persistent pattern of behavior that reflects a disregard for the feelings of others as well as a general lack of empathy. Previous research has shown that children with CU traits display highly persistent and serious antisocial behavior problems (Frick, Cornell, Barry, Bodin, & Dane, 2003; Frick, Stickle, Dandreaux, Farrell, & Kimonis, 2005) and also exhibit clear deficits in emotional processing, which could become manifest in their emotional memory. The study presented in Chapter 5 examined whether a deficit in emotional memory could be found in children high in CU traits. Furthermore, it was explored whether deficits in emotional processing could also affect emotional false memory (Chapter 6). The main findings of the studies described in this dissertation will be discussed below. We will address methodological issues and implications for future research and clinical practice. The effectiveness of PMTO Because longitudinal research has demonstrated a high degree of stability of conduct problems in childhood and a possible aggravation into criminal and violent behavior in adulthood (Moffitt, 1993; Nixon, 2002), early interventions seem to be of the highest importance. In The Netherlands, there existed a need for effective treatment programs for children 4-12 years with antisocial conduct problems. Therefore, the Ministry of Health, Welfare and Sport decided in 2005 to fund the implementation of Parent Management Training Oregon model (PMTO). PMTO is a theory-driven, evidence- 107 7 Chapter 7 | General discussion based therapy for parents of children between 4 and 12 years of age who have serious externalizing behavior problems. The intervention has proven to be effective in the US, Norway, and Iceland (e.g., Forgatch & DeGarmo, 1999; Ogden & Amlund-Hagen, 2008; Sigmarsdóttir, Thorlacius, Guðmundsdóttir, & DeGarmo, 2014), but it cannot be automatically assumed that it would also be effective in The Netherlands. According to the Social Interaction Learning (SIL) model, parents have a major influence on the development of their child (Forgatch, Bullock, & Patterson, 2004). By use of effective parenting, parents stimulate a healthy, prosocial development in their children. However, when parents use ineffective parenting strategies (i.e., coercion) they stimulate the development of antisocial behavior. Therefore, parents are the main agents of change in PMTO. The assumption is that by improving the parenting skills, the behavior of the child will also improve. The study presented in Chapter 2 examined whether PMTO is more effective in reducing children’s problem behaviors, by increasing the use of effective parenting practices, and reducing parental psychopathology and stress, as compared to CAU. In this study, 146 clinically referred children aged 4 to 11 years and their parents were included of which 91 received PMTO and 55 CAU. Outcome data were collected via parent- and teacher-rated questionnaires and by means of observations of structured parent-child interactions. Families were assessed at four time points: pretreatment, and after 6, 12, and 18 months. Results showed no statistically significant differences in the effectiveness of the two interventions on the primary outcome measures relating to parent-reported externalizing behaviors. That is, both treatment conditions showed a significant decrease in parent-rated externalizing behavior problems over time. Also on parenting stress and parental psychopathological symptoms, no significant differences between PMTO and CAU were found, with parents experiencing significantly less parenting stress and psychopathological symptoms over the course of both interventions. For observed parenting skills, no significant effects were documented. For self-report, the only parenting practice that significantly improved for both PMTO and CAU was the use of discouragement of undesirable behavior (i.e., discipline). For both conditions, improvements were most evident during the first six months of the study and then remained stable until 18 months after baseline. These findings imply that PMTO is effective in a clinically referred sample of children with externalizing problems in the Netherlands, but not more effective than CAU. When examining factors that may impact the effect of PMTO, we found that the treatment effect was moderated by the level of externalizing behavior problems at baseline: especially children with serious externalizing behavior problems appeared to benefit more from PMTO. One could argue that children with higher levels of problem behavior might be more difficult to treat. However, our results suggest the opposite, which may have been due to the fact that there was simply more room for improvement 108 General discussion | Chapter 7 in these children. However, a recent review by Shelleby and Shaw (2014) showed that this finding was demonstrated in more studies evaluating parenting interventions, such as Incredible Years and Family Check-Up. These findings together demonstrate that children with higher levels of antisocial behavior problems, which are thought to be at greater risk for the persistence and exacerbation of these behaviors (Broidy et al., 2003; Nagin & Tremblay, 1999), benefit from early intervention. Improvements in behavior problems were only documented for parent ratings but not for teacher ratings, which is a finding not uncommon in research on parenting interventions for children with externalizing behavior problems (e.g., Kjøbli, Hukkelberg, & Ogden, 2013; Kjøbli & Ogden, 2012). One explanation for this result might be that children’s behavior problems are less apparent in the more structured school setting and that, therefore, change was less noticeable for the teachers. Indeed, our data indicated that teachers overall reported lower levels of behavior problems compared to parents. This is also in accordance with previous empirical findings showing a fairly low agreement between parents and teachers with regard to these problem behaviors, with teachers reporting lower behavior problem scores as compared to parents (Youngstrom, Loeber, & Stouthamer-Loeber, 2000). Another explanation could be that parents and teachers observe children’s problem behavior in different settings and roles. Teachers are more likely to make norm-referenced assessments by comparing the behavior problems of the child to behaviors of other children. Parents on the other hand are more likely to make ipsative assessments, where the children are compared to themselves over time (Hukkelberg & Ogden, 2013). Notwithstanding these considerations, the most conservative interpretation of our results is that the positive treatment effects did not generalize to the school setting and that these parent-based interventions may only be effective in the context where they are implemented (i.e., at home). This was the first study that showed that PMTO did not produce better effects compared to a treatment control group. Although most of the earlier studies conducted in the US compared PMTO to a waiting list control condition, families included in the control condition of studies performed in Norway and Iceland also received treatment. When comparing our effect sizes for PMTO and CAU with those obtained in the other European studies, the effect size found for PMTO in our study was as high (Norwegian study) or even higher (Icelandic study). However, the effect size of CAU in our study seemed to be somewhat higher than the effect sizes of CAU in Norway and Iceland, which indicates that the regular treatments offered to children with externalizing behavior problems in The Netherlands appear to be of relatively good quality. This is likely the result of the fact that many psychologists in our country are trained to apply cognitivebehavioral techniques, which are an important ingredient of effective interventions for children with externalizing problems (Greene, Ablon, Goring, Fazio, & Morse, 2004), 109 7 Chapter 7 | General discussion including PMTO. In contrast to other European countries such as Norway and Iceland, implementation of evidence-based interventions is fairly common in The Netherlands. Moreover, PMTO is not the only evidence-based treatment for externalizing behavior problems that was not found to be more effective than CAU in The Netherlands. For example, the Triple-P Positive Parenting Program developed by Sanders (2008) produced similar effects to CAU as delivered in The Netherlands (Kleefman, Jansen, Stewart, & Reijneveld, 2014; Spijkers, Jansen, & Reijneveld, 2013). The study of Ogden and Amlund-Hagen (2008) showed that discipline is one of the most significant parenting strategies contributing to improvements in child behavior. The time-out procedure is an important part of PMTO treatment that helps parents to effectively discipline their child. In the past five decades, numerous studies have shown that the use of time-out is effective in reducing antisocial behavior problems, such as non-compliance, yelling, and aggression (e.g., Donaldson & Vollmer, 2011; Everett et al., 2007; Scarboro & Forehand, 1975). This procedure is not only used in PMTO, but also in other evidence-based parent management training programs such as the Incredible Years Program (Webster-Stratton & Reid, 2010) and the above mentioned Triple-P program (Sanders, 2008). However, recently, there has been some debate about the application of the time-out procedure. An article in the popular American magazine TIME even claimed that the time-out procedure would be harmful for children (Siegel & Bryson, 2014, September 23). This conclusion was based on research showing that the experience of relational pain (e.g., rejection) activates the same brain areas as the experience of physical pain (Eisenberger, Lieberman, & Williams, 2003). However, a subsequent check of the original research paper (Siegel & Bryson, 2014, September 23) indicates that the conclusions drawn in the TIME article were far too liberal in that relational pain is not fully equal to what happens during a time-out (Siegel & Bryson, 2014, October 21). More precisely, a time-out is not emotionally damaging when applied properly. Within PMTO, parents are taught to give a time-out in a calm way using a neutral tone of voice. Parents only give a time-out when the child has not complied with a request that has been repeated two times. Furthermore, it is clearly stated to the child why he/she receives a time-out. The procedure has been explained to the child beforehand, and therefore the child knows what to expect and what to do to comply with his/her parent’s request in order to avoid a time-out. Most importantly, when the time-out is finished, the issue is dropped. Especially when given occasionally and combined with a lot of praise and positive attention for appropriate, prosocial behavior, the time-out procedure is a safe, predictable, and effective strategy that has little to do with (interpersonal) rejection and isolation. 110 General discussion | Chapter 7 Determinants of the effectiveness of PMTO The fact that PMTO was not more effective in reducing externalizing behavior problems than CAU raises questions about possible factors that may have influenced the effectiveness of PMTO in The Netherlands. For example, prior research found evidence to suggest that higher PMTO treatment fidelity, which can be defined as the delivery of the therapy according to the treatment principles, predicted larger improvements in parenting skills and externalizing behavior problems (Forgatch & DeGarmo, 2011; Forgatch, Patterson, & DeGarmo, 2005; Hukkelberg & Ogden, 2013). Within PMTO, treatment fidelity is measured by means of the Fidelity of Implementation Rating System (FIMP; Knutson, Forgatch, Rains, & Sigmarsdóttir, 2009). In Chapter 3, treatment fidelity was examined as a possible determinant of the effectiveness of PMTO. Data of the 86 PMTO families from the effectiveness study presented in Chapter 2 were used. FIMP scores of certified PMTO therapists were significantly and meaningfully associated with treatment outcome measures. Results showed that treatment fidelity was related to outcome measures of externalizing behavior problems, parenting practices, parenting stress, and parental psychopathological symptoms. The higher the fidelity scores of the therapists, the larger the improvements on these outcome measures. In addition, significant differences in treatment fidelity scores were found between treatment completers and dropouts: parents who completed the treatment had a more adherent therapist than families who dropped out. However, no significant correlations were found between treatment fidelity and treatment satisfaction. Previous studies on treatment fidelity of PMTO only examined whether this factor was predictive of outcome with regard to parenting practices or child behavior problems. Our study showed that fidelity scores were also related to other outcome measures, such as parental stress and psychopathological symptoms. Therefore, even for certified PMTO therapists, all of whom had obtained at least a sufficient FIMP score, treatment fidelity varied systematically with effectiveness. Furthermore, the present study showed that most associations were significant for the outcome assessment at 18-months follow-up. This indicates that while the PMTO treatment is ongoing, the effects of PMTO are similar for families who have a therapist with sufficient fidelity and those who have a therapist with excellent fidelity. However, for an enduring effect of PMTO, it seems to be important that the therapist adheres very closely to the PMTO method. The assessment of parenting practices In our effectiveness study (Chapter 2), it appeared to be difficult to measure change in parenting practices. Since multi-method assessment has been shown to possess incremental validity over the reliance on only one measurement method (Harvey, Danforth, Ulaszek, & Eberhardt, 2001; Meyer et al., 2001), both self-report and observation 111 7 Chapter 7 | General discussion of structured child-parent interactions were used to assess parenting practices. Both methods have their advantages and limitations. Self-report questionnaires provide a more general picture of parents’ rearing behaviors and interactions with the child, thereby giving an impression of parenting behaviors over longer time periods and across diverse situations (Zaslow et al., 2006). However, it may be quite difficult for parents to evaluate their own parenting behaviors for a time frame of three or six months (Morsbach & Prinz, 2006). Furthermore, parents may give socially desirable answers or simply lack insight into their own behaviors towards their child (Schwarz, 1999). Structured observations of parent-child interactions in a laboratory situation or at home may be an improvement in this regard: this method of assessing parenting is more objective and indeed has been found to be more predictive of child outcomes than selfreport questionnaires (Duijster et al., 2015; Zaslow et al., 2006). Although observational methods are often seen as the gold standard in the assessment of parenting practices, their use in clinical settings is limited. Structured observations are more complex, time consuming and, therefore, more expensive as compared to self-report scales (Hawes & Dadds, 2006). Moreover, the observed rearing behaviors are obviously only a sample of the behaviors parents show in real life and some important parenting practices may be quite difficult to assess, because they rarely occur during structured observation tasks (Peterson, Tremblay, Ewigman, & Popkey, 2002; Shelton, Frick, & Wootton, 1996). The latter also seemed to be the case in our study. The observational tasks we used did not elicit particularly high levels of negative behaviors in the child, so parents hardly had to engage in disciplining their child during the assessments. Even at baseline, when children were expected to show clear signs of externalizing behavior problems, the frequency of such behaviors was quite low. This may partly explain why we did not find evidence for a treatment effect on parenting skills as assessed by means of observations. Points of critique can also be raised regarding the self-report measure of parenting skills, the Caregiver Wish List (CWL), that we used in our study. This relatively new measure was employed because it was specifically developed to measure the core parenting practices of the SIL model. In our effectiveness study, the internal consistency of the CWL appeared to be unsatisfactory for five out of the six a priori domains of parenting. Nevertheless, with the two scales of the CWL (discouraging undesirable behavior and CWL total score) that did show sufficient reliability, positive treatment effects could be documented. So far, no study on the psychometric properties of the CWL has been conducted. Therefore, the aim of the study reported in Chapter 4 was to explore the factor structure of the CWL as a first step in validating this instrument. Furthermore, its reliability and validity were examined. A sample of 348 parents of children aged between 4 and 11 years was used for this purpose, of which 220 were parents of children from the general population and 128 were parents of children with externalizing 112 General discussion | Chapter 7 behavior problems that were clinically referred. The factor analysis conducted on the data of the total sample revealed five factors, which were named Adequate discipline, Controlled responding, Focus on positive behavior, Consistency, and Monitoring, and deviated substantially from the original domains. Only 23 of the 53 original CWL items were included in the five factors. Furthermore, three of the five factors contained items from one original domain (i.e., discouraging undesirable behavior) and no items from the problems solving domain were represented in any of the factors. The content of the extracted factors for the community and clinical samples was quite comparable. Therefore, the extracted CWL factor structure seems to be applicable to parents from both the general and clinical populations. The extracted factor structure of the CWL can be considered an improvement over the original domains. Although the factors contained a lower number of items, four of the five factors exhibited acceptable internal consistency coefficients (i.e., α > .70). In contrast, of the six original domains, the internal consistency was acceptable for only one domain. Therefore, in the PMTO effectiveness study, only this domain and the CWL total score could be used to examine change in parenting and these measures indeed showed significant change over time. Furthermore, parents from the community sample displayed significantly higher scores on the five factors than parents from the clinical sample, indicating that parents in the community sample reported a higher quality of parenting skills than parents in the clinical sample. Results also showed that all extracted factors were negatively related to child behavior problems, with significantly higher associations for externalizing behaviors than for internalizing behaviors thereby supporting the external validity of the CWL. The use of the reliable factors could provide insight into change on various dimensions of parenting. For research purposes, however, it would probably be better to rely on another, better validated self-report measure for assessing parenting skills, such as the Alabama Parenting Questionnaire (APQ; Essau, Sasagawa, & Frick, 2006), which has been shown to possess adequate psychometric properties in various studies (e.g., Dadds, Maujean, & Fraser, 2003; Escribano, Aniorte, & Orgilés, 2013; Shelton et al., 1996; Zlomke, Lamport, Bauman, Garland, & Talbot, 2014) and which also measures parenting skills similar to the ones defined by the SIL model. Nevertheless, for clinical purposes, the CWL could still be more suitable than the APQ, because this interview-based instrument has been designed to help therapists gather information on strengths and weaknesses in parenting practices as seen by parents themselves, and is likely to enhance the therapeutic alliance and positive engagement between parents and therapists (Hodges, 2005). Thus, the CWL may have higher clinical utility than self-report paper and pencil tests such as the APQ, and further examination of the psychometric properties of the CWL seems warranted. Probably, items could be added to some of the factors to improve their internal consistency. 113 7 Chapter 7 | General discussion Inflexible clinical practice vs. a Randomized Controlled trial as ideal A randomized controlled trial (RCT) is considered to be the gold standard for evaluating the effectiveness of interventions. The randomization procedure ensures that there are no differences between the experimental and control group in known and unknown factors that might influence the outcome. This minimizes confounding, which is the bias that occurs when one treatment group exhibits certain (known or unknown) features affecting the relationship between the intervention and outcome, that the other group does not possess (Sibbald & Roland, 1998; Viera & Bangdiwala, 2007). Furthermore, expectancy effects and other heuristic biases are controlled through randomization. Therefore, the observed differences between the experimental and control group can be attributed to the effects of the treatment instead of other factors (Eccles, Grimshaw, Campbell, & Ramsay, 2003). However, RCTs have a number of practical and ethical limitations. First, they are expensive and time consuming. Furthermore, assigning patients to a control treatment that is believed to be inferior to the target intervention is often regarded as unethical (Sibbald & Roland, 1998). Therapists may refuse to include patients who should be allocated at random (Fairhurst & Dowrick, 1996). Patients may also have a strong preference for one of the treatments, which limits recruitment and could bias the outcomes (Brewin & Bradley, 1989). In our study of the effectiveness of PMTO, we encountered some of the problems described above. Our study was originally designed as a RCT, but we had to continue as a quasi-experimental study, which is of course a major limitation. Allocation to the treatment conditions (i.e., PMTO and CAU) was random at three of the five child service agencies. At the other two agencies no randomization took place, because after the study had started, one child service agency decided to no longer offer CAU and thus only recruited families for the PMTO condition. Another agency was then specifically included in the study to compensate for the missing CAU families. Although the families in the PMTO and CAU condition did not differ in terms of demographic characteristics and almost all of the outcome measures at baseline, the possibility cannot be ruled out that there were still unknown factors that may have affected the outcome. Emotional memory in children with callous-unemotional traits Some children with externalizing behavior problems are characterized by so-called callous-unemotional (CU) traits, similar to the emotional detachment component of psychopathy (Frick, 2006; Frick et al., 2003). Children with CU traits are more likely to show persistent and more serious antisocial behavior during adolescence and adulthood and they are at greater risk of developing psychopathy (Fontaine, McCrory, Boivin, Moffitt, & Viding, 2011). Children with CU traits have low fear, high impulsivity, are unconcerned with other people’s feelings, and do not exhibit feelings of guilt after misbehaving 114 General discussion | Chapter 7 (Frick, 2006). Previous research has shown that these children have a specific deficit in the processing of negative emotional stimuli. For example, they have more difficulty in identifying fearful and sad facial expressions and are less physiologically responsive to distressing and threatening pictures, films and words (Anastassiou-Hadjicharalambous & Warden, 2008; Blair, 1999; Blair, Colledge, Murray, & Mitchell, 2001; Kimonis, Frick, Fazekas, & Loney, 2006; Loney, Frick, Clements, Ellis, & Kerlin, 2003; Stevens, Charman, & Blair, 2001; Woodworth & Waschbusch, 2008), which might also affect their emotional memory. Impairments in emotional memory could be an explanation for why children with CU traits learn less from negative emotional experiences, such as being disciplined for misbehavior (see Dadds & Salmon, 2003). Previous research has shown that children with CU traits are unresponsive to discipline, which is an important parenting strategy in PMTO (DeBaryshe, Patterson, & Capaldi, 1993; Nix et al., 1999; Ogden & AmlundHagen, 2008; Oliver, Guerin, & Coffman, 2009). The assumption is that they are less likely to remember the experience of being disciplined, which makes them less likely to inhibit their bad behavior in future situations. Studies in adults and adolescents showed that, in contrast to normally functioning healthy individuals, individuals with psychopathic traits do not seem to exhibit enhanced memory for (negative) emotional events (Christianson et al., 1996; Dolan & Fullam, 2005, 2010). This implies that psychopathic people experience adverse emotional situations as neutral, which could explain why these people use antisocial behavior instrumentally to fulfill their desires without feelings of guilt over the other person’s distress (Dolan & Fullam, 2010). This finding has not yet been replicated in children. Therefore, emotional memory was examined in children high and low on CU traits. We also examined whether the deficit in processing emotional stimuli in children with high CU traits could have an effect on their emotional false memory. If children with high CU traits are found to be at increased risk of developing false memories for emotional information, this could provide information on the involvement of memory in the learning and maintenance of antisocial behavior. Emotional memory distortions can influence the child’s own behavior as well as the interpretation of others’ behaviors. Chapter 5 presents results of a self-developed peripheral memory test in which memory performance for central and peripheral components of neutral and emotionally negative pictures in children with high CU traits was compared with the performance of children with low CU traits. The peripheral memory test is based on the weapon focus effect, which refers to the phenomenon that attention is being drawn to the weapon (central aspect) and consequently less attention is devoted to the other details (peripheral aspects). Therefore, memory is generally poorer for the peripheral details surrounding the weapon in comparison to memory for the weapon itself (Pickel, Narter, Jameson, & Lenhardt, 2008). From a community sample of children between 8 and 12 115 7 Chapter 7 | General discussion years of age, children with high (n=24) and low CU traits (n=18) were selected. We found that overall recognition rates did not differ between the high- and low-CU groups. For negative pictures, both groups demonstrated better recognition of the central component at the expense of the peripheral component, while for neutral pictures, the peripheral component was better recognized than the central component. Next, these children also performed the Deese-Roediger-McDermott (DRM) paradigm (Deese, 1959; Roediger & McDermott, 1995) to examine both true and false memory formation. In this paradigm, children are presented with a list of words that are all semantically related to a non-presented theme word, which is called the critical lure. A false memory occurs when the critical lure is incorrectly recalled as being presented in the word list (see for an overview, Brainerd, Reyna, & Ceci, 2008). For the study described in Chapter 6, five neutral and five emotionally negative word lists were used. Analyses revealed no significant difference between the high and low CU groups in their performance on true recall. For both groups, true recall was higher for the neutral word lists than for the negative word lists. For false recall, there was no difference between the high and low CU groups on the neutral word lists. However, a difference between the groups was found on the negative word lists: the high CU group recalled fewer critical lures than the low CU group. Moreover, the low CU group showed no difference in false recall between the neutral and negative word lists, while the high CU group children had fewer false memories on the negative word lists compared to the neutral word lists. Taken together, the results of these two studies suggest that children with high CU traits do not differ from children with low CU traits in their emotional true memory. Therefore, children with CU traits do not seem to have a deficit in emotional memory. This implies that impairments in emotional memory may not be an explanation for why children with CU traits are less responsive to effective parental correction. These findings are in contrast with the results obtained previously in adolescents and adults (Christianson et al., 1996; Dolan & Fullam, 2005, 2010), where psychopathic offenders did not show enhanced memory for central emotionally negative details over peripheral details while healthy controls did. Possibly, the impairment in emotional memory has not developed yet in children of this age. Salekin, Debus, and Jackson (2008) noted that with regard to performance tasks, the observed association between emotional processing deficits and psychopathy in adults is not as robust in children. Therefore, psychopathy might not fully develop until adulthood and the impaired emotional processing in children with high CU traits may thus not have affected memory for emotional material yet. However, we did find a difference between children with high and low CU traits on emotionally negative false memories. The results indicate that children with high CU traits are better at differentiating between true and false memories concerning negative material than children with low CU traits. Conclusively, 116 General discussion | Chapter 7 it seems that having high CU traits lowers the risk of inaccurate memories, while leaving true emotional memory untouched. This finding is in contrast to the predictions of the hostile attribution bias and the general aggression model for aggression. People with a hostile attribution bias interpret intentions of others as hostile, whether they are or not (Dodge, 1980). According to the general aggression model, aggression in people with this bias is the result of the activation and application of aggression-related schemas and scripts stored in memory (Anderson & Bushman, 2002). These aggression-related knowledge structures are more extensive and elaborate, and operate automatically in aggressive-prone individuals, causing them to choose to behave aggressively in response to another person’s comment or behavior (Laney & Takarangi, 2013). Based on this cognitive model, one would expect that aggressive-prone individuals are more susceptible to falsely remember aggression-related material. Indeed, by use of a modified DRM-paradigm, Takarangi and colleagues (2008) found that aggressive participants were more likely to falsely recall unpresented aggressive words as compared to less aggressive participants. In a similar vein, studies of Vannucci and colleagues (2012; 2014) showed that violent false memories were positively associated with delinquent behaviors and bullying in adolescents. Recently, Laney and Takarangi (2013) even demonstrated that individuals with aggressive tendencies were more prone to falsely remember having committed aggressive acts. All in all, these results suggest that aggressive people are more likely to remember aggressive information they have never encountered (Vannucci et al., 2014). Although aggressive behavior is only one component of antisocial behavior displayed by children with CU traits, it could still be expected that these same results would have been found in our study, since some of our word lists also contained aggressive-related words (e.g., murder, pain). Obviously, this post-hoc explanation is speculative and should be examined in future studies. However, our findings are more in line with results found in a study conducted by Mirandola and colleagues (2014). Their findings demonstrated that elaboration of an experienced emotional event increases the risk of false memories. This can be explained in light of the associative activation theory (AAT; Howe, Wimmer, Gagnon, & Plumpton, 2009), which states that semantically related words, events, and experiences are connected in our memory and share associative relations (Wimmer & Howe, 2009). Because associated relatedness is higher among emotionally negative material than among neutral material (Howe, Candel, Otgaar, Malone, & Wimmer, 2010; Howe et al., 2009; Talmi & Moscovitch, 2004), spreading activation is more automatic for negative than for neutral material, which would lead to an increased risk of false memories for negative material. Thus, when children without CU traits encounter a negative emotional situation, they may react with empathy or shame to the emotion-provoking 117 7 Chapter 7 | General discussion event. Therefore, they are likely to reflect on this experience and to elaborate on it, which puts them at risk for memory errors (Brainerd, Stein, Silveira, Rohenkohl, & Reyna, 2008; Christianson et al., 1996; Mirandola et al., 2014). Since children with CU traits seem to experience emotionally negative information as neutral, they may be less likely to reevaluate a negative event, which protects them against erroneous memories. The interpretation of the results of these studies should be considered in light of their methodological limitations, which could also be an explanation for not replicating previous findings regarding emotional memory in individuals with psychopathic traits. First, the low and high CU groups in our studies were selected from the general population, while previous studies relied on clinical samples. It is possible that there would be a difference in true recall when the high CU group is selected from the clinical population, since they can be expected to be more extreme in their CU traits. Second, our studies used different memory tasks than the previous studies that used a slide show together with a short narrative (Christianson et al., 1996; Dolan & Fullam, 2005, 2010). Possibly, the memory test with narratives relies more on relational information, while our memory tests (PMT and DRM) likely depended on item-specific information (Hunt & Einstein, 1981). Relational information refers to features that are similar between items that have to be remembered and item-specific information refers to features that differentiate one item from the other items that have to be memorized. False memories are more easily induced when relying on relational information (Cirelli, Dickinson, & Poirier, 2014). Since persons with psychopathic traits seem to have particular deficits in their relational processing of emotional memories (Christianson et al., 1996), one might argue that our memory tests were not sensitive enough to detect these relational processing deficits. Finally, our sample size was fairly small, which may have led to insufficient power to detect small between group differences. Callous-unemotional traits in relation to parent management training programs Unfortunately, we were not able to assess CU traits in the children included in the PMTO effectiveness study. By the time the idea of incorporating CU traits emerged, the effectiveness study was already too far under way. Furthermore, we did not want to increase the burden on the parents more by adding another questionnaire to the already quite extensive battery of tests included in the study. Nevertheless, it would have been interesting to examine whether the effects of PMTO (and CAU) are different for children high and low on CU traits. As mentioned in the introduction of this dissertation, the frequency and severity of externalizing behavior problems in children with CU traits seem to be independent of the quality of parenting (Hipwell et al., 2007; Oxford, Cavell, & Hughes, 2003; Viding, Fontaine, Oliver, & Plomin, 2009; Vitacco, Neumann, Ramos, & 118 General discussion | Chapter 7 Roberts, 2003; Wootton, Frick, Shelton, & Silverthorn, 1997). Surprisingly, little research has been conducted on the association between CU traits and parent management training outcome. The few trials that have been conducted show that children with high CU traits have poor treatment outcomes as compared with children low on CU traits (e.g., Hawes & Dadds, 2005; Hawes, Dadds, Brennan, Rhodes, & Cauchi, 2013; Högström, Enebrink, & Ghaderi, 2013; Kimonis, Bagner, Linares, Blake, & Rodriguez, 2014). Importantly, parents of high and low CU children did not differ in improvement of parenting skills during the treatment (Högström et al., 2013). However, this does not mean that the level of behavior problems in children with CU traits is immutable. Research has found that the level of externalizing behavior problems in CU children decreased after parent management training, especially when these treatments were based on social learning theory (Hawes, Price, & Dadds, 2014). Furthermore, it has been found that parent management training for children with high CU traits is particularly capable of improving CU traits when delivered early in childhood (i.e., preschool age; Hawes et al., 2014). These findings clearly suggest that pre-treatment data on CU traits are clinically informative regarding the prognostic status of children referred for treatment for externalizing behavior problems (Hawes et al., 2014). Some components of parent management training might be more effective in children with CU traits than others. Since children with CU traits appear to have a reward dominance response style and seem to be relatively insensitive to discipline, they are expected to be more responsive to positive reinforcement than discipline. Therefore, to achieve the most change in externalizing behavior problems in children with CU traits, a focus on positive reinforcement and parental warmth may be particularly indicated in parent management training (Hawes et al., 2014). In accordance with this notion, Hawes and Dadds (2005) found that boys with high CU traits were less responsive to discipline by means of a time-out and responded with less affect to this type of parenting than boys without CU traits. Furthermore, a review by Frick and colleagues (2014) showed that children with high and low CU traits responded equally well to the part of the parent training that focused on positive reinforcement, while children with low levels of CU traits demonstrated added improvement with the part of the intervention that focused on teaching parents to use more effective discipline strategies. It should be noted, however, that the focus should not be exclusively on warm and positive parenting. Parents should also be equipped with non-coercive ways of discipline, because coercive parenting has been associated with increases in CU traits (Waller et al., 2012). 119 7 Chapter 7 | General discussion Concluding remarks and directions for future research The findings of this dissertation suggest that PMTO is effective in The Netherlands in improving child externalizing behavior problems, effective parenting practices, parenting stress, and parental psychopathological symptoms, but not more effective than CAU. Before drawing any conclusions on the future of PMTO in The Netherlands, it should first be examined for whom and under which circumstances the treatment is most effective. Therefore, a number of possible moderators and determinants of the effectiveness of PMTO were investigated. Of several child, parent, and family factors, only externalizing behavior problems at baseline appeared to be a moderator: children with more severe externalizing behavior problems at baseline showed more improvement than children with less severe problems. The current dissertation also showed that the effectiveness of PMTO is significantly related to the treatment integrity of the PMTO therapist. When PMTO therapists adhered closely to the program principles, more positive long-term treatment effects were obtained. Moderate adherence to the treatment model appears to be insufficient to achieve this. Furthermore, research by others has shown that children with CU traits respond less well to effective parenting. As a possible explanation for this, emotional memory was examined in children high and low on CU traits, because the assumption is that they are less likely to remember the experience of being disciplined, which makes them less likely to inhibit their bad behavior in future situations. However, no differences in emotional (true) memory were found between children high and low on CU traits, indicating that an impaired emotional memory is not an explanation for why children with high CU traits are less responsive to an intervention that focuses on effective parenting than children with low CU traits. Although PMTO was not more effective than CAU, it could be that PMTO is more cost-effective than CAU, especially when long term societal effects are taken into account. Evidence of cost-effectiveness is important, because it can assist in decisions concerning which intervention to use, at what cost and with what benefit (Charles, Bywater, & Edwards, 2011). Along with our study on the effectiveness of PMTO, data concerning child and parental health, and the use of health care (benefits) as well as time investment and other costs of the interventions (costs) were collected. These data are currently being analyzed by researchers of the Medical Center of the University of Groningen (UMCG). If these data show that PMTO is more cost-effective than CAU, PMTO would be the preferable intervention for children with externalizing behavior problems. When no differences between PMTO and CAU in cost-effectiveness are found, PMTO could still be an improvement over CAU for some children with externalizing behavior problems. In combination with previous studies, the results of the research presented in this dissertation suggest that PMTO could be more beneficial for children 120 General discussion | Chapter 7 with more severe externalizing behavior problems. Further, especially when delivered by strongly adherent therapists, PMTO could be more effective than CAU. The ultimate test of PMTO’s effectiveness, compared to CAU, lies in its long-term ability to curb a child’s development into an adolescent and adult with an antisocial lifestyle. We strongly recommend follow-up research in our sample of treated children, to monitor their developmental trajectory. More research on the factors that predict, influence, or account for variations in the treatment effectiveness of PMTO is needed, and would contribute to our knowledge about how to adjust the treatment to the specific needs of the individual family in order to improve the intervention. Furthermore, the presence of CU traits should be investigated in PMTO effectiveness trials, because research has shown that the presence of CU traits in children with externalizing behavior problems seems to have a negative impact on treatment outcome and there are tentative indications that some components of parent management training are more effective in these children than others. In order to meet the specific needs of children with CU traits, accurate screening of CU traits in children referred for parent management training is necessary. Considering the fact that CU traits are more likely to improve when addressed early in childhood, screening of these traits at a young age is recommended. Finally, the effectiveness of PMTO has primarily been investigated in Caucasian samples and should also be examined in other ethnic populations. This is relevant as PMTO has not been specifically developed for families with other ethnic backgrounds and research has demonstrated that ethnic minority families are less likely to be referred to or attend interventions (Lau, 2006). Therefore, cultural adaptation of the PMTO manual is needed to improve engagement and acceptability in order to achieve better retention rates of ethnic minority families. The PMTO manual has been adapted to meet the cultural patterns, meanings, and values of Mexican (Baumann, Rodríguez, Amador, Forgatch, & Parra-Cardona, 2014) and of Somali and Pakistani (Bjørknes & Manger, 2013) families. The effectiveness of the adapted version of PMTO has not yet been examined for the Mexican families, but has been demonstrated to improve parenting skills and mother-reported child problem behavior in Somali and Pakistani families in Norway (Bjørknes & Manger, 2013). 121 7 Summary S Summary | Externalizing behavior problems in children can have significant negative long-term consequences and constitute a major social and financial burden for the community. Hence, it is important to intervene as soon as these behavior problems occur. Ineffective parenting is one of the well-established risk factors for externalizing behavior problems in children. Therefore, many interventions for childhood externalizing problems focus on improving parenting skills. Parent Management Training –Oregon model (PMTO) is an intervention that focuses on parents of children between 4 and 12 years of age with severe externalizing behavior problems. In the current thesis, a series of studies is described that examined the effectiveness of PMTO in The Netherlands as well as factors that could have an influence on the effectiveness of parent management trainings in general, and PMTO in particular. The thesis consists of two parts. Part I contains three chapters on the effectiveness of PMTO in reducing externalizing problem behavior in children in The Netherlands. Part II consists of two chapters concerning emotional memory in children with callous-unemotional traits. The first part concerns research that examined the effectiveness of PMTO as an intervention for clinically referred children with externalizing behavior problems in The Netherlands. In Chapter 2, PMTO was compared with Care As Usual (CAU), which consisted of a mix of other viable interventions employed for treating children with externalizing behavior problems. Families were recruited at five different youth care agencies throughout The Netherlands. Of the 146 children and their parents included in the study, 91 received PMTO while 55 were given CAU. Families were assessed at four time points: at pretreatment, and after 6, 12, and 18 months. Results showed that PMTO was effective in reducing children’s problem behaviors, increasing the use of effective parenting practices, and alleviating parenting stress and psychopathological symptoms of the parents. However, PMTO appeared to be no more effective than CAU. For both treatments, the improvements were most evident during the first six months of the study and remained stable until the 18-months follow-up assessment. Additionally, we found that especially children with serious externalizing behavior problems at baseline benefited from PMTO. Comparing the effect size of PMTO delivered in The Netherlands with previous PMTO effectiveness studies in Norway and Iceland, we demonstrated that PMTO had a similar, large effect size as shown in previous studies. From these findings it can be concluded that PMTO is effective in a clinically referred sample of children with externalizing problems in the Netherlands, although it seems to be no more effective than CAU. The finding that PMTO was not more effective in reducing externalizing behavior problems than CAU raises questions about possible factors that may have influenced the effectiveness of PMTO in The Netherlands. For example, prior research found evidence to suggest that higher PMTO treatment fidelity might predict larger improvements 125 S | Summary in parenting skills and externalizing behavior problems. In Chapter 3, treatment fidelity was examined as a possible determinant of the effectiveness of PMTO. Data of 86 PMTO families from the effectiveness study presented in Chapter 2 were used. Results showed that treatment fidelity scores of certified PMTO therapists as obtained before treatment (i.e., at certification) were significantly associated with the outcome measures. The higher the fidelity scores of the therapists, the larger the improvements on externalizing behavior problems, parenting practices, parenting stress, and parental psychopathological symptoms. Interestingly, the strongest associations with fidelity were found for the outcome assessment at the 18-months follow-up. This indicates that very close adherence to the PMTO method is especially important for the longer term sustenance of the treatment effects of PMTO. In addition, we found that parents who completed the treatment more often had a therapist who adhered more closely to the PMTO protocol than families who dropped out. However, parents were not more satisfied with their therapist and the intervention when their therapist adhered more strictly to PMTO guidelines, revealing that treatment satisfaction is largely independent of how well the therapist delivers the PMTO treatment. In our study examining the effectiveness of PMTO (Chapter 2), we found that the parenting domains measured by our self-report measure of parenting strategies, the Caregiver Wish List (CWL), showed poor internal consistency. That is, reliability was only found to be adequate for the CWL total score and for one of its domains. Thus, we decided to examine the psychometric properties of the CWL in more detail in the study presented in Chapter 4. More precisely, the factor structure of the instrument as well as its reliability and validity were explored in a sample of 348 parents of children aged between 4 and 11 years, of which 220 were parents of children from the general population and 128 were parents of clinically referred children with externalizing behavior problems. Factor analyses indicated that a five factor solution (obtained after eliminating quite a number of unsatisfactory items) provided the best fit. The five factors were labelled as Adequate discipline, Controlled responding, Focus on positive behavior, Consistency, and Monitoring, and all showed adequate internal consistency coefficients. The correlations among the factors were rather low, suggesting that they represent fairly independent parenting domains. Furthermore, the factors differentiated between the community and clinical sample, with significantly better parenting skills observed for the community sample. Finally, all factors were significantly and negatively related to child behavior problems, with stronger correlations found for externalizing compared to internalizing problems, which provides support for the external validity of this parenting measure. 126 Summary | The second part of this dissertation describes two studies that focus on the emotional memory of children with CU traits, that is, children who are particularly prone to develop externalizing psychopathology. In Chapter 5, a self-developed memory test was administered to a community sample of children between 8 and 12 years of age with high (n = 24) and low (n = 18) CU traits. The peripheral memory test assesses memory performance for central and peripheral components of emotionally neutral and negative pictures. We found that overall recognition rates did not differ between the high and low CU groups. For negative pictures, both groups demonstrated better recognition of the central components at the expense of the peripheral components, while for neutral pictures, the peripheral components were better recognized than the central components. These results suggest that there are no differences in emotional memory between children with high and low CU traits, implying that children with CU traits do not have an impaired emotional memory. Chapter 6 describes a study in which emotional false memory was examined by use of the Deese-Roediger-McDermott paradigm. Children high and low on CU traits were subjected to five neutral and five emotionally negative word lists. Analyses revealed that true recall was higher for the neutral word lists than for the negative word lists in both the high and low CU groups. For false recall, memory performance did not differ between the high and low CU groups on the neutral word lists, but the high CU group recalled fewer unpresented words than the low CU group on the negative word lists. Moreover, the low CU group showed no difference in false recall between the neutral and negative word lists, while the high CU group children had fewer false memories on the negative word lists compared to the neutral word lists. The results indicate that children with high CU traits are better at differentiating between true and false memories concerning negative material than children with low CU traits. Therefore, children with high CU traits seem to be at lower risk of inaccurate memories. The final chapter of this dissertation (Chapter 7) contains a general discussion of the main findings of the studies in the preceding chapters. Furthermore, this chapter outlines methodological issues of the research presented in this dissertation and implications for clinical practice that follow from these results. The main message of this chapter, and of this dissertation as a whole, is that PMTO is effective in improving child externalizing behavior problems, effective parenting practices, parenting stress, and parental psychopathological symptoms, but not more effective than CAU. However, the effectiveness of PMTO is related to the treatment integrity of the PMTO therapist.When PMTO therapists adhered closely to the program principles, more positive long-term treatment effects were obtained. Furthermore, this chapter concludes that an impaired emotional memory is not an explanation for why children with high CU traits are less responsive to an intervention that focuses on effective parenting than children with low CU traits. 127 S Samenvatting S Samenvatting | Kinderen met externaliserende gedragsproblemen worden gekenmerkt door gedragingen als vechten, vernielen, ongehoorzaamheid, spijbelen, liegen en stelen. Deze gedragsproblemen kunnen op de lange termijn leiden tot verdere ontsporing in crimineel gedrag en vormen zo een grote sociale en financiële last voor de samenleving. Daarom is het belangrijk om zo vroeg mogelijk in te grijpen wanneer deze gedragsproblemen zich voordoen. Ineffectieve opvoeding is één van de meest onderzochte en vastgestelde risicofactoren voor externaliserende gedragsproblemen bij kinderen. Veel behandelingen richten zich dan ook op het verbeteren van opvoedingsvaardigheden van ouders. Parent Management Training – Oregon model (PMTO) is een training die ontwikkeld is voor ouders van kinderen tussen de 4 en 12 jaar met ernstige externaliserende gedragsproblemen. PMTO is gebaseerd op het Social Interaction Learning (SIL) model. Het SIL model veronderstelt dat de opvoedingsvaardigheden van ouders een direct effect hebben op de ontwikkeling van het kind. Contextuele factoren, zoals een lage socio-economische status en psychopathologie bij ouders, zouden een negatieve invloed hebben op de kwaliteit van opvoeding en daarmee een indirect negatief effect op de ontwikkeling van het kind. In dit proefschrift wordt een aantal studies beschreven naar de effectiviteit van PMTO in Nederland en naar factoren die mogelijk de effectiviteit van PMTO beïnvloeden. Dit proefschrift bestaat uit twee delen. Het eerste deel bevat drie hoofdstukken over de effectiviteit van PMTO in Nederland. Het tweede deel bestaat uit twee hoofdstukken over het emotioneel geheugen van kinderen met psychopathische trekken. In Hoofdstuk 2 wordt PMTO vergeleken met Care As Usual (CAU). CAU bestaat uit een mix van andere interventies die worden aangeboden aan kinderen met externaliserende gedragsproblemen en hun ouders. Van de 146 gezinnen die aan de studie deelnamen, kregen er 91 PMTO en 55 CAU. De uitkomstmetingen van deze studie vonden op vier momenten plaats: vóór de start van de behandeling (voormeting) en vervolgens na 6, 12 en 18 maanden. PMTO bleek effectief in het verminderen van externaliserende gedragsproblemen bij kinderen, het verbeteren van de effectieve opvoedingsvaardigheden en het verlichten van de opvoedingsstress en lichamelijke en psychische klachten (bijv. angst, depressie, slaapproblemen) bij de ouders. PMTO bleek echter niet significant effectiever dan CAU. Voor beide behandelingen waren de verbeteringen het grootste tijdens de eerste zes maanden en deze bleven vervolgens stabiel tot de laatste meting na 18 maanden. Verder vonden we dat vooral kinderen met ernstige externaliserende gedragsproblemen, zoals gemeten bij de voormeting, meer baat hadden bij de PMTO behandeling. Uit deze resultaten kan geconcludeerd worden dat PMTO effectief is voor kinderen met externaliserende gedragsproblemen in Nederland, maar PMTO blijkt niet effectiever dan het al bestaande behandelaanbod voor deze kinderen (CAU). 131 S | Samenvatting Aangezien PMTO niet effectiever bleek te zijn dan CAU in het verminderen van externaliserende gedragsproblemen, hebben we onderzocht welke factoren mogelijk invloed hebben op de effectiviteit van PMTO. Eén van die factoren is de behandelintegriteit van therapeuten. Behandelintegriteit is de mate waarin de therapeut de behandelprincipes van een interventie toepast zoals deze bedoeld zijn. Eerder onderzoek heeft aangetoond dat een grotere behandelintegriteit gepaard gaat met sterkere verbeteringen in opvoedingsvaardigheden en externaliserende gedragsproblemen. In Hoofdstuk 3 werd onderzocht of behandelintegriteit invloed heeft op de effectiviteit van PMTO in Nederland. Hier werden gegevens van 86 PMTO gezinnen gebruikt die hadden deelgenomen aan de effectiviteitsstudie uit Hoofdstuk 2. Alle PMTO therapeuten die de behandelingen gaven waren in het kader van hun certificering, voorafgaand aan ons effectiviteitsonderzoek, gescoord op behandelintegriteit. Deze behandelintegriteitsscores hingen samen met de uitkomstmaten: hoe hoger de scores van therapeuten op behandelintegriteit, hoe sterker de verbeteringen op externaliserende gedragsproblemen van de kinderen, en opvoedingsvaardigheden, opvoedingsstress en lichamelijk en psychische klachten van de ouders. Een opmerkelijk resultaat was dat de sterkste samenhang met behandelintegriteit werd gevonden voor de laatste meting na 18 maanden. Dit geeft aan dat het sterk vasthouden aan de behandelprincipes vooral belangrijk is voor het lange termijn behoud van het behandeleffect van PMTO. Verder vonden we dat ouders die de behandeling afgemaakt hebben een therapeut met hogere behandelintegriteit hadden dan ouders die voortijdig waren gestopt. Ouders bleken echter niet meer tevreden over hun therapeut wanneer de therapeut zich strakker aan de PMTO richtlijnen hield. Dit betekent dat de behandeltevredenheid van ouders los staat van hoe goed de therapeut de PMTO behandeling geeft. In de studie waarin de effectiviteit van PMTO werd onderzocht (Hoofdstuk 2) maakten we onder andere gebruik van de Caregiver Wish List (CWL), een gestandaardiseerd interview met ouders voor het meten van de kwaliteit van hun opvoedingsvaardigheden, zoals door henzelf beoordeeld. De CWL meet zes verschillende opvoedingsdomeinen, zoals toezicht houden en positieve betrokkenheid bij het kind, maar deze bleken een lage betrouwbaarheid (interne consistentie) te hebben. Dit betekent dat de vragen in de verschillende opvoedingsdomeinen niet hetzelfde opvoedingsaspect meten. In onze effectiviteitsstudie bleek de betrouwbaarheid van de CWL alleen voor de CWL totaal score en voor één van de zes opvoedingsdomeinen adequaat te zijn. Daarom besloten we om de psychometrische kenmerken van de CWL meer gedetailleerd te onderzoeken in de studie beschreven in Hoofdstuk 4. De factor structuur, betrouwbaarheid en validiteit van het instrument werden onderzocht in een steekproef van 348 ouders van kinderen tussen de 4 en 11 jaar. Hiervan kwamen 220 ouders uit de algemene 132 Samenvatting | bevolking (zij werden via basisscholen geworven) en 128 ouders van kinderen die waren doorverwezen vanwege externaliserende gedragsproblemen (klinische populatie). Factoranalyses gaven aan dat een vijf-factoren-oplossing het beste bij de CWL-gegevens paste. De vijf factoren kregen de labels Adequate discipline, Beheerst reageren, Focus op positief gedrag, Consistentie en Toezicht houden en lieten allemaal een goede interne betrouwbaarheid zien. De correlaties tussen de factoren waren redelijk laag, wat betekent dat ze grotendeels onafhankelijke opvoedingsvaardigheden weergeven. Verder vonden we een verschil in scores op de vijf factoren tussen de twee groepen ouders. Ouders uit de algemene populatie rapporteerden, zoals verwacht, betere opvoedingsvaardigheden dan ouders uit de klinische populatie. Bovendien waren alle factoren negatief gerelateerd aan gedragsproblemen: hoe beter de opvoedingsvaardigheden, hoe minder de gedragsproblemen. Deze verbanden waren sterker voor externaliserende dan voor internaliserende (angstig, teruggetrokken, depressief ) gedragsproblemen. Deze resultaten bieden ondersteuning voor de externe validiteit van de CWL: het instrument lijkt vijf belangrijke opvoedingsvaardigheden accuraat te kunnen meten. Het tweede deel van dit proefschrift beschrijft studies die ingaan op het emotioneel geheugen van kinderen met psychopathische trekken. Kinderen met deze trekken kennen weinig tot geen angst, zijn impulsief, bekommeren zich niet om de gevoelens van anderen en ervaren geen schuldgevoelens. Deze kinderen lopen een verhoogd risico op het ontwikkelen van serieuze antisociale gedragsproblemen in de adolescentie en volwassenheid. Eerder onderzoek heeft aangetoond dat effectieve opvoedingsvaardigheden ongerelateerd zijn aan de gedragsproblemen van kinderen met psychopathische trekken. Een beperking van het emotioneel geheugen zou een verklaring kunnen zijn voor waarom deze kinderen minder goed leren van straf. Voor de studie in Hoofdstuk 5 is een zelfontworpen geheugentest afgenomen bij kinderen tussen de 8 en 12 jaar met een hoge (n = 24) versus een lage (n = 18) score op psychopathische trekken. De geheugentest meet de geheugenprestatie voor centrale en perifere (aan de rand) details van neutrale en emotioneel negatief beladen plaatjes. We vonden dat herkenning van de details niet verschilde tussen de hoge en lage psychopathie groepen. Voor de negatieve plaatjes herkenden beide groepen de centrale details beter in vergelijking met de perifere details. Bij de neutrale plaatjes werden juist de perifere details beter herkend dan de centrale details. Deze resultaten suggereren dat er geen verschil is in emotioneel geheugen tussen kinderen met hoge en lage psychopathische trekken en daarmee dat kinderen met psychopathische trekken geen beperkt emotioneel geheugen hebben. Dit suggereert dat een beperkt emotioneel geheugen niet als verklaring kan dienen voor waarom effectieve opvoedingsvaardigheden minder effect hebben op kinderen met psychopathische trekken. 133 S | Samenvatting Hoofdstuk 6 beschrijft een studie waarin emotioneel geheugen werd onderzocht met behulp van het Deese-Roediger-McDermott paradigma. In dit paradigma krijgen kinderen een lijst met woorden te horen (bijv. bakker, boter, beleg, etc.) die ze moeten proberen te onthouden. Deze woorden zijn allemaal geassocieerd aan een nietgepresenteerd themawoord (brood). Wanneer kinderen vervolgens, op de vraag welke woorden zij gehoord hebben, het thema woord noemen (brood), is er sprake van een pseudoherinnering. We lieten kinderen met hoge en lage psychopathische trekken vijf emotioneel neutrale en vijf negatieve woordlijsten horen. Uit de analyses kwam naar voren dat beide groepen meer neutrale dan negatieve woorden hadden onthouden. We vonden geen verschillen tussen de groepen voor de neutrale woordlijsten wat betreft het aantal pseudoherinneringen. Voor de negatieve woordlijsten vonden we echter dat de kinderen met hoge psychopathische trekken minder pseudoherinneringen opnoemden dan de kinderen met lage psychopathische trekken. Verder liet de laag psychopathische groep geen verschil zien in pseudoherinneringen op de negatieve en neutrale woordlijsten, terwijl de hoog psychopathische groep minder pseudoherinneringen had voor de negatieve woordlijsten in vergelijking met de neutrale woordlijsten. Deze resultaten geven aan dat kinderen met hoge psychopathische trekken beter zijn in het onderscheiden van echte en pseudoherinneringen voor negatief materiaal dan kinderen met lage psychopathische trekken. Kinderen met hoge psychopathische trekken lijken daarmee minder vatbaar te zijn voor het ontwikkelen van inaccurate herinneringen voor negatieve gebeurtenissen. Een verklaring hiervoor kan zijn dat kinderen met psychopathische trekken een negatieve gebeurtenis als neutraal beschouwen en daardoor minder waarschijnlijk zijn om de gebeurtenis te herevalueren, waardoor ze beschermd worden tegen pseudoherinneringen. Het laatste hoofdstuk (Hoofdstuk 7) bevat een discussie van de belangrijkste bevindingen van de studies uit de voorgaande hoofdstukken. De kernboodschap van dit hoofdstuk, en daarmee van dit gehele proefschrift, is dat PMTO effectief is in het verminderen van externaliserende gedragsproblemen bij kinderen, vergroten van effectieve opvoedingsvaardigheden, en verminderen van opvoedingsstress en lichamelijke en psychische klachten bij ouders, maar niet effectiever dan standaard zorgaanbod. De effectiviteit van PMTO hangt samen met de behandelintegriteit van de PMTO therapeut. Wanneer de PMTO therapeut zich streng aan de behandelprincipes houdt, leidt dat tot grotere lange termijn effecten. Verder concluderen we dat een zwak emotioneel geheugen niet als verklaring kan dienen voor de bevinding uit eerdere onderzoeken dat kinderen met hoge psychopathische trekken minder goed reageren op een behandeling die gericht is op het versterken van de effectieve opvoedingsvaardigheden bij hun ouders. 134 Valorization addendum V Valorization addendum | The present dissertation describes the results of a large multi-center study examining the effectiveness of Parent Management Training - Oregon model (PMTO) for parents of children with externalizing behavior problems. This dissertation also studied factors that could have an influence on the effectiveness of parent management training in general, and PMTO in particular. This valorization addendum discusses the findings in a broader societal and economical context. Relevance of the research More than 13% of the children and adolescents in The Netherlands show externalizing behavior problems (de Looze et al., 2014). These youth are at increased risk of later antisocial behavior (Coté, Vaillancourt, LeBlanc, Nagin, & Tremblay, 2006; Simonoff et al., 2004). Moreover, the earlier externalizing behavior problems begin, the higher the risk of serious problems during adulthood (Moffitt, 1993; Patterson, DeBaryshe, & Ramsey, 1989). When externalizing behavior problems are not properly addressed, about 50 to 75% of the youth will continue to show serious problem behaviors (Nixon, 2002). Behavior problems are not only a social burden to the community, but also a financial one. For example, Scott and colleagues (2001) found that by the age of 28, the costs for individuals who had conduct disorder as a child were 10 times higher than for persons without psychiatric problems and 3.5 times higher than for individuals who as a child had displayed only mild behavioral problems. Therefore, it is important to intervene as soon as a child shows signs of externalizing behavior problems. In The Netherlands, there was a need for effective treatment programs for children 4-12 years with antisocial conduct problems. Already in 2005, the Ministry of Health, Welfare and Sport decided to fund the implementation of Parent Management Training - Oregon model (PMTO), but required that its effectiveness would be evaluated. This treatment had been found to be effective in other countries, but it could not be automatically assumed that it would also be effective in The Netherlands. Further, it is not only important to examine if a certain treatment is effective, but also for whom and under which circumstances. For example, research has shown that children with externalizing behavior problems who also exhibit callous-unemotional (CU) traits have an earlier onset of offending, display higher levels of aggression and delinquency, and more frequently come into contact with the police as compared to children with externalizing behaviour problems who do not exhibit these traits (Andershed, Gustafson, Kerr, & Stattin, 2002; Frick, Cornell, Barry, Bodin, & Dane, 2003; Frick, Stickle, Dandreaux, Farrell, & Kimonis, 2005; Frick & White, 2008; Lawing, Frick, & Cruise, 2010). Although not examined directly in this dissertation, there is evidence to suggest that youth with a combination of externalizing problems and CU traits are less responsive to effective parenting strategies and to treatments focusing on the improvement of parenting 137 V | Valorization addendum skills. Further exploration of these issues would enable us to improve interventions for children with externalizing problems and if necessary to adjust treatments to the specific needs of these youngsters and their families. The studies in this thesis empirically tested a number of factors that might influence the effectiveness of PMTO specifically and parenting trainings in general. The general discussion in Chapter 7 gives a more detailed description of the theoretical, clinical, and research implications of our research and provides directions for future research. To whom the results of this dissertation may be relevant First, the results are relevant for parents of children with externalizing behavior problems and their social environment. Our studies show that PMTO and other interventions available in The Netherlands are effective in reducing externalizing behavior problems of children. Furthermore, this dissertation contributes to the knowledge about factors that could influence the effectiveness of such treatments, which could be used to optimize evidence-based interventions for children with externalizing behavior problems. Second, our results are also interesting for PMTO therapists and PMTO supervisors, especially the findings on treatment fidelity of PMTO as described in Chapter 3. Although previous research had already demonstrated that fidelity scores were predictive of positive change in parenting skills and externalizing behavior problems, our findings show that high treatment fidelity is especially important for the long-term effectiveness of PMTO. These results may also be relevant for clinicians conducting other interventions for externalizing behavior problems or even clinicians carrying out treatments for other psychological problems, such as depression, anxiety disorders, and substance use disorders. In The Netherlands, clinicians are used to working according to a treatment protocol. However, only a few of these treatments have integrated fidelity checks as part of the program. This means that therapists may drift from the treatment principles and/or may only use elements of the protocol. The results of our study indicate that it is important to check treatment adherence in order to optimize the long-term effectiveness of the interventions. Third, since the findings from this dissertation may contribute to a reduction of the social and financial burden caused by children’s externalizing behavior problems on the community, the gained insights are also of interest to policy makers and health insurance companies. Finally, this dissertation may also be relevant for researchers in child and adolescent mental health. A randomized controlled trial (RCT) is considered the gold standard for evaluating the effectiveness of treatment interventions. However, the study presented in Chapter 2 shows how difficult it is to conduct a RCT in regular clinical practice. Our study started as a RCT, but we encountered several organizational problems which eventually forced us to continue our study with a quasi-experimental design. These problems are not 138 Valorization addendum | uncommon when conducting intervention research in a clinical setting and underline the need to consider alternative research designs (Langendam, Hooft, Heus, Scholten, & Bossuyt, 2013). Activities and services The findings of this dissertation show that PMTO is effective in improving child externalizing behavior problems and that it seems to be as effective as other interventions for children with externalizing behavior problems in The Netherlands. This knowledge can be used in decision making about which treatment to recommend for which family. For example, the choice of the most suitable intervention could be based on the resources of the families, such as the availability of time and transportation. However, this recommendation is still premature, since evidence of the cost-effectiveness and the long-term (i.e., > 10 years) effectiveness of the treatments is needed before firm conclusions about relative effectiveness can be drawn. Furthermore, in Chapter 7 of this dissertation, the importance of screening for CU traits is addressed, since this characteristic could influence the effects of parenting treatment. To achieve optimal change in externalizing behavior problems in children with CU traits, a focus on positive reinforcement and parental warmth may be particularly indicated. This dissertation further demonstrates that the quality of interventions for children with externalizing behavior problems in The Netherlands is quite good. This implies that the focus should not necessarily be on implementing new interventions for this target group, but on improving and optimizing the current treatments and in optimizing their treatment fidelity. For example, the study presented in Chapter 3 provides evidence for the influence of treatment integrity on the effectiveness of PMTO. This knowledge could also be applied to other interventions in order to further increase their effectiveness. The present dissertation found support for Social Interaction Learning (SIL) theory. This model states that parents have a direct influence on the behaviors of the child, implying that by improving the parenting strategies of parents, the behavior problems of the child will be reduced. This insight underlines the influence that parents have on their children and this might not only apply to externalizing behavior problems, but to other problems as well. For example, it has already been shown that the quality of parenting is related to dental caries in children (de Jong-Lenters, Duijster, Bruist, Thijssen, & de Ruiter, 2014; Duijster et al., 2015). Therefore, improving effective parenting skills could also be an effective mechanism in improving other child outcomes. 139 V | Valorization addendum Innovation of the research The study on the effectiveness of PMTO in The Netherlands (Chapter 2) is not necessarily innovative. The effectiveness of PMTO had already been examined in other countries and our research design is quite similar to that applied in these other studies. We used a longitudinal design and compared families receiving PMTO with other available interventions by using a multimethod and multi-informant approach. This research was necessary because the effectiveness of interventions should also be evaluated in The Netherlands, in order to be registered in the Dutch Youth Institute (Nederlands Jeugdinstituut; NJi) database of effective interventions. Furthermore, the similarity in research design with other PMTO studies allows for a direct comparison of the results. The other studies described in this dissertation are more innovative. First, the study on treatment integrity of PMTO therapists (Chapter 3) used a different approach compared to previous studies. For example, earlier studies examined whether treatment fidelity was predictive of only one outcome measure (i.e., parenting practices or child behavior problems), while we also assessed the association between fidelity and parental stress and psychopathological symptoms. In addition, our study used fidelity scores of certified PMTO therapists, obtained prior to treatment, and examined the association with effectiveness at different time points. Based on this innovative design, we were able to show that even for certified PMTO therapists, treatment fidelity varied systematically with effectiveness on all our outcome measures and that very close adherence to the PMTO method is especially important for the endurance of the effects produced by PMTO. Second, the study on the Caregiver Wish List (CWL) described in Chapter 4 is innovative because no other study can be found in the literature examining the psychometric properties of this instrument. Therefore, our study was the first to examine its factor structure, reliability, and validity. Finally, the studies described in Chapters 5 and 6 also used an innovative approach. So far, only three studies examined emotional memory in participants with psychopathic (CU) traits, but none of them focused on children. The study presented in Chapter 5 was the first to theoretically relate emotional memory in children with CU-traits to unresponsiveness to effective parenting and parenting training, and used a self-developed memory test which already received interest from other researchers who investigate emotional memory in children. For example, our peripheral memory test has been used to examine whether induced stress affects emotional memory in 5- to 9-year old children (R. Neuenschwander, personal communication, June 10, 2015). Emotional false memory in children with CU-traits (see Chapter 6) is a new research area, which could contribute to our knowledge on the role of memory in learning and maintenance of antisocial cognitions. 140 Valorization addendum | Knowledge dissemination The outcomes of the studies described in this dissertation have been, and will be, communicated in several ways. The studies presented in Chapters 2, 3, and 4 have recently been finished and are submitted to international peer reviewed journals. The studies described in Chapters 5 and 6 have already been published in international journals. Furthermore, the research in Chapter 2, 5, and 6 has been presented at several national and international conferences, which were attended by both researchers and professionals from the clinical field. Additionally, results of the studies presented in this dissertation were communicated through lectures given to university students at Maastricht University. Communication of the findings in this thesis outside the scientific field will be done through a press release. V 141 References R References | Achenbach, T. M. (1991). Manual for the Child Behavior Checklist/4-18. Burlington: University of Vermont, Department of Psychiatry. Alink, L. R. A., Mesman, J., van Zeijl, J., Stolk, M. N., Juffer, F., Koot, H. M., . . . van Ijzendoorn, M. H. (2006). The early childhood aggression curve: Development of physical aggression in 10- to 50-month-old children. Child Development, 77, 954-966. doi: 10.1111/j.1467-8624.2006.00912.x American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: American Psychiatric Association. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: American Psychiatric Association. Amlund-Hagen, K., Ogden, T., & Bjørnebekk, G. (2011). Treatment outcomes and mediators of parent management training: A one-year followup of children with conduct problems. Journal of Clinical Child and Adolescent Psychology, 40, 165178. doi: 10.1080/15374416.2011.546050 Anastassiou-Hadjicharalambous, X., & Warden, D. (2008). Physiologically-indexed and selfperceived affective empathy in conductdisordered children high and low on callousunemotional traits. Child Psychiatry and Human Development, 39, 503-517. doi: 10.1007/s10578008-0104-y Andershed, H. A., Gustafson, S. B., Kerr, M., & Stattin, H. k. (2002). The usefulness of self-reported psychopathy-like traits in the study of antisocial behaviour among non-referred adolescents. European Journal of Personality, 16, 383-402. doi: 10.1002/per.455 Anderson, C. A., & Bushman, B. J. (2002). Human aggression. Annual Review of Psychology, 53, 27-51. doi: 10.1146/annurev. psych.53.100901.135231 Arrindell, W. A., & Ettema, J. M. H. (1986). SCL90: Handleiding bij een Multidimensionele Psychopathologie-Indicator [Manual for a Multidimensional Psychopathology Indicator]. Lisse, The Netherlands: Swets & Zeitlinger. Arrindell, W. A., & Ettema, J. M. H. (2003). Symptom Checklist. Handleiding bij een multidimensionele psychopathologie-indicator [Manual for a Multidimensional Psychopathology Indicator] (2nd edition). Lisse, The Netherlands: Swets Test Publishers. Bank, L., Marlowe, J. H., Reid, J. B., & Patterson, G. R. (1991). A comparative evaluation of parent-training interventions for families of chronic delinquents. Journal of Abnormal Child Psychology, 19, 15-33. doi: 10.1007/BF00910562 Barzi, F., & Woodward, M. (2004). Imputation of missing values in practice: Results from imputation of serum cholestorol in 28 cohort studies. American Journal of Epidemiology, 160, 34-45. doi: 10.1093/aje/kwh175 Bates, J. E., Pettit, G. S., Dodge, K. A., & Ridge, B. (1998). Interaction of temperamental resistance to control and restrictive parenting in the development of externalizing behavior. Developmental Psychology, 34, 982-995. doi: 10.1037/0012-1649.34.5.982 Bauer, S., Lambert, M. J., & Nielsen, S. L. (2004). Clinical significance methods: A comparison of statistical techniques. Journal of Personality Assessment, 82, 60-70. doi: 10.1207/s15327752jpa8201_11 Baumann, A. A., Rodríguez, M. M. D., Amador, N. G., Forgatch, M. S., & Parra-Cardona, J. R. (2014). Parent Management Training-Oregon model (PMTOTM) in Mexico City: Integrating cultural adaptation activities in an implementation model. Clinical Psychology: Science and Practice, 21, 32-47. doi: 10.1111/cpsp.12059 Bennett, K. J., Lipman, E. L., Brown, S., Racine, Y., Boyle, M. H., & Offord, D. R. (1999). Predicting conduct problems: Can high-risk children be identified in kindergarten and Grade 1? Journal of Consulting and Clinical Psychology, 67, 470480. doi: 10.1037/0022-006X.67.4.470 Berger, M., & van Everdingen, J. (2006). De positionering van PMTO ten opzichte van andere intensieve ouderinterventies in Nederland. Utrecht: NIZW. Bierman, K. L., & Smoot, D. L. (1991). Linking family characteristics with poor peer relations: The mediating role of conduct problems. Journal of Abnormal Child Psychology, 19, 341-356. doi: 10.1007/BF00911236 Bijttebier, P., & Decoene, S. (2009). Assessment of psychopathic traits in children and adolescents: Further validation of the Antisocial Process Screening Device and the Childhood Psychopathy Scale. European Journal of Psychological Assessment, 25, 157-163. Bjørknes, R., & Manger, T. (2013). Can parent training alter parent practice and reduce conduct problems in ethnic minority children? A randomized controlled trial. Prevention Science, 14, 52-63. doi: 10.1007/s11121-012-0299-9 145 R | References Blair, R. J. R. (1999). Responsiveness to distress cues in the child with psychopathic tendencies. Personality and Individual Differences, 27, 135145. Blair, R. J. R., Colledge, E., Murray, L., & Mitchell, D. G. V. (2001). A selective impairment in the processing of sad and fearful expressions in children with psychopathic tendencies. Journal of Abnormal Child Psychology, 29, 491-498. doi: 10.1023/A:1012225108281 Brainerd, C. J., Holliday, R. E., Reyna, V. F., Yang, Y., & Toglia, M. P. (2010). Developmental reversals in false memory: Effects of emotional valence and arousal. Journal of Experimental Child Psychology, 107, 137-154. doi: 10.1016/j.jecp.2010.04.013 Brainerd, C. J., & Reyna, V. F. (2012). Reliability of children’s testimony in the era of developmental reversals. Developmental Review. doi: 10.1016/j. dr.2012.06.008 Brainerd, C. J., Reyna, V. F., & Ceci, S. J. (2008a). Developmental reversals in false memory: A review of data and theory. Psychological Bulletin, 134, 343-382. doi: 10.1037/0033-2909.134.3.343 Brainerd, C. J., Reyna, V. F., & Zember, E. (2011). Theoretical and forensic implications of developmental studies of the DRM illusion. Memory & Cognition, 39, 365-380. doi: 10.3758/ s13421-010-0043-2 Brainerd, C. J., Stein, L. M., Silveira, R. A., Rohenkohl, G., & Reyna, V. F. (2008b). How does negative emotion cause false memories? Psychological Science, 19, 919-925. doi: 10.1111/j.14679280.2008.02177.x Brewin, C. R., & Bradley, C. (1989). Patient preferences and randomised clinical trials. British Medical Journal, 299, 313-315. Broidy, L. M., Nagin, D. S., Tremblay, R. E., Bates, J. E., Brame, B., Dodge, K. A., . . . Vitaro, F. (2003). Developmental trajectories of childhood disruptive behaviors and adolescent delinquency: A six-site, cross-national study. Developmental Psychology, 39, 222-245. doi: 10.1037/0012-1649.39.2.222 Buchanan, T. W. (2007). Retrieval of emotional memories. Psychological Bulletin, 133, 761-779. doi: 10.1037/0033-2909.133.5.761 Burke, A., Heuer, F., & Reisberg, D. (1992). Remembering emotional events. Memory & Cognition, 20, 277-290. 146 Burke, J. D., Loeber, R., & Lahey, B. B. (2007). Adolescent conduct disorder and interpersonal callousness as predictors of psychopathy in young adults. Journal of Clinical Child and Adolescent Psychology, 36, 334-346. doi: 10.1080/15374410701444223 Cahill, L., & McGaugh, J. L. (1995). A novel demonstration of enhanced memory associated with emotional arousal. Consciousness and Cognition, 4, 410-421. doi: 10.1006/ ccog.1995.1048 Charles, J. M., Bywater, T., & Edwards, R. T. (2011). Parenting interventions: A systematic review of the economic evidence. Child: Care, Health and Development, 37, 462-474. doi: 10.1111/j.13652214.2011.01217.x Christianson, S.-Ã. (1992). Emotional stress and eyewitness memory: A critical review. Psychological Bulletin, 112, 284-309. doi: 10.1037//0033-2909.112.2.284 Christianson, S.-Ã., Forth, A. E., Hare, R. D., Strachan, C., Lidberg, L., & Thorell, L.-H. (1996). Remembering details of emotional events: A comparison between psychopathic and nonpsychopathic offenders. Personality and Individual Differences, 20, 437-443. doi: 10.1016/0191-8869(95)00220-0 Christianson, S.-Ã., Loftus, E. F., Hoffman, H., & Loftus, G. R. (1991). Eye fixations and memory for emotional events. Journal of Experimental Psychology: Learning, Memory, and Cognition, 17, 693-701. Chronis, A. M., Lahey, B. B., Pelham, W. E., Williams, S. H., Baumann, B. L., Kipp, H., . . . Rathouz, P. J. (2007). Maternal depression and early positive parenting predict future conduct problems in young children with attention-deficit/hyperactivity disorder. Developmental Psychology, 43, 70-82. doi: 10.1037/0012-1649.43.1.70 Cirelli, L. K., Dickinson, J., & Poirier, M. (2014). Using implicit instructional cues to influence false memory induction. Journal of Psycholinguistic Research, 44, 485-494. doi: 10.1007/s10936-0149301-y Conger, R. D., Patterson, G. R., & Ge, X. (1995). It takes two to replicate: A mediational model for the impact of parents’ stress on adolescent adjustment. Child Development, 66, 80-97. doi: 10.2307/1131192 Corson, Y., & Verrier, N. (2007). Emotions and false memories: Valence or arousal? Psychological Science, 18, 208-211. doi: 10.1111/j.14679280.2007.01874.x References | Coté, S. M., Vaillancourt, T., LeBlanc, J. C., Nagin, D. S., & Tremblay, R. E. (2006). The development of physical aggression from toddlerhood to preadolescence: A nation wide longitudinal study of Canadian children. Journal of Abnormal Child Psychology, 34, 71-85. doi: 10.1007/s10802-0059001-z Cunningham, C. E., & Boyle, M. H. (2002). Preschoolers at risk for attention-deficit hyperactivity disorder and oppositional defiant disorder: Family, parenting, and behavioral correlates. Journal of Abnormal Child Psychology, 30, 555-569. doi: 10.1023/A:1020855429085 Dadds, M. R., Fraser, J., Frost, A., & Hawes, D. J. (2005). Disentangling the underlying dimensions of psychopathy and conduct problems in childhood: A community study. Journal of Consulting and Clinical Psychology, 73, 400-410. doi: 10.1037/0022-006X.73.3.400 Dadds, M. R., Maujean, A., & Fraser, J. A. (2003). Parenting and conduct problems in children: Australian data and psychometric properties of the Alabama Parenting Questionnaire. Australian Psychologist, 38, 238-241. doi: 10.1080/00050060310001707267 Deese, J. (1959). On the prediction of occurrence of particular verbal intrusions in immediate recall. Journal of Experimental Psychology, 58, 17-22. DeGarmo, D. S., & Forgatch, M. S. (2005). Early development of delinquency within divorced families: Evaluating a randomized preventive intervention trial. Developmental Science, 8, 229239. doi: 10.1111/j.1467-7687.2005.00412.x DeGarmo, D. S., & Forgatch, M. S. (2007). Efficacy of parent training for stepfathers: From playful spectator and polite stranger to effective stepfathering. Parenting: Science and Practice, 7, 331-355. doi: 10.1080/15295190701665631 DeGarmo, D. S., Patterson, G. R., & Forgatch, M. S. (2004). How do outcomes in a specified parent training intervention maintain or wane over time? Prevention Science, 5, 73-89. doi: 10.1023/B:PREV.0000023078.30191.e0 Derogatis, L. R. (1977). SCL-90-R: Administration, scoring & procedures manual-II for the R(evised) version. Baltimore: Clinical Psychometric Research. Dodge, K. A. (1980). Social cognition and children’s aggressive behavior. Child Development, 51, 162170. doi: 10.2307/1129603 Dadds, M. R., & Salmon, K. (2003). Punishment insensitivity and parenting: Temperament and learning as interacting risks for antisocial behavior. Clinical Child and Family Psychology Review, 6, 69-86. doi: 10.1023/A:1023762009877 Dolan, M. C., & Fullam, R. (2005). Memory for emotional events in violent offenders with antisocial personality disorder. Personality and Individual Differences, 38, 1657-1667. doi: 10.1016/j.paid.2004.09.028 de Brock, A. J. L. L., Vermulst, A. A., Gerris, J. R. M., & Abidin, R. R. (1992). Nijmeegse Ouderlijke Stress Index. Opvoeding in gezin en andere primaire leefvormen [Nijmegen Parenting Stress Index: Caregiving in the family and other primary care systems]. Lisse, The Netherlands: Swets & Zeitlinger. Dolan, M. C., & Fullam, R. (2010). Emotional memory and psychopathic traits in conduct disordered adolescents. Personality and Individual Differences, 48, 327-331. doi: 10.1016/j. paid.2009.10.029 de Jong-Lenters, M., Duijster, D., Bruist, M. A., Thijssen, J., & de Ruiter, C. (2014). The relationship between parenting, family interaction and childhood dental caries: A case-control study. Social Science & Medicine, 116, 49-55. doi: doi: 10.1016/j.socscimed.2014.06.031 and without release contingencies. Journal of Applied Behavior Analysis, 44, 693-705. doi: 10.1901/ jaba.2011.44-693 de Looze, M., Van Dorsselaer, S., de Roos, S., Verdurmen, J., Stevens, G., Gommans, R., . . . Vollebergh, W. (2014). HBSC 2013. Gezondheid, welzijn en opvoeding van jongeren in Nederland. Utrecht: Universiteit Utrecht. Parental and family-related influences on dental caries in children of Dutch, Moroccan and Turkish origin. Community Dentistry and Oral Epidemiology, 43, 152-162. doi: 10.1111/ cdoe.12134 DeBaryshe, B. D., Patterson, G. R., & Capaldi, D. M. (1993). A performance model for academic achievement in early adolescent boys. Developmental Psychology, 29, 795-804. doi: 10.1037/0012-1649.29.5.795 Eccles, M., Grimshaw, J., Campbell, M., & Ramsay, C. (2003). Research designs for studies evaluating the effectiveness of change and improvement trategies. Quality & Safety in Health Care, 12, 4752. doi: 10.1136/qhc.12.1.47 Donaldson, J. M., & Vollmer, T. R. (2011). An evaluation and comparison of time-out procedures with Duijster, D., de Jong-Lenters, M., de Ruiter, C., Thijssen, J., van Loveren, C., & Verrips, G. (2015). 147 R | References Ehrensaft, M. K., Wasserman, G. A., Verdelli, L., Greenwald, S., Miller, L. S., & Davies, M. (2003). Maternal antisocial behavior, parenting practices, and behavior problems in boys at risk for antisocial behavior. Journal of Child and Family Studies, 12, 27-40. doi: 10.1023/A:1021302024583 Eisenberger, N. I., Lieberman, M. D., & Williams, K. D. (2003). Does rejection hurt? An fMRI study of social exclusion. Science, 302, 290-292. doi: 10.1126/science.1089134 Escribano, S., Aniorte, J., & Orgilés, M. (2013). Factor structure and psychometric properties of the Spanish version of the Alabama Parenting Questionnaire (APQ) for children. Psicothema, 25, 324-329. doi: 10.7334/psicothema2012.315 Essau, C. A., Sasagawa, S., & Frick, P. J. (2006). Psychometric properties of the Alabama Parenting Questionnaire. Journal of Child and Family Studies, 15, 597-616. doi: 10.1007/s10826006-9036-y Everett, G. E., Olmi, D. J., Edwards, R. P., Tingstrom, D. H., Sterling-Turner, H. E., & Christ, T. J. (2007). An empirical investigation of timeout with and without escape extinction to treat escape-maintained noncompliance. Behavior Modification, 31, 412-434. doi: 10.1177/0145445506297725 Fairhurst, K., & Dowrick, C. (1996). Problems with recruitment in a randomized controlled trial of counselling in general practice: Causes and implications. Journal of Health Services Research and Policy, 1, 77-80. doi: 10.1177/135581969600100205 Fisher, P. A., Burraston, B., & Pears, K. (2005). The Early Intervention Foster Care Program: Permanent placement outcomes from a randomized trial. Child Maltreatment, 10, 61-71. doi: 10.1177/1077559504271561 Fite, P. J., Greening, L., Stoppelbein, L., & Fabiano, G. A. (2009). Confirmatory factor analysis of the Antisocial Process Screening Device with a clinical inpatient population. Assessment, 16, 103-114. doi: 10.1177/1073191108319893 Fontaine, N. M. G., McCrory, E. J. P., Boivin, M., Moffitt, T. E., & Viding, E. (2011). Predictors and outcomes of joint trajectories of callous-unemotional traits and conduct problems in childhood. Journal of Abnormal Psychology, 120, 730-742. doi: 10.1037/ a0022620 148 Forgatch, M. S., Bullock, B. M., & Patterson, G. R. (2004). From theory to practice: Increasing effective parenting through role-play. In H. Steiner (Ed.), Handbook of mental health interventions in children and adolescents: An integrated developmental approach. (pp. 782813). San Francisco: Jossey-Bass. Forgatch, M. S., & DeGarmo, D. S. (1999). Parenting through change: An effective prevention program for single mothers. Journal of Consulting and Clinical Psychology, 67, 711-724. doi: 10.1037/0022-006X.67.5.711 Forgatch, M. S., & DeGarmo, D. S. (2011). Sustaining fidelity following the nationwide PMTO™ implementation in Norway. Prevention Science, 12, 235-246. doi: 10.1007/s11121-011-0225-6 Forgatch, M. S., DeGarmo, D. S., & Beldavs, Z. G. (2005a). An efficacious theory-based intervention for stepfamilies. Behavior Therapy, 36, 357365. doi: 10.1016/S0005-7894(05)80117-0 Forgatch, M. S., Knutson, N. M., & Mayne, T. (1992). Coder impressions of ODS lab tasks. Eugene: Oregon Social Learning Center. Forgatch, M. S., & Patterson, G. R. (2010). Parent Management Training - Oregon model: An intervention for antisocial behavior in children and adolescents. In J. R. Weisz & A. E. Kazdin (Eds.), Evidence-based psychotherpies for children and adolescents (pp. 159-178). New York, NY: Guilford. Forgatch, M. S., Patterson, G. R., & DeGarmo, D. S. (2005b). Evaluating fidelity: Predictive validity for a measure of competent adherence to the Oregon Model of Parent Management Training. Behavior Therapy, 36, 3-13. doi: 10.1016/S00057894(05)80049-8 Forgatch, M. S., Patterson, G. R., DeGarmo, D. S., & Beldavs, Z. G. (2009). Testing the Orgeon delinquency model with 9-year follow-up of the Oregon Divorce study. Development and Psychopathology, 21, 637-660. doi: 10.1017/ S0954579409000340 Frick, P. J. (2006). Developmental pathways to conduct disorder. Child and Adolescent Psychiatric Clinics of North America, 15, 311-331. doi: 10.1016/j.chc.2005.11.003 Frick, P. J., Bodin, S. D., & Barry, C. T. (2000). Psychopathic traits and conduct problems in community and clinic-referred samples of children: Further development of the Psychopathy Screening Device. Psychological Assessment, 12, 382-393. doi: 10.1037/1040-3590.12.4.382 References | Frick, P. J., Cornell, A. H., Barry, C. T., Bodin, S. D., & Dane, H. E. (2003a). Callous-unemotional traits and conduct problems in the prediction of conduct problem severity, aggression, and self-report of delinquency. Journal of Abnormal Child Psychology, 31, 457-470. doi: 10.1023/A:1023899703866 Greene, R. W., Ablon, J. S., Goring, J. C., Fazio, V., & Morse, L. R. (2004). Treatment of oppositional defiant disorder in children and adolescents. In P. M. Barrett & T. H. Ollendick (Eds.), Handbook of interventions that work with children and adolescents: Prevention and treatment. (pp. 369393). New York, NY, US: John Wiley & Sons Ltd. Frick, P. J., & Hare, R. D. (2001). Antisocial Process Screening Device. Toronto: Multi-Health Systems. Hair, J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2009). Multivariate data analysis. A global perspective. New Jersey: Pearson Education Inc. Frick, P. J., Kimonis, E. R., Dandreaux, D. M., & Farell, J. M. (2003b). The 4 year stability of psychopathic traits in non-referred youth. Behavioral Sciences & the Law, 21, 713-736. doi: 10.1002/bsl.568 Frick, P. J., Lahey, B. B., Loeber, R., Stouthamer-Loeber, M., Christ, M. A. G., & Hanson, K. (1992). Familial risk factors to oppositional defiant disorder and conduct disorder: Parental psychopathology and maternal parenting. Journal of Consulting and Clinical Psychology, 60, 49-55. doi: 10.1037/0022006X.60.1.49 Frick, P. J., Ray, J. V., Thornton, L. C., & Kahn, R. E. (2014). Can callous-unemotional traits enhance the understanding, diagnosis, and treatment of serious conduct problems in children and adolescents? A comprehensive review. Psychological Bulletin, 140, 1-57. doi: 10.1037/ a0033076 Frick, P. J., Stickle, T. R., Dandreaux, D. M., Farrell, J. M., & Kimonis, E. R. (2005). Callous-unemotional traits in predicting the severity and stability of conduct problems and delinquency. Journal of Abnormal Child Psychology, 33, 471-487. doi: 10.1007/s10648-005-5728-9 Frick, P. J., & White, S. F. (2008). Research review: The importance of callous-unemotional traits for developmental models of aggressive and antisocial behavior. Journal of Child Psychology and Psychiatry, 49, 359-375. doi: 10.1111/j.14697610.2007.01862.x Gallo, D. A. (2010). False memories and fantastic beliefs: 15 years of the DRM illusion. Memory & Cognition, 38, 833-848. doi: 10.3758/MC.38.7.833 Glass, S. J., & Newman, J. P. (2009). Emotion processing in the criminal psychopath: The role of attention in emotion-facilitated memory. Journal of Abnormal Psychology, 118, 229-234. doi: 10.1037/a0014866 Hamilton, W., & Burns, T. G. (2003). Review of WPPSIIII: Wechsler Preschool and Primary Scale of Intelligence (3rd Ed.). Applied Neuropsychology, 10, 188-190. Hare, R. D. (1991). Manual for the Hare Psychopathy Checklist-revised. Toronto: Multi-Health Systems. Harvey, E., Danforth, J. S., Ulaszek, W. R., & Eberhardt, T. L. (2001). Validity of the parenting scale for parents of children with attention-deficit/ hyperactivity disorder. Behaviour Research and Therapy, 39, 731-743. doi: 10.1016/S00057967(00)00052-8 Hawes, D. J., & Dadds, M. R. (2005). The treatment of conduct problems in children with callousunemotional traits. Journal of Consulting and Clinical Psychology, 73, 737-741. doi: 10.1037/0022-006x.73.4.737 Hawes, D. J., & Dadds, M. R. (2006). Assessing parenting practices through parent-report and direct observation during parent-training. Journal of Child and Family Studies, 15, 555-568. doi: 10.1007/s10826-006-9029-x Hawes, D. J., Dadds, M. R., Brennan, J., Rhodes, T., & Cauchi, A. (2013). Revisiting the treatment of conduct problems in children with callousunemotional traits. Australian and New Zealand Journal of Psychiatry, 47, 646-653. doi: 10.1177/0004867413484092 Hawes, D. J., Price, M. J., & Dadds, M. R. (2014). Callousunemotional traits and the treatment of conduct problems in childhood and adolescence: A comprehensive review. Clinical Child and Family Psychology Review, 17, 248-267. doi: 10.1007/ s10567-014-0167-1 Hetherington, E. M., Bridges, M., & Insabella, G. M. (1998). What matters? What does not? Five perspectives on the association between marital transitions and children’s adjustment. American Psychologist, 53, 167-184. doi: 10.1037/0003066X.53.2.167 149 R | References Heuer, F., & Reisberg, D. (1990). Vivid memories of emotional events: The accuracy of remembered minutiae. Memory & Cognition, 18, 496-506. doi: 10.3758/BF03198482 Hipwell, A. E., Pardini, D. A., Loeber, R., Sembower, M., Keenan, K., & Stouthamer-Loeber, M. (2007). Callous-unemotional behaviors in young girls: Shared and unique effects relative to conduct problems. Journal of Clinical Child and Adolescent Psychology, 36, 293-304. doi: 10.1080/15374410701444165 Hodges, K. (2002). Caregiver Wish List. Ypsilanti: Eastern Michigan University. Hodges, K. (2005). Caregiver Wish List Handbook. Ypsilanti: Eastern Michigan University. Hodges, K., de Ruiter, C., & Thijssen, J. (2009). Authorized Dutch translation of the Caregiver Wish List. Maastricht, The Netherlands: Department of Clinical Psychological Science, Maastricht University. Högström, J., Enebrink, P., & Ghaderi, A. (2013). The moderating role of child callous-unemotional traits in an Internet-based parent-management training program. Journal of Family Psychology, 27, 314-323. doi: 10.1037/a0031883 Hogue, A., Henderson, C. E., Dauber, S., Barajas, P. C., Fried, A., & Liddle, H. A. (2008). Treatment adherence, competence, and outcome in individual and family therapy for adolescent behavior problems. Journal of Consulting and Clinical Psychology, 76, 544-555. doi: 10.1037/0022-006x.76.4.544 Howe, M. L., Candel, I., Otgaar, H., Malone, C., & Wimmer, M. C. (2010). Valence and the development of immediate and long-term false memory illusions. Memory, 18, 58-75. doi: 10.1080/09658210903476514 Howe, M. L., Toth, S. L., & Cicchetti, D. (2011). Can maltreated children inhibit true and false memories for emotional information? Child Development, 82, 967-981. doi: 10.1111/j.14678624.2011.01585.x Howe, M. L., Wimmer, M. C., Gagnon, N., & Plumpton, S. (2009). An associative-activation theory of children’s and adults’ memory illusions. Journal of Memory and Language, 60, 229-251. doi: 10.1016/j.jml.2008.10.002 Hu, L., & Bentler, P. M. (1999). Cutoff criteria fo fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal, 6, 1-55. doi: 10.1080/10705519909540118 150 Huey, S. J., Jr., Henggeler, S. W., Brondino, M. J., & Pickrel, S. G. (2000). Mechanisms of change in multisystemic therapy: Reducing delinquent behavior through therapist adherence and improved family and peer functioning. Journal of Consulting and Clinical Psychology, 68, 451467. doi: 10.1037/0022-006x.68.3.451 Hukkelberg, S. S., & Ogden, T. (2013). Working alliance and treatment fidelity as predictors of externalizing problem behaviors in parent management training. Journal of Consulting and Clinical Psychology, 81, 1010-1020. doi: 10.1037/ a0033825 Hunt, R. R., & Einstein, G. O. (1981). Relational and item-specific information in memory. Journal of Verbal Learning & Verbal Behavior, 20, 497-514. Jacobson, N. S., Roberts, L. J., Berns, S. B., & McGlinchey, J. B. (1999). Methods for defining and determining the clinical significance of treatment effects: Description, application, and alternatives. Journal of Consulting and Clinical Psychology, 67, 300-307. doi: 10.1037/0022006x.67.3.300 Johnston, C., & Jassy, J. S. (2007). Attention-Deficit/ Hyperactivity Disorder and Oppositional/ Conduct problems: Links to parent-child interactions. Journal of the Canadian Academy of Child and Adolescent Psychiatry, 16, 74-79. Kaufman, A. S., Kaufman, J. C., Balgopal, R., & McLean, J. E. (1996). Comparison of three WISC-III short forms: Weighing psychometric, clinical, and practical factors. Journal of Clinical Child Psychology, 25, 97-105. doi: 10.1207/ s15374424jccp2501_11 Kazdin, A. E. (2007). Progress in treating children referred for severe aggressive and antisocial behavior. NYS Psychologist, 19, 7-12. Kimonis, E. R., Bagner, D. M., Linares, D., Blake, C. A., & Rodriguez, G. (2014). Parent training outcomes among young children with callous– unemotional conduct problems with or at risk for developmental delay. Journal of Child and Family Studies, 23, 437-448. doi: 10.1007/s10826013-9756-8 Kimonis, E. R., Frick, P. J., Fazekas, H., & Loney, B. R. (2006). Psychopathy, aggression, and the processing of emotional stimuli in non-referred girls and boys. Behavioral Sciences & the Law, 24, 21-37. doi: 10.1002/bsl.668 References | Kjøbli, J., Hukkelberg, S., & Ogden, T. (2013). A randomized trial of group parent training: Reducing child conduct problems in real-world settings. Behaviour Research and Therapy, 51, 113-121. doi: 10.1016/j.brat.2012.11.006 Ledingham, J. E., & Schwartzman, A. E. (1984). A 3-year follow-up of aggressive and withdrawn behavior in childhood: Preliminary findings. Journal of Abnormal Child Psychology, 12, 157168. doi: 10.1007/BF00913467 Kjøbli, J., & Ogden, T. (2012). A randomized effectiveness trial of brief parent training in primary care settings. Prevention Science, 13, 616-626. doi: 10.1007/s11121-012-0289-y Lee, I. A., & Preacher, K. J. (2013, September). Calculation for the test of the difference between two dependent correlations with one variable in common [Computer software]. Retrieved from http://quantpsy.org. Kleefman, M., Jansen, D. E., Stewart, R. E., & Reijneveld, S. A. (2014). The effectiveness of Stepping Stones Triple P parenting support in parents of children with borderline to mild intellectual disability and psychosocial problems: a randomized controlled trial. BMC Medicine, 12, 191-198. doi: 10.1186/ s12916-014-0191-5 Knutson, N. M., Forgatch, M. S., Rains, L. A., & Sigmarsdóttir, M. (2009). Fidelity of implementation rating system (FIMP): The manual for PMTO. Eugene, Oregon: Implementation Sciences International, Inc. Kramer, T. H., Buckhout, R., & Eugenio, P. (1990). Weapon focus, arousal, and eyewitness memory: Attention must be paid. Law and Human Behavior, 14, 167-184. doi: 10.1007/bf01062971 LaBar, K. S., & Cabeza, R. (2006). Cognitive neuroscience of emotional memory. Nature Reviews Neuroscience, 7, 54-64. doi: 10.1038/ nrn1825 Laney, C., & Takarangi, M. K. T. (2013). False memories for aggressive acts. Acta Psychologica, 143, 227234. doi: 10.1016/j.actpsy.2013.04.001 Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1998). International affective picture system (IAPS): Technical manual and affective ratings. Centre for research in psychophysiology, University of Florida: Gainesville. Langendam, M. W., Hooft, L., Heus, P., Scholten, R. J. P. M., & Bossuyt, P. M. M. (2013). Alternatieven voor Randomized Controlled Trials in onderzoek naar de effectiviteit van interventies. Amsterdam: Dutch Cochrane Centre. Lau, A. S. (2006). Making the case for selective and directed cultural adaptations of evidence-based treatments: Examples from parent training. Clinical Psychology: Science and Practice, 13, 295310. doi: 10.1111/j.1468-2850.2006.00042.x Lawing, K., Frick, P. J., & Cruise, K. R. (2010). Differences in offending patterns between adolescent sex offenders high or low in callous-unemotional traits. Psychological Assessment, 22, 298-305. doi: 10.1037/a0018707 Levine, L. J., & Edelstein, R. S. (2009). Emotion and memory narrowing: A review and goal-relevance approach. Cognition and Emotion, 23, 833-875. doi: 10.1080/02699930902738863 Little, R. (1988). Missing-data adjustments in large surveys. Journal of Business and Economic Statistics, 6, 287-296. Loeber, R., & Burke, J. D. (2011). Developmental pathways in juvenile externalizing and internalizing problems. Journal of Research on Adolescence, 21, 34-46. doi: 10.1111/j.15327795.2010.00713.x Loftus, E. F. (2005). Planting misinformation in the human mind: A 30-year investigation of the malleability of memory. Learning & Memory, 12, 361-366. doi: 10.1101/lm.94705 Loney, B. R., Frick, P. J., Clements, C. B., Ellis, M. L., & Kerlin, K. (2003). Callous-unemotional traits, impulsivity, and emotional processing in adolescents with antisocial behavior problems. Journal of Clinical Child and Adolescent Psychology, 32, 66-80. doi: 10.1207/S15374424JCCP3201_07 Martinez, C. R., & Forgatch, M. S. (2001). Preventing problems with boys’ noncompliance: Effects of a parent training intervention for divorcing mothers. Journal of Consulting and Clinical Psychology, 69, 416-428. doi: 10.1037/0022006X.69.3.416 McCart, M. R., Priester, P. E., Davies, W. H., & Azen, R. (2006). Differential effectiveness of Behavioral Parent-Training and Cognitive-Behavioral Therapy for antisocial youth: A meta-analysis. Journal of Abnormal Child Psychology, 34, 527543. doi: 10.1007/s10802-006-9031-1 McCoy, M. G., Frick, P. J., Loney, B. R., & Ellis, M. L. (1999). The potential mediating role of parenting practices in the development of conduct problems in a clinic-referred sample. Journal of Child and Family Studies, 8, 477-494. doi: 10.1023/A:1021907905277 151 R | References McGraw, K. O., & Wong, S. P. (1996). Forming inferences about some intraclass correlation coefficients. Psychological Methods, 1, 30-46. doi: 10.1037/1082-989x.1.1.30 Nixon, R. D. V. (2002). Treatment of behavior problems in preschoolers: A review of parent training programs. Clinical Psychology Review, 22, 525-546. doi: 10.1016/S0272-7358(01)00119-2 McMahon, R. J., Witkiewitz, K., & Kotler, J. S. (2010). Predictive validity of callous-unemotional traits measured in early adolescence with respect to multiple antisocial outcomes. Journal of Abnormal Psychology, 119, 752-763. doi: 10.1037/ a0020796 Obradovic, J., Pardini, D. A., Long, J. D., & Loeber, R. (2007). Measuring interpersonal callousness in boys from childhood to adolescence: An examination of longitudinal invariance and temporal stability. Journal of Clinical Child and Adolescent Psychology, 36, 276-292. doi: 10.1080/15374410701441633 Meyer, G. J., Finn, S. E., Eyde, L. D., Kay, G. G., Moreland, K. L., Dies, R. R., . . . Reed, G. M. (2001). Psychological testing and psychological assesment: A review of evidence and issues. American Psychologist, 56, 128-165. doi: 10.1037/0003-066X.56.2.128 Minshew, N. J., Turner, C. A., & Goldstein, G. (2005). The application of short forms of the Wechsler Intelligence Scales in adults and children with high functioning autism. Journal of Autism and Developmental Disorders, 35, 45-52. doi: 10.1007/ s10803-004-1030-x Mirandola, C., Toffalini, E., Grassano, M., Cornoldi, C., & Melinder, A. (2014). Inferential false memories of events: Negative consequences protect from distortions when the events are free from further elaboration. Memory, 22, 451-461. doi: 10.1080/09658211.2013.795976 Moffitt, T. E. (1993). Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychological Review, 100, 674-701. doi: 10.1037/0033-295X.100.4.674 Morsbach, S. K., & Prinz, R. J. (2006). Understanding and improving the validity of self-report of parenting. Clinical Child and Family Psychology Review, 9, 1-21. doi: 10.1007/s10567-006-0001-5 Nagin, D., & Tremblay, R. E. (1999). Trajectories of boys’ physical aggression, opposition, and hyperactivity on the path to physically violent and nonviolent juvenile delinquency. Child Development, 70, 1181-1196. doi: 10.1111/14678624.00086 Nicholson, B. C., Fox, R. A., & Johnson, S. D. (2005). Parenting young children with challenging behaviour. Infant and Child Development, 14, 425-428. doi: 10.1002/icd.403 Nix, R. L., Pinderhughes, E. E., Dodge, K. A., Bates, J. E., Pettit, G. S., & McFadyen-Ketchum, S. A. (1999). The relation between mothers’ hostile attribution tendencies and children’s externalizing behavior problems: The mediating role of mothers’ harsh discipline practices. Child Development, 70, 896909. doi: 10.1111/1467-8624.00065 152 Ogden, T., & Amlund-Hagen, K. (2008). Treatment effectiveness of parent management training in Norway: A randomized controlled trial of children with conduct problems. Journal of Consulting and Clinical Psychology, 76, 607-621. doi: 10.1037/0022-006X.76.4.607 Oliver, P. H., Guerin, D. W., & Coffman, J. K. (2009). Big five parental personality traits, parenting behaviors, and adolescent behavior problems: A mediation model. Personality and Individual Differences, 47, 631-636. doi: 10.1016/j. paid.2009.05.026 Otani, H., Libkuman, T. M., Widner, R. L., Jr., & Graves, E. I. (2007). Memory for emotionally arousing stimuli: A comparison of younger and older adults. Journal of General Psychology, 134, 23-42. doi: 10.3200/GENP.134.1 Otgaar, H., Peters, M., & Howe, M. L. (2012). Dividing attention lowers children’s but increases adults’ false memories. Journal of Experimental Psychology: Learning, Memory, and Cognition, 38, 204-210. doi: 10.1037/a0025160 Oxford, M., Cavell, T. A., & Hughes, J. N. (2003). Callous/unemotional traits moderate the relation between ineffective parenting and child externalizing problems: A partial replication and extension. Journal of Clinical Child and Adolescent Psychology, 32, 577-585. doi: 10.1207/ S15374424JCCP3204_10 Patterson, G. R. (1986). Performance models for antisocial boys. American Psychologist, 41, 432444. doi: 10.1037/0003-066X.41.4.432 Patterson, G. R. (2005). The next generation of PMTO models. The Behavior Therapist, 28, 25-32. Patterson, G. R., Chamberlain, P., & Reid, J. B. (1982). A comparative evaluation of a parent-training program. Behavior Therapy, 13, 638-650. doi: 10.1016/S0005-7894(82)80021-X References | Patterson, G. R., DeBaryshe, B. D., & Ramsey, E. (1989). A developmental perspective on antisocial behavior. American Psychologist, 44, 329-335. doi: 10.1037/0003-066X.44.2.329 Patterson, G. R., DeGarmo, D., & Forgatch, M. S. (2004). Systematic changes in families following prevention trials. Journal of Abnormal Child Psychology, 32, 621-633. doi: 10.1023/B:JACP.0000047211.11826.54 Patterson, G. R., Dishion, T. J., & Yoerger, K. (2000). Adolescent growth in new forms of problem behavior: Macro- and micro-peer dynamics. Prevention Science, 1, 3-13. doi: 10.1023/A:1010019915400 Patterson, G. R., Forgatch, M. S., & DeGarmo, D. S. (2010). Cascading effects following intervention. Development and Psychopathology, 22, 949-970. doi: 10.1017/S0954579410000568 Patterson, G. R., Forgatch, M. S., Yoerger, K. L., & Stoolmiller, M. (1998). Variables that initiate and maintain an early-onset trajectory for juvenile offending. Development and Psychopathology, 10, 531-547. doi: 10.1017/S0954579498001734 Patterson, G. R., Reid, J. B., & Eddy, J. M. (2002). A brief history of the Oregon model. In J. B. Reid, G. R. Patterson & J. Snyder (Eds.), Antisocial behavior in children and adolescents: A developmental analysis and model for intervention. (pp. 3-20). Washington, DC US: American Psychological Association. Peterson, L., Tremblay, G., Ewigman, B., & Popkey, C. (2002). The parental daily diary: A sensitive measure of the process of change in a child maltreatment prevention program. Behavior Modification, 26, 627-647. doi: 10.1177/014544502236654 Reef, J., Diamantopoulou, S., van Meurs, I., Verhulst, F. C., & van der Ende, J. (2011). Developmental trajectories of child to adolescent externalizing behavior and adult DSM-IV disorder: Results of a 24-year longitudinal study. Social Psychiatry and Psychiatric Epidemiology, 46, 1233-1241. doi: 10.1007/s00127-010-0297-9 Riggs, L., McQuiggan, D. A., Farb, N., Anderson, A. K., & Ryan, J. D. (2011). The role of overt attention in emotion-modulated memory. Emotion, 11, 776785. doi: 10.1037/a0022591 Roediger, H. L., & McDermott, K. B. (1995). Creating false memories: Remembering words not presented in lists. Journal of Experimental Psychology: Learning, Memory, and Cognition, 21, 803-814. doi: 10.1037/0278-7393.21.4.803 Roediger, H. L., Watson, J. M., McDermott, K. B., & Gallo, D. A. (2001). Factors that determine false recall: A multiple regression analysis. Psychonomic Bulletin & Review, 8, 385-407. doi: 10.3758/BF03196177 Romeo, R., Knapp, M., & Scott, S. (2006). Economic cost of severe antisocial behaviour in children-And who pays it. British Journal of Psychiatry, 188, 547-553. doi: 10.1192/bjp.bp.104.007625 Rubin, D. B. (1986). Statistical matching using file concatenation with adjusted weights and multiple imputations. Journal of Business and Economic Statistics, 4, 87-94. Rubin, D. B. (1987). Multiple imputation for nonresponse in surveys. New York: John Wiley. Salekin, R. T., Debus, S. A., & Jackson, R. (2008). Assessing child and adolescent psychopathy Learning forensic assessment. (pp. 347-383). New York, NY US: Routledge/Taylor & Francis Group. Pickel, K. L., Narter, D. B., Jameson, M. M., & Lenhardt, T. T. (2008). The weapon focus effect in child eyewitnesses. Psychology, Crime & Law, 14, 61-72. doi: 10.1080/10683160701391307 Sanders, M. R. (2008). Triple P-Positive Parenting Program as a public health approach to strengthening parenting. Journal of Family Psychology, 22, 506-517. doi: 10.1037/08933200.22.3.506 Prior, M., Smart, D., Sanson, A., & Oberklaid, F. (1999). Relationships between learning difficulties and psychological problems in preadolescent children from a longitudinal sample. Journal of the American Academy of Child & Adolescent Psychiatry, 38, 429-436. doi: 10.1097/00004583199904000-00016 Scarboro, M. E., & Forehand, R. (1975). Effects of two types of response-contingent time-out on compliance and oppositional behavior of children. Journal of Experimental Child Psychology, 19, 252-264. doi: 10.1016/00220965(75)90089-2 R Core Team. (2014). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. Schwarz, N. (1999). Self-reports: How the questions shape the answers. American Psychologist, 54, 93-105. doi: 10.1037/0003-066x.54.2.93 153 R | References Scott, S., Knapp, M., Henderson, J., & Maughan, B. (2001). Financial cost of social exclusion: Follow up study of antisocial children into adulthood. British Medical Journal, 323, 1-5. doi: 10.1136/ bmj.323.7306.191 Shaffer, D., Fisher, P., Lucas, C. P., Dulcan, M. K., & Schwab-Stone, M. E. (2000). NIMH Diagnostic Interview Schedule for Children Version IV (NIMH DISC-IV): Description, differences from previous versions, and reliability of some common diagnoses. Journal of the American Academy of Child & Adolescent Psychiatry, 39, 28-38. doi: 10.1097/00004583-200001000-00014 Sharma, V., & Sandhu, G. K. (2006). A community study of association between parenting dimensions and externalizing behaviors. Journal of Indian Association for Child and Adolescent Mental Health, 2, 48-58. Shelleby, E. C., & Shaw, D. S. (2014). Outcomes of parenting interventions for child conduct problems: A review of differential effectiveness. Child Psychiatry and Human Development, 45, 628-645. doi: 10.1007/s10578-013-0431-5 Shelton, K. K., Frick, P. J., & Wootton, J. (1996). Assessment of parenting practices in families of elementary school-age children. Journal of Clinical Child Psychology, 25, 317-329. doi: 10.1207/s15374424jccp2503_8 Sibbald, B., & Roland, M. (1998). Understanding controlled trials. Why are randomised controlled trials important? British Medical Journal, 316, 201. doi: 10.1136/bmj.316.7126.201 Siegel, D. J., & Bryson, T. P. (2014, October 21). You said WHAT about time-outs? The Huffington Post, retrieved from http://www.huffingtonpost. com/daniel-j-siegel-md/time-outsoverused_b_6006332.html. Siegel, D. J., & Bryson, T. P. (2014, September 23). Time-outs are hurting your child. TIME. Sigmarsdóttir, M., DeGarmo, D. S., Forgatch, M. S., & Guðmundsdóttir, E. V. (2013). Treatment effectiveness of PMTO for children’s behavior problems in Iceland: Assessing parenting practices in a randomized controlled trial. Scandinavian Journal of Psychology, 54, 468-476. doi: 10.1111/sjop.12078 Sigmarsdóttir, M., Thorlacius, Ö., Guðmundsdóttir, E. V., & DeGarmo, D. S. (2014). Treatment effectiveness of PMTO for children’s behavior problems in iceland: Child outcomes in a nationwide randomized controlled trial. Family Process, 1-20. doi: 10.1111/famp.12109 154 Simonoff, E., Elander, J., Holmshaw, J., Pickles, A., Murray, R., & Rutter, M. (2004). Predictors of antisocial personality: Continuities from childhood to adult life. British Journal of Psychiatry, 184, 118-127. doi: 10.1192/ bjp.184.2.118 Snyder, J. (2002). Reinforcement and coercion mechanisms in the development of antisocial behavior: Peer relationships. In J. B. Reid, G. R. Patterson & J. Snyder (Eds.), Antisocial behavior in children and adolescents: A developmental analysis and model for intervention. (pp. 101-122). Washington, DC US: American Psychological Association. Snyder, J., Edwards, P., McGraw, K., Kilgore, K., & Holton, A. (1994). Escalation and reinforcement in mother-child conflict: Social processes associated with the development of physical aggression. Development and Psychopathology, 6, 305-321. doi: 10.1017/S0954579400004600 Snyder, J., & Patterson, G. R. (1995). Individual differences in social aggression: A test of a reinforcement model of socialization in the natural environment. Behavior Therapy, 26, 371391. doi: 10.1016/S0005-7894(05)80111-X Spijkers, W., Jansen, D. E., & Reijneveld, S. A. (2013). Effectiveness of Primary Care Triple P on child psychosocial problems in preventive child healthcare: a randomized controlled trial. BMC Medicine, 11, 240-247. doi: 10.1186/1741-701511-240 Steiger, J. H. (1980). Tests for comparing elements of a correlation matrix. Psychological Bulletin, 87, 245-251. doi: 10.1037/0033-2909.87.2.245 Steinmetz, K. R. M., Addis, D. R., & Kensinger, E. A. (2010). The effect of arousal on the emotional memory network depends on valence. NeuroImage, 53, 318-324. doi: 10.1016/j. neuroimage.2010.06.015 Stevens, D., Charman, T., & Blair, R. J. R. (2001). Recognition of emotion in facial expressions and vocal tones in children with psychopathic tendencies. The Journal of Genetic Psychology, 162, 201-211. doi: 10.1080/00221320109597961 Takarangi, M. K. T., Polaschek, D. L. L., Hignett, A., & Garry, M. (2008). Chronic and temporary aggression causes hostile false memories for ambiguous information. Applied Cognitive Psychology, 22, 39-49. doi: 10.1002/acp.1327 Talmi, D., & Moscovitch, M. (2004). Can semantic relatedness explain the enhancement of memory for emotional words? Memory & Cognition, 32, 742-751. doi: 10.3758/BF03195864 References | Thijssen, J., Albrecht, G., Muris, P., & de Ruiter, C. (2015a). Treatment fidelity as a determinant of the effectiveness of Parent Management Training - Oregon model in The Netherlands. Manuscript submitted for publication. Thijssen, J., Muris, P., & de Ruiter, C. (2015b). Initial validation of the Dutch translation of the Caregiver Wish List, an interview-based scale for measuring parenting practices. Manuscript submitted for publication. Thijssen, J., Otgaar, H., Howe, M. L., & de Ruiter, C. (2013). Emotional true and false memories in children with callous-unemotional traits. Cognition and Emotion, 27, 761-768. doi: 10.1080/02699931.2012.744300 Thijssen, J., Otgaar, H., Meijer, E. H., Smeets, T., & de Ruiter, C. (2012). Emotional memory for central and peripheral details in children with callousunemotional traits. Behavioral Sciences & the Law, 30, 506-515. doi: 10.1002/bsl.2021 Thijssen, J., Vink, G., Muris, P., & de Ruiter, C. (2015c). The effectiveness of Parent Management Training - Oregon model for children with externalizing behavior problems in The Netherlands. Manuscript accepted with revisions. Tracey, T. J., & Kokotovic, A. M. (1989). Factor structure of the Working Alliance Inventory. Psychological Assessment, 1, 207-210. doi: 10.1037/10403590.1.3.207 Tremblay, R. E. (2000). The development of aggressive behaviour during childhood: What have we learned in the past century? International Journal of Behavioral Development, 24, 129-141. doi: 10.1080/016502500383232 van Buuren, S., Boshuizen, H. C., & Knook, D. L. (1999). Multiple imputation of missing blood pressure covariates in survival analysis. Statistics in Medicine, 18, 681-694. doi: 10.1002/ (SICI)1097-0258(19990330)18:6<681::AIDSIM71>3.0.CO;2-R van Buuren, S., & Groothuis-Oudshoorn, K. (2011). Mice: Multivariate Imputation by Chained Equations in R. Journal of Statistical Software, 45, 1-67. van Leeuwen, E., & Albrecht, G. (2008). Parent Management Training Oregon. Ernstig probleemgedrag neemt af met PMTO. Amsterdam: PI Research. Vannucci, M., Nocentini, A., Chiorri, C., & Menesini, E. (2014). Brief report: Violent false memories and engagement in aggressive and delinquent behavior: An investigation in adolescents. Journal of Adolescence, 37, 1329-1333. doi: 10.1016/j.adolescence.2014.09.007 Vannucci, M., Nocentini, A., Mazzoni, G., & Menesini, E. (2012). Recalling unpresented hostile words: False memories predictors of traditional and cyberbullying. European Journal of Developmental Psychology, 9, 182-194. doi: 10.1080/17405629.2011.646459 Verhulst, F. C., van der Ende, J., & Koot, H. M. (1996). Handleiding voor de CBCL/4-18 [Maunal for the CBCL/4-18]. Rotterdam: Department of child and adolescent psychiatry, Erasmus Medical Centre/ Sophia. Viding, E., Fontaine, N. M. G., Oliver, B. R., & Plomin, R. (2009). Negative parental discipline, conduct problems and callous-unemotional traits: Monozygotic twin differences study. British Journal of Psychiatry, 195, 414-419. doi: 10.1192/bjp.bp.108.061192 Viera, A. J., & Bangdiwala, S. I. (2007). Eliminating bias in randomized controlled trials: Importance of allocation concealment and masking. Family Medicine, 39, 132-137. Vitacco, M. J., Neumann, G. S., Ramos, V., & Roberts, M. K. (2003a). Ineffective parenting: A precursor to psychopathic traits and delinquency in Hispanic females. Annals of the New York Academy of Sciences, 1008, 300-303. Vitacco, M. J., Rogers, R., & Neumann, C. S. (2003b). The Antisocial Process Screening Device: An examination of its construct and criterionrelated validity. Assessment, 10, 143-150. doi: 10.1177/1073191103010002005 Waller, R., Gardner, F., Hyde, L. W., Shaw, D. S., Dishion, T. J., & Wilson, M. N. (2012). Do harsh and positive parenting predict parent reports of deceitfulcallous behavior in early childhood? Journal of Child Psychology and Psychiatry, 53, 946-953. doi: 10.1111/j.1469-7610.2012.02550.x Walter, H. I., & Gilmore, S. K. (1973). Placebo versus social learning effects in parent training procedures designed to alter the behavior of aggressive boys. Behavior Therapy, 4, 361-377. doi: 10.1016/S0005-7894(73)80116-9 155 R | References Webster-Stratton, C. H., & Reid, M. J. (2010). The incredible years program for children from infancy to pre-adolescence: Prevention and treatment of behavior problems. In R. C. Murrihy, A. D. Kidman & T. H. Ollendick (Eds.), Clinical handbook of assessing and treating conduct problems in youth. (pp. 117-138). New York, NY, US: Springer Science + Business Media. Wechsler, D. (1991). Manual for the Wechsler Intelligence Scale for Children - Third edition. San Antonio: The Psychological Corporation. Wechsler, D. (2002). Wechsler Preschool and Primary Scale of Intelligence - Third edition. San Antonio: The Psychological Corporation. Williams, L. R., Degnan, K. A., Perez-Edgar, K. E., Henderson, H. A., Rubin, K. H., Pine, D. S., . . . Fox, N. A. (2009). Impact of behavioral inhibition and parenting style on internalizing and externalizing problems from early childhood through adolescence. Journal of Abnormal Child Psychology, 37, 1063-1075. doi: 10.1007/s10802009-9331-3 Wiltz, N. A., & Patterson, G. R. (1974). An evaluation of parent training procedures designed to alter inappropriate aggressive behavior of boys. Behavior Therapy, 5, 215-221. doi: 10.1016/ S0005-7894(74)80137-1 Wimmer, M. C., & Howe, M. L. (2009). The development of automatic associative processes and children’s false memories. Journal of Experimental Child Psychology, 104, 447-465. doi: 10.1016/j.jecp.2009.07.006 156 Woodworth, M., & Waschbusch, D. (2008). Emotional processing in children with conduct problems and callous/unemotional traits. Child: Care, Health and Development, 34, 234-244. doi: 10.1111/j.1365-2214.2007.00792.x Wootton, J. M., Frick, P. J., Shelton, K. K., & Silverthorn, P. (1997). Ineffective parenting and childhood conduct problems: The moderating role of callous-unemotional traits. Journal of Consulting and Clinical Psychology, 65, 301-308. doi: 10.1037/0022-006X.65.2.292.b Youngstrom, E., Loeber, R., & Stouthamer-Loeber, M. (2000). Patterns and correlates of agreement between parent, teacher, and male adolescent ratings of externalizing and internalizing problems. Journal of Consulting and Clinical Psychology, 68, 1038-1050. doi: 10.1037/0022006x.68.6.1038 Zaslow, M. J., Weinfield, N. S., Gallagher, M., Hair, E. C., Ogawa, J. R., Egeland, B., . . . De Temple, J. M. (2006). Longitudinal prediction of child outcomes from differing measures of parenting in a low-income sample. Developmental Psychology, 42, 27-37. doi: 10.1037/0012-1649.42.1.27 Zlomke, K. R., Lamport, D., Bauman, S., Garland, B., & Talbot, B. (2014). Parenting adolescents: Examining the factor structure of the Alabama Parenting Questionnaire for adolescents. Journal of Child and Family Studies, 23, 1484-1490. doi: 10.1007/s10826-013-9803-5 Dankwoord (acknowledgments) D Dankwoord | Ik kan bijna niet geloven dat dit proefschrift er dan eindelijk is. Toch wel gek om na al die jaren eindelijk op dit punt te zijn gekomen. Het was een lang en moeizaam project, maar wat ben ik trots op het eindresultaat! Dit was me absoluut niet in mijn eentje gelukt. Gelukkig heb ik hier de ruimte om iedereen te bedanken. Corine, de belangrijkste bijdrage komt natuurlijk van jou. Je was nog maar een week in dienst bij de UM, toen ik als derdejaarsstudent op jouw deur klopte om te vragen of je me wilde begeleiden bij mijn bachelorthese. Toen was je al bezig met het implementeren van PMTO in Nederland. Na mijn master gaf je me de mogelijkheid om de effectiviteit van PMTO in Nederland als promovenda te onderzoeken. Een fantastische kans die ik niet kon laten liggen. Dat het onderzoek zo moeizaam zou verlopen hadden we beiden niet verwacht. Meerdere keren heb ik op het punt gestaan om de handdoek in de ring te gooien. Steeds wist jij te voorkomen dat ik het opgaf. Je gaf me de gelegenheid om extra studies uit te voeren en je hebt me de ruimte gegeven om klinische ervaring op te doen. Uiteindelijk heb ik met hervonden enthousiasme en motivatie het project kunnen afmaken. Bedankt dat je in mij en in het project bent blijven geloven. Peter, jij werd in de loop van het promotietraject als tweede promotor bij mijn project betrokken en daar ben ik altijd erg blij mee geweest. Je was altijd razendsnel met je feedback en je keek met een zeer kritische blik naar mijn stukken. Hoewel ik dat ook wel eens heb vervloekt, moet ik toegeven dat de stukken daar wel altijd beter van zijn geworden. Vooral in de laatste fase van mijn proefschrift ben je van onschatbare waarde geweest. Bedankt! Ik ben veel dank verschuldigd aan de instellingen die hun medewerking hebben verleend aan het onderzoek: De Bascule, Cardea Jeugdzorg, GGZ kinderen en Jeugd Rivierduinen, Yorneo en Lindenhout. Jullie hebben veel geïnvesteerd om dit onderzoek mogelijk te maken. Procedures moesten soms worden aangepast en personeel moest beschikbaar worden gesteld. In tijden van bezuinigingen was het moeilijk om hier tijd en geld voor vrij te maken, maar het is uiteindelijk gelukt! Bedankt voor jullie doorzettingsvermogen en dank aan de bestuurders dat jullie naast de zorg voor cliënten, het belang en de noodzaak van continue onderzoek steeds meer inzien. Het uitzetten en vervolgens draaiende houden van onderzoek op meerdere locaties in het land is een behoorlijke uitdaging en bijna onmogelijk om alleen te doen vanuit Maastricht. Ik ben dan ook enorm blij dat we op iedere locatie een onderzoekscoördinator hadden. Inès, Carlijn, Romana, Judith en Erna: ontzettend bedankt voor jullie inzet voor het onderzoek! Het is voor jullie zeker niet gemakkelijk geweest om de belangen van 159 D | Dankwoord zowel Maastricht als jullie eigen instelling op één lijn te krijgen en te houden. Ik ben dan ook enorm dankbaar dat jullie de moed niet hebben opgegeven. Ook jullie voorgangers verdienen een dikke pluim. Ik zou ook graag de gezinnen willen bedanken die aan het PMTO-onderzoek hebben deelgenomen. Er is heel wat moed voor nodig om hulp te zoeken wanneer je kind gedragsproblemen heeft. Op zo’n moment staat het water vaak al tot aan de lippen. Om dan naast de behandeling ook nog deel te nemen aan een onderzoek, waarbij er ieder half jaar behoorlijk wat vragenlijsten ingevuld moesten worden, is bewonderingswaardig. Mijn dank aan alle ouders is groot. I would like to thank my colleagues from the Atferdssenteret in Oslo and from the OSLC in Oregon. Kristin Nordahl, thank you for teaching me the microsocial coding of the Structured Interaction Tasks. Julie Stubbs, thank you for your help with improving our Coder Impressions. Finally, thank you Marion Forgatch, for your interest, suggestions, and support for our research over the years. Gonnie, bedankt voor de jarenlange samenwerking vanuit PI Research/Kenniscentrum PMTO en voor je hulp bij de treatment fidelity studie. Ferko, jouw enorme enthousiasme tovert iedere keer weer een glimlach op mijn gezicht. Bedankt voor al je steun aan mij en Corine de afgelopen jaren. Jij hebt altijd in dit project geloofd en dat werkte aanstekelijk. Gerko, de multipele geïmputeerde dataset van jou heeft de statistische analyses er niet gemakkelijker op gemaakt voor mij. Gelukkig kon ik altijd bij je terecht met mijn vragen en nam je de tijd voor me, zelfs als je een papa-dag had. Zo kon ik snel weer verder. Bedankt hiervoor! Lieve Linsey en Nele. Al sinds het begin zijn we kamergenootjes en door de tijd echte vriendinnen geworden. We hebben samen lief en leed gedeeld met betrekking tot ons promotieonderzoek, maar ook op privégebied. Ontzettend bedankt voor jullie emotionele steun op beide fronten. Ik ben enorm trots dat jullie op deze belangrijke dag mijn paranimfen zijn. Peggy, hoewel je pas sinds een paar maanden ons nieuwe kamergenootje bent, ben je een leuke aanwinst. Aangezien je recent jouw proefschrift heb afgerond, kon ik bij jou terecht voor mijn vele vragen over het proefschrift en alle procedures. Bedankt dat je de tijd voor mij nam. 160 Dankwoord | Lieve (oud-)collega’s van de sectie Forensische Psychologie. Jullie hebben mijn AIO-tijd extra aangenaam gemaakt. Bedankt voor jullie wetenschappelijke inzichten, leerzame retraites en gezellige lunches. Ik ben dankbaar dat ik met zo’n deskundige, talentvolle en vooral leuke groep heb mogen werken. Saskia, ook al werk je al een tijdje niet meer aan de UM, het was supergezellig met jou als collega. Bedankt voor de vele, vooral nietwerk gerelateerde gesprekken. Ook mijn vriendinnen ben ik dank verschuldigd. Lieve Coby’s, hoewel jullie niet direct betrokken zijn geweest bij mijn promotietraject, hebben jullie wel voor de nodige gezelligheid en vertier gezorgd. Hierdoor was ik in staat om mijn werk los te laten en helemaal te ontspannen. Al vanaf het begin van de middelbare school is onze groep gevormd en ik vind het fantastisch om jullie te hebben zien ontwikkelen van puberende meiden naar feestende studenten tot uiteindelijk serieus werkende vrouwen en, in sommige gevallen, moeders. Iedereen heeft zijn eigen pad gekozen, maar het contact is gebleven. Ik kijk met veel plezier terug op de vele uitstapjes, vakanties, stedentripjes en jaarlijkse vriendinnenweekendjes. Deze verzetjes zijn echt nodig geweest, vooral op momenten dat ik het niet meer zag zitten met mijn project. Ik ben dankbaar voor onze leuke club meiden (of moet ik dames zeggen, nu we de 30 zijn gepasseerd J). Dat we dit nog lang mogen volhouden! Pap en mam, bedankt voor jullie steun de afgelopen jaren. Hoewel jullie je regelmatig afvroegen waarom ik nog doorging met het project, zijn jullie wel altijd achter mij blijven staan. Jullie hebben me geleerd om af te maken waar je aan begint. Ook mijn schoonouders wil ik bedanken voor hun interesse en betrokkenheid. Allerliefste Guyon, bedankt voor al je steun, hulp en geduld! Alle hoogte- en dieptepunten van het project heb ik met je gedeeld. Jij hielp me te relativeren wanneer ik er doorheen zat en je vierde enthousiast ieder hoogtepunt mee. Ik mag me gelukkig prijzen dat ik jou al zo vroeg in mijn leven ben tegengekomen. We zijn een goed team. Ik hou heel veel van je. Lieve Roan, wat heb jij mijn leven verrijkt! Iedere ochtend als ik jouw vrolijke koppie weer zie, kan mijn dag niet meer stuk. Ik ben supertrots op jou! Liefste Alix, jij hebt ervoor gezorgd dat ik de laatste maanden een eindsprint heb gemaakt. Mijn proefschrift moest en zou voor jouw komst af zijn, zodat ik in mijn verlof alle rust zou hebben om van jou te kunnen genieten. Ik ben blij dat het is gelukt. Samen met papa en jouw grote broer maak je mij de gelukkigste vrouw op aarde. 161 D Curriculum Vitae C Curriculum vitae | Jill Thijssen werd geboren op 10 januari 1985 te Deurne. In juli 2004 behaalde zij haar vwo-diploma aan het Bouwens van der Boijecollege te Panningen. In september van dat jaar begon zij aan haar studie Psychologie aan de Universiteit Maastricht en in augustus 2007 behaalt ze haar masterdiploma in Psychology & Law cum laude. Van september 2007 tot en met juni 2008 heeft zij als onderzoekmedewerker aan de Faculteit der Psychologie en Neurowetenschappen van de Universiteit Maastricht gewerkt aan het opzetten van de effectiviteitsstudie naar PMTO. Vanaf 1 juli 2008 werd zij aangesteld als promovenda op de effectiviteitsstudie naar PMTO. Jill Thijssen was born on January 10, 1985 in Deurne. In July 2004, she finished her secondary school at the Bouwens van der Boijecollege in Panningen. In September of that year, she started studying Psychology at Maastricht University and in August 2007 she received her master’s diploma in Psychology & Law, cum laude. From September 2007 until June 2008, she worked as a research-assistant at the Faculty of Psychology and Neuroscience of Maastricht University. From July 2008, she was appointed at the same Faculty as a PhD student on the effectiveness study of PMTO. C 165 | Curriculum vitae International journal articles Duijster, D., de Jong-Lenters, M., de Ruiter, C., Thijssen, J., van Loveren, C., & Verrips, G. H. W. (2015). Parental and family-related influences on dental caries in children of Dutch, Moroccan and Turkish origin. Community Dentistry and Oral Epidemiology, 43, 152-162. doi: 10.1111/cdoe.12134 de Jong-Lenters, M., Duijster, D., Bruist, M. A., Thijssen, J., & de Ruiter, C. (2014). The relationship between parenting, family interaction and childhood dental caries: A case-control study. Social Science & Medicine, 116, 49-55. doi: 10.1016/j. socscimed.2014.06.031 Thijssen, J., Otgaar, H., Howe, M.L., & de Ruiter, C. (2013). Emotional true and false memories in children with callous-unemotional traits. Cognition & Emotion, 27, 761768. doi: 10.1080/02699931.2012.744300. Thijssen, J., Otgaar, H., Meijer, E.H., Smeets, T., & de Ruiter, C. (2012). Emotional memory for central and peripheral details in children with callous-unemotional traits. Behavioral Sciences and the Law, 30, 506-515. doi: 10.1002/bsl.2021 Thijssen, J., & de Ruiter, C. (2011). Identifying subtypes of spousal assaulters using the B-SAFER. Journal of Interpersonal Violence, 26, 1307-1321. doi: 10.1177/0886260510369129 Thijssen, J., & de Ruiter, C. (2011). Instrumental and expressive violence in Belgian homicide perpetrators. Journal of Investigative Psychology and Offender Profiling, 8, 58-73. doi: 10.1002/jip.130 Dutch journal articles Thijssen, J., & de Ruiter, C. (2010). De relatie tussen subtypen relationeel geweldplegers en de adviezen van de reclassering. PROCES, 89, 416-428. Thijssen, J., & de Ruiter, C. (2010). Strenge bewaking van kwaliteit bij PMTO. Kennis, 4, 8-14. Book chapter Thijssen, J., de Ruiter, C., & Albrecht, G. (2008). Preventie van antisociaal gedrag bij kinderen: Parent Management Training Oregon. [Prevention of antisocial behavior in children: Parent Management Training Oregon]. In J.R.M. Gerris en R.C.M.E. Engels (Eds.), Vernieuwingen in jeugd en gezin: Beleidsvisies, gezinsrelaties en interventies (p. 125 - 140). Assen: Van Gorcum. 166 Curriculum vitae | Submitted articles Thijssen, J., Vink, G., Muris, P., & de Ruiter, C. (2016). The effectiveness of Parent Management Training - Oregon Model in clinically referred children with externalizing behavior problems in The Netherlands. Accepted with revisions Thijssen, J., Albrecht, G., Muris, P., & de Ruiter, C. (2016). Treatment fidelity as a determinant of the effectiveness of Parent Management Training-Oregon model in The Netherlands. Submitted for publication Thijssen, J., Broers, N., Muris, P., & de Ruiter, C. (2016). Initial validation of the Dutch translation of the Caregiver Wish List, an interview-based scale for measuring parenting practices. Submitted for publication Conference presentations Thijssen, J., Vink, G., Muris, P., & de Ruiter, C. (2015). The effectiveness of PMTO in The Netherlands. Paper presented at the Association for Psychological Science, New York, United States. Thijssen, J., & de Ruiter, C. (2013). PMTO. Een behandeling voor kinderen met externaliserende gedragsproblemen. Paper presented at the ‘Forensic Psychology Update 2.0’, Maastricht, The Netherlands. Thijssen, J., de Ruiter, C., & Muris, P. (2013). Associations between parenting stress and self-rated and observed parenting behavior in parents of children with externalizing behavior problems. Poster presented at the Association for Psychological Science, Washington DC, United States. Thijssen, J., Otgaar, H., Meijer, E.H., Smeets, T., Howe, M.L., & de Ruiter, C. (2012). Emotional memory in children with callous-unemotional traits. Paper presented at the European Conference on Psychology and Law, Nicosia, Cyprus. Thijssen, J., Otgaar, H., Meijer, E.H., Smeets, T., & de Ruiter, C. (2011). Emotioneel geheugen in kinderen met psychopathische trekken. Paper presented at ‘Forensic Psychology Update’, Maastricht, The Netherlands. Thijssen, J., & de Ruiter, C. (2010). The predictive validity of the B-SAFER for spousal assault. Paper presented at the American Psychology-Law Society, Vancouver, Canada. Van Leeuwen, E., & Thijssen, J. (2009). Preventie van antisociaal gedrag bij kinderen: Parent Management Training Oregon (PMTO). [Prevention of antisocial behavior of children: Parent Management Training Oregon (PMTO)]. Poster presented at ‘Jeugd in Onderzoek’, Nieuwegein, The Netherlands. Thijssen, J., & de Ruiter, C. (2008). Risk assessment of spousal assaulters using the B-SAFER. Paper presented at the European Conference on Psychology and Law, Maastricht, The Netherlands. 167 C In a poor and infertile environment it is difficult to develop, but within the shadow of the parent’s love and attention, a child is able to bloom and grow to its full potential.