STUDYING CHANGE IN YOUNG PEOPLE WITH A MILD INTELLECTUAL DISABILITY: Time to take it personal! Daan Hulsmans
Studying change in young people with a mild intellectual disability: Time to take it personal! Daan Hulsmans
ISBN: 978-94-6506-904-3 Cover design and lay-out: Joey Roberts | www.ridderprint.nl Print: Ridderprint | www.ridderprint.nl © Copyright 2025: Daan Hulsmans, Nijmegen, The Netherlands All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, by photocopying, recording, or otherwise, without the prior written permission of the author.
Studying change in young people with a mild intellectual disability: Time to take it personal! Proefschrift ter verkrijging van de graad van doctor aan de Radboud Universiteit Nijmegen op gezag van de rector magnificus prof. dr. J.M. Sanders, volgens besluit van het college voor promoties in het openbaar te verdedigen op vrijdag 28 maart 2025 om 10.30 uur precies door Daniël Henricus Gerardus Hulsmans geboren op 23 september 1991 te Venlo
Promotoren: Prof. dr. R. Otten Prof. dr. A. Lichtwarck-Aschoff (Rijksuniversiteit Groningen) Prof. dr. E.A.P. Poelen Manuscriptcommissie: Prof. dr. A.M.T. Bosman Prof. dr. H.C.M. Didden Prof. dr. C. Schuengel (Vrije Universiteit Amsterdam) Dr. J.L. Pouwels Dr. H. Riese (Universitair Medisch Centrum Groningen)
Studying change in young people with a mild intellectual disability: Time to take it personal! Dissertation to obtain the degree of doctor from Radboud University Nijmegen on the authority of the Rector Magnificus prof. dr. J.M. Sanders, according to the decision of the Doctorate Board to be defended in public on Friday, March 28, 2025 at 10.30 am by Daniël Henricus Gerardus Hulsmans born on September 23, 1991 in Venlo (the Netherlands)
Supervisors: Prof. dr. R. Otten Prof. dr. A. Lichtwarck-Aschoff (University of Groningen) Prof. dr. E.A.P. Poelen Manuscript Committee: Prof. dr. A.M.T. Bosman Prof. dr. H.C.M. Didden Prof. dr. C. Schuengel (Vrije Universiteit Amsterdam) Dr. J.L. Pouwels Dr. H. Riese (University Medical Center Groningen)
Table of contents Chapter 1. General Introduction 9 Young people with a mild intellectual disability 10 Studying change 21 My own change towards this PhD 30 Chapter 2. The effectiveness of Take it Personal! on substance use 35 Chapter 3. Behavioral problems in Take it Personal! 53 Chapter 4. The feasibility of daily monitoring 73 Chapter 5. Idiographic personality networks 97 Chapter 6. Case study challenging behavior 125 Chapter 7. General Discussion 153 Summary of main findings 154 Discussion of findings 155 Epilogue: who have I studied? 172 Appendices 183 Appendix A: Supplement material for Chapter 2 184 Appendix B: Supplement material for Chapter 3 185 Appendix C: Supplement material for Chapter 5 186 Appendix D: Supplement material for Chapter 6 189 Appendix E: Supplement material for Chapter 6 190 References 191 Dutch summary (Nederlandse samenvatting) 208 List of publications 212 Research data management 214 About the author 216 Dankwoord (acknowledgments) 218
Chapter 1 General Introduction
10 Chapter 1 “The only constant in life is change” – Heraclitus. The concept of change has fascinated people since time immemorial. Ancient Greek philosopher Heraclitus argued that change is fundamental to the nature of reality, famously stating that one can never step into the same river twice. At the time, Heraclitus’ views challenged the works of Parmenides, who argued observable change to be deceptive because he considered the true nature of existence as static. This ancient debate between static and dynamic perspectives still resonates in modern social sciences. The degree to which one perceives reality as static or dynamic influences research methodologies, the interpretation of data, and the development of theories about human behavior. This dissertation more closely aligns with Heraclitus' dynamic perspective. No two people are fully identical, no two days are fully identical, and therefore no person is fully identical to who they were yesterday. The question remains: how should change be studied empirically? This dissertation describes different ways to study how and why problem behaviors change over time, focusing solely on young people who are diagnosed with a mild intellectual disability and receive specialized care. This is a specific group of people and a specific care setting for whom the study of change is especially relevant. I will therefore start this general introduction by describing the people that this dissertation is about and the care setting in which the research is situated. After that, this general introduction will proceed to the overarching topic of this dissertation: different ways of studying change. Young people with a mild intellectual disability With an example, I would like to describe a day in the life of Kevin, a fictitious (yet realistic) youngster living in a residential care facility specialized for mild intellectual disability. This example aims to illustrate three key features of this dissertation: 1) the care setting that this dissertation's research concerns, 2) the clinical relevance and difficulty of understanding problem behavioral change, and 3) that conclusions about change can differ substantially when drawn at the group- versus the individual level.
11 General Introduction 1 On a Wednesday afternoon in the living room of a residential care facility, 16-year-old Kevin and his care professional Eric get into an argument. Kevin was found smoking weed in his room. Because it is not the first time Kevin has broken this house rule, Eric decides to revoke Kevin’s weekend furlough. Kevin argues: “That’s totally unfair! It is Wednesday afternoon, and school is done early. Normally it is okay to smoke one on Wednesdays!". Eric counters that rules are rules, regardless of which day it is. Infuriated, Kevin asserts "Yeah, rules are rules whenever it suits you! Your rules only apply to me, not to others. Smoking weed helps people to relax after school. As long as it’s only one joint, all is well. You’re just doing this because it’s me!”. Later that week, Eric joins the weekly staff meeting, where all caretakers and clinicians gather to evaluate the group's overall progress. Despite the recent incident involving Kevin and Eric, the general consensus was positive. There have been a few incidents, the atmosphere within the group has been reported to be good, and the peace has noticeably increased compared to previous weeks. Several youngsters in the group have shown improvements, which is considered a sign that the daily routine schedule adjustments are working. However, there is a particular concern regarding Kevin. Eric brings up the recent incident involving Kevin's violation of the house rule on smoking weed, expressing his confusion and disappointment, noting that Kevin had
12 Chapter 1 been making progress in the past weeks. Eric feels at a loss for what to do with Kevin, wishing he understood Kevin’s motivations for such decisions better. The incident raises questions about how to support Kevin effectively and ensure progress in the future. In the above-mentioned (fictitious, yet realistic) scenario, three things stand out that are relevant to this dissertation. First, (problem) behaviors are not static entities, but patterns that can change over time. During the weekly meeting, the caretakers reflected on how clients were doing before, how they are doing now, and what might happen in the future. Second, the example of Kevin shows that preventing undesirable behavioral change is a difficult challenge for both clients and staff. A comprehensive understanding of why problems occur is likely necessary to help Kevin achieve desirable behavioral change. Third, one can draw conclusions about behavioral problems and preventive strategies at either the group level or for specific individuals. Those conclusions may differ substantially. As the staff meeting showed: there were concerns about Kevin's progress (the individual), while the general opinion about the group was positive. The chapters in this dissertation are dedicated to studying how and why problematic behaviors of young people with a mild intellectual disability change over time, both at the group level and the individual level. Before I further elaborate on how and why behaviors may possibly change over time, it is necessary to further introduce who this dissertation is about. I will first do so at the individual-level, by introducing the three other (again fictitious but realistic) vignettes of young people who, like Kevin, had a rough start in life or otherwise endured hardships early on. All of them receive care because they show behaviors that are deemed problematic, in a manner that is considerate of their (dis)abilities. After this person-specific introduction, I will provide a more general group-level description of people with a mild intellectual disability and the challenges they may face.
13 General Introduction 1 The individuals John is an 18-year-old who has not had an easy childhood. After his father left the family, he grew from a shy boy to a temperamental, unruly teenager. He had some run-ins with the police after getting caught stealing from the store. At home, the arguments with his mother had become increasingly intense and shortly after his 14th birthday, the situation escalated - at which point he moved to a residential youth care facility. Once there, his situation did not improve, as he felt the caretakers there worked against him, rather than with him. When he did not understand the instructions or routine provided by his caretakers, John took matters into his own hands by adopting a hot-headed, angry, and occasionally aggressive demeanor. John finished his vocational training but for a long time struggled to find a job. More and more, he felt he could not live up to his environment’s demands – a feeling that he could escape when being with friends, with whom he sometimes used cannabis. He also recently moved to a different apartment and found a job. At the moment, things are looking better, so this time around he is determined to succeed.
14 Chapter 1 Naomi is a 21-year-old, friendly and somewhat timid woman. She had difficulty making friends at school but had a loving family environment. At 15 years old, she became the victim of human trafficking by a lover boy. Two years of repeated sexual abuse followed. Ashamed, Naomi pretended to be all right for her friends and family. In reality, however, she struggled with severe depression and eventually became suicidal. After a failed suicide attempt, she moved to a residential youth care facility close to where she grew up. In the care facility, Naomi increasingly self-harms. Although she does well when she receives affectionate one-onone attention from caretakers, such contact cannot be provided 24/7. Most of the time, she self-harms when she is alone. Naomi goes through several therapies, without real improvement. Recently, she moved again to a new residential care facility on the other side of the country. Further away from the places where she experienced trauma makes her feel more at ease. Recently, she picked up new hobbies, such as knitting and taking care of horses. At times, she still self-harms, but less so than before.
15 General Introduction 1 Kyla is a 17-year-old girl. She was brought up in a household with two older siblings and an alcoholic father who kept a neglectful parenting style towards Kyla. Her mother was mostly absent. For as long as she can remember, she has been labeled a troublemaker. She attended a school for special education where she frequently fought with other kids and teachers. Kyla is headstrong and has a disregard for authority. This often leads to conflict. In a way, she is her own enemy. Eventually, youth care placed her with a foster family. After an escalation in her foster family, she lived with another foster family, after which she moved to residential youth care. Despite the caretakers’ best efforts, Kyla does not perceive the support she receives to be genuine. For a time she doubled down on her rule-breaking habits because "everything is broken already". She has a boyfriend, with whom she started to try out various illicit drugs. Recently she has been increasingly experiencing paranoia when being intoxicated. Her drug-using habits now frighten her somewhat. After having seen the devastating consequences of suffering from addiction, she wants to prevent developing one at all costs. The group: definition, prevalence rates and “common” problems Kevin, John, Naomi, and Kyla are a few examples of the thousands of youngsters who receive support and treatment for such problems in specialized (youth) care. Every person’s life story is different. As anyone who worked in youth- or mental healthcare will know: no two clients are the same. They do, however, have certain similarities that define them as a group. One thing that many individuals – like Kevin, John, Naomi, and Kyla – have in common: they have been diagnosed with a mild intellectual disability and receive mental healthcare specialized for this
16 Chapter 1 diagnosis. Now that I have introduced some of the individuals that this dissertation is about, it is time to describe who they are as a group. In this section, I will outline the definition of a mild intellectual disability, the prevalence rates of this diagnosis in the Netherlands, and "common" behavioral problems in this group. Definition A mild intellectual disability is – unlike severe and profound intellectual disability – not recognizable just by a person’s looks. According to the DSM-5, there are three criteria for diagnosing a mild intellectual disability (American Psychiatric Association [APA], 2013). The first criterion is a deficit in intellectual functioning, such as reasoning, problem-solving, abstract thinking, and judgment – as evidenced by intelligence tests and clinical judgment. In the Netherlands (and worldwide) there are many healthcare practices that require this criterion to be met through standardized tests revealing an intelligence quotient (IQ) between 50 and 69 (Landelijk Kenniscentrum LVB, 2024). However, the DSM-5 (APA, p. 33) as well as the ICD-11 (World Health Organization, 2022) discourage a strong emphasis on full-scale IQ scores. Instead, it is recommended that the second criterion – deficits in adaptive functioning – is decisive in the diagnostic process. Adaptive functioning deficits make a person fail to meet standards of social responsibility in one or more aspects of daily life, including communication, social participation, academic or occupational achievement, and personal independence at home or in community settings. Third, the aforementioned intellectual and adaptive deficits already manifested before the age of 18 (not as a result of acquired brain injury; APA, 2013). A mild intellectual disability often goes unrecognized during childhood or adolescence. As such, when diagnosing an adult, problems that happened during their youth can in hindsight be attributed to a mild intellectual disability. In the Netherlands, people with a so-called borderline intellectual functioning diagnosis have access to the same specialized care as their peers who, solely based on IQ might otherwise be considered as having a mild intellectual disability. The only difference between these diagnoses is that those with borderline intellectual functioning have a slightly higher IQ – typically ranging between 70 and 85. Similar to those with mild intellectual disability (i.e., IQ below the 70 cut-off) they struggle with the adaptive skills to meet the demands of everyday life, and thus need care that is considerate of their limited adaptive and intellectual abilities (APA, 2013; Emerson, 2011; Wieland & Zitman, 2016). Due to them both receiving the same specialized care, researchers typically study people with mild intellectual disability and borderline intellectual functioning as one group. This dissertation will do the same. To avoid an excess of the lengthy term ‘mild intellectual disability or
17 General Introduction 1 borderline intellectual functioning’, the target group is referred to as ‘mild intellectual disability’ throughout this dissertation. This is done to improve readability, but please bear in mind that, formally speaking, those with borderline intellectual functioning are included. Prevalence rates People with a mild intellectual disability constitute a significant portion of the Dutch mental healthcare system. Under the current definition of a mild intellectual disability, a roughly estimated 1.1 million Dutch people have a mild intellectual disability. Of course, not all are diagnosed as such, nor do this many people receive care specialized youth or adult mental healthcare for it. In the Netherlands, youth care is possible until the maximum age of 23. Adult mental healthcare facilities, some of which are also specialized for mild intellectual disability, are available for ages 18 and above. Approximately 110,000 people have a formal diagnosis and receive a form of (specialized) ambulatory or residential mental healthcare for people with disabilities ("Wet Langdurige Zorg"; CBS, 2020). Approximately 2,000 Dutch youth with a mild intellectual disability receive intensive residential care for their severe behavioral problems in the so-called “Orthopedagogische Behandelcentra” (Buysse et al., 2021). Also in regular youth care facilities, where a total of approximately 479,000 youth receive care, those with a mild intellectual disability diagnosis are overrepresented (CBS, 2024). That is, whereas 0.5% of youth who are not in youth care are diagnosed with a mild intellectual disability, this relative percentage is higher within the population that receives ambulatory youth care (1.4%). Especially in residential youth care, which includes open, semi-secure, and secure facilities (Harder et al., 2006), the overrepresentation of youngsters with a mild intellectual disability is most apparent (7.6%). Similar to their overrepresentation in youth care, there are disproportionally many adults with a mild intellectual disability in psychiatric care. Recent screening by Nieuwenhuis et al. (2022) showed that a staggering 44% of people in regular Dutch psychiatric care meet the criteria for a mild intellectual disability, suggesting that roughly 64,000 people could profit from specialized care for mild intellectual disability. Taken together, these prevalence rates indicate that the target group takes up a large part of the mental healthcare system. “Common” problems Young people with a mild intellectual disability often face a range of challenges that can manifest as behavioral problems. It can be overwhelming to process information, organize thoughts, and complete tasks, leading to a sense of failure that may manifest in disruptive behaviors. Many of them struggle with regulating
18 Chapter 1 their emotions and inhibiting impulses (Bexkens et al., 2014; Littlewood et al., 2018) – potentially leading to impulsive actions that are seen as externalizing problems, such as aggression towards peers or authority figures. Alternatively, emotion dysregulation can lead to internalizing problems like depressive feelings, self-injurious behavior, or suicidality. Social interactions can pose another challenge. For some, it may for example be difficult to express themselves effectively, which can lead to misunderstandings, frustration, and aggressive outbursts. Consequently, it can be hard to establish and/or maintain meaningful relationships. Some people with a mild intellectual disability fail to navigate subtle social cues, making them extra susceptible to peer influence (Larkin et al., 2013; Wagemaker et al., 2022). This contributes to an increased risk for minor or severe criminal offenses (Segeren et al., 2018). For some it is even hard to recognize general social norms, possibly leading to sexually transgressive behaviors. People with a mild intellectual disability are thus a highly heterogeneous group in terms of the problems they show. It frequently happens that a mild intellectual disability remains undiagnosed or undetected, and consequently, care is not adapted to the diagnosis (Nieuwenhuis et al., 2022; Wieland et al., 2020). Behavioral problems that may be inherent to an unrecognized mild intellectual disability are often met with little empathy from the environment. Constant comparisons to peers without intellectual disabilities – for example at home or at school – may lead to a negative self-perception, fostering a cycle of low self-esteem and potentially further challenging behaviors. This lack of understanding from others may cause feelings of inadequacy in these youngsters, negatively affecting their selfimage. For some, this leads to feelings of isolation or actual socially withdrawn behavior. Parental and societal attitudes potentially can exacerbate these issues. A considerable amount of young people with a mild intellectual disability find solace in using alcohol or drugs (VanDerNagel et al., 2014) – a behavior (a coping mechanism for some) that their environment may neither understand nor condone. Taken together, the nature of the challenges that someone with a mild intellectual disability faces, can be extremely wide-ranging. This range of challenges can also be recognized in the vignettes of John, Naomi, and Kyla. These case descriptions illustrate that although a mild intellectual disability can be a common factor, people with this diagnosis vary immensely in terms of backgrounds, their (dis)abilities, personalities, and behavioral problems. They differ in where they come from: their care-, educational- and familial histories (Segeren et al., 2018; Soenen et al., 2012; Nouwens et al., 2017). Some youngsters, like Kyla, have hopped from foster family to foster family, while others, like John and Naomi, received only residential care. Whereas some youngsters, like John, managed to obtain a
19 General Introduction 1 diploma, others have barely been in school. Some youngsters, like Naomi, have enjoyed a supportive familial background, while others, like Kyla, come from broken- or even abusive households. A commonality of people with a mild intellectual disability is impaired intellectual functioning, but how those (dis)abilities exactly manifest can vary widely between people. That is, research consistently shows that youngsters with the same overall IQ can have very different intelligence profiles: combinations of high/low (non)verbal working memory, knowledge, reasoning, and visual-spatial processing abilities (Bertelli et al., 2018; Márquez‑Caraveo et al., 2021; Sajewicz-Radtke et al., 2022; van der Molen et al., 2009). The consequences are, for example, that John may struggle to understand verbal instructions but excel in day-planning, while for Kyla this is the opposite, even though their average IQs may be the same. Heterogeneity in backgrounds and cognitive abilities contribute to a variety of adaptive problems youngsters with a mild intellectual disability can face. They typically show (several) behavioral problems, which vary in nature and severity (de Bildt et al., 2005; Embregts et al., 2010; Soenen et al., 2009; Nouwens et al., 2017). John and Kyla may, for example, find solace in the effects of using illicit substances. John’s substance use is mostly recreational but with Kyla, this may be becoming a more hard-wired pattern. Alternatively, Naomi is fully abstinent but struggles with issues like self-harm and suicidality. With such a wide variety of behavioral problems, it comes as no surprise that comorbid diagnoses to the mild intellectual disability diagnosis are the rule rather than the exception (Hesapcioglu et al., 2019; Whitaker & Read, 2006). Psychopathologies like anxiety-, attention deficit hyperactivity-, autism spectrum-, depressive-, oppositional defiant-, post-traumatic stress-, and substance use disorders are among the most common in addition to a mild intellectual disability. An important consequence of the immense diversity of this target group is that it implies highly individualized support needs. Every person and her/his context is unique, which is why staff are encouraged to provide person-specific care – tailored to the individual’s unique needs (Embregts et al., 2019).
20 Chapter 1 Figure 1 The AAIDD theoretical model of intellectual disability Note. This is an English translation that I made, based on the adapted version of the AAIDD-model (Embregts et al., 2019) Mapping out these needs is not an easy task. An integrated understanding of someone’s problematic behaviors requires taking both a broad and person-specific perspective. The American Association of Intellectual and Developmental Disorders (AAIDD; Schalock et al., 2010) developed a well-known model that can guide a broad inquiry of characteristics and support needs for people with a mild intellectual disability. Figure 1 shows a recently adapted version of the AAIDD-model by Embregts et al. (2019), who complemented this model with elements of the biopsychosocial model (Došen et al., 2008) and ecological systems model (Bronfenbrenner, 2005). Within this adapted AAIDD-model, the multifaceted nature of problematic behaviors and/or impeded quality of life can be traced back to five distinct but interrelated dimensions. Three dimensions reflect characteristics that are described as within-person factors (the onto system, Bronfenbrenner, 2005): her/his intellectual abilities, adaptive skills, and health (both mental and physical). In addition to such within-person factors (micro-sytem) and the effects the meso- and exosystem have on the person, there is always an interaction between the person and her/his environment. In the adapted AAIDD model, such features are subdivided into dimensions such as social participation (e.g., school or work) and care organizational context. Problematic behaviors (and/or impeded quality of life) indicate a disbalance
21 General Introduction 1 between five distinct but related dimensions (Embregts et al., 2019). Problem behavior may be sustained or even exacerbated when staff demand too much, too little, or the wrong things of the individual. In such cases, there exists a disbalance between the (dis)abilities of a person (left panel in Figure 1) and the provided support (oval in the middle). The AAIDD model presented in Figure 1 shows an immense amount of psychological, biological, and social factors contributing to problematic behaviors. The model is broad and comprehensive, which illustrates the complex task of understanding these behaviors. What complicates the matter even further is that – in addition to their multifaceted nature – behaviors change over time in person-specific ways. Studying change We have now arrived at the central topic of this dissertation: studying change. Stimulating youngsters to change their problematic behaviors for the better, or at least preventing change for the worse, is the essence of (youth) care practice. It constitutes the main goal of thousands of care professionals. They develop and implement prevention and intervention programs. They build relationships with their clients. They listen, educate, empower, and guide them – all with the aim of achieving improvements over time. All chapters in this dissertation are dedicated to answering the question of how and why problematic behaviors of people with a mild intellectual disability change over time. Below, I illustrate one fundamental contrast in approaches to studying change. Nomothetic and idiographic There are essentially two routes for studying change in people: the nomothetic approach and the idiographic approach. Nomothetic research looks broadly at whether a group of individuals change over time. The aim here is to identify lawlike change patterns within a sample, that then apply (or generalize) to a population. Idiographic research, on the other hand, is not about understanding how samples or populations change but attempts to understand the individual's change trajectory. The emphasis of an idiographic approach is on the uniqueness of individuals, rather than their sameness. That is, mapping out their own personspecific historical, psychological, and social context in relation to change. This dissertation will describe studies on change processes that use the nomothetic and
22 Chapter 1 idiographic approach. Both approaches have clear advantages and disadvantages – which I will introduce using a microscope metaphor1. Imagine someone looking through a microscope. Under the microscope’s lens is a large group of individuals. The person looking through the microscope can use different lenses. For example, choosing a lens that allows for a wide field of view on all individuals. By taking this broad perspective you can actually see what the sample looks like: which people look more or less alike and which people seem to differ. The conclusions drawn by looking through this lens ideally apply to many people. This broad perspective, however, also makes the view kind of blurry. The whole group is visible, but it is difficult to see much detail on any of the individuals within the group. To be able to see any individual in detail, the lens needs to be adjusted to one with a smaller field of view. Doing so allows a focus on a smaller subgroup or even one single person. This focus means that you have lost sight of all the others you might otherwise have been seeing using the wide-angle lens but the advantage of the narrow-angle lens, is that the view is much clearer and one can see detail on individuals. The conclusions you make about an individual may not apply to other individuals but do describe that individual better. Looking through the wide-angle lens is nomothetic science, while looking through the lens with a smaller field of view is idiographic science. Essentially, this dissertation is an illustration of how nomothetic and idiographic science can both contribute to care for young people with a mild intellectual disability. Reading guide to the chapters of this dissertation In the remainder of this introduction, I will provide a reading guide to the various chapters of this dissertation. This dissertation contains five chapters which – at face value – may seem quite different. The first two chapters assess the effectiveness of a substance abuse intervention. Then we evaluate the feasibility of daily self-monitoring. Another chapter concerns dynamic personality theory and network modeling. The last chapter then describes a case study about selfinjurious and aggressive patterns. I suspect that their coherence may not be glaringly obvious when reading the chapters separately, but there is an important commonality. The common denominator is that they all evaluate when and why problematic behaviors of young people with a mild intellectual disability change over time. In each chapter, the methodology with which change is assessed (i.e., the microscope lens) is different. The first chapter will start out nomothetic, 1 The inspiration for this microscope metaphor came to me from Anna Bosman, who had it from Thelen and Smith (1994).
23 General Introduction 1 looking through the lens that provides a full view. With each chapter that follows, the microscope will be zoomed in step-by-step. The last chapter's perspective will be idiographic: fully zoomed in. In the next paragraphs, I will introduce the lens of each chapter in more detail. In Chapter 2 we look through a wide-angle lens. This nomothetic approach is common practice in intellectual disability research and social sciences in general because it permits doing that which most scientists consider to be the ultimate goal: drawing conclusions that generalize to the population. Typically, the question of whether an intervention program is effective or not is approached with a group-level research design. Essentially, randomized controlled studies ask whether or not some form of treatment elicited behavioral change beyond what may be expected based on chance. Conclusions will be drawn for the group, so recruiting a large sample size is important for power considerations. The sample’s characteristics do need to be representative of the population you wish to make claims about. Change can then be assessed with at least two measurements per person, for example comparing a baseline screening to a follow-up assessment. The most typical design to test the effectiveness of an intervention program is then to compare if two groups differ from each other in their change rate. Figure 2 A hypothetical example of Chapter 2’s lens for studying change Note. A hypothetical example of differences in average change of an intervention group (undashed line) and a control group (dashed line) between a baseline and follow-up screening. Figure 2 illustrates a hypothetical result of a study that compares groups. The two lines summarize how all individuals changed between two time
24 Chapter 1 points: the baseline screening and a follow-up screening two months later. Scores on baseline and follow-up screening are averaged across individuals. Because there are just two measurement occasions, the change trajectory between them is then visualized with a straight line. The undashed line reflects the average change of a group of people who received a specific treatment. Their problem behavior was reduced more than a group of people who did not receive that treatment (the dashed line). In this example, we may conclude the intervention is effective. That conclusion (ideally) generalizes to other people, which implies that the nomothetically obtained knowledge can be used to help others (Hekler et al., 2018). Chapter 2 of this dissertation describes a two-arm intervention trial for a major problem in specialized care facilities, but for which there is a dearth of effective programs: alcohol and drug use. The intervention program that was evaluated in this chapter targets personality traits in relation to substance use, and is called ‘Take it Personal!’. More detail on the program and background about substance use as a problem in care settings for those with mild intellectual disability is provided in the chapter. What is most relevant for now, is that this chapter uses the full-view lens that aims to find out whether Take it Personal! – on average – reduces alcohol and drug use in people with mild intellectual disability. Figure 3 A hypothetical example of Chapter 3's lens for studying change Note. A hypothetical example of how subgroup change, in this example men (blue) and women (pink), may differ from their group average change (black lines). In Chapter 3 the same sample from Chapter 2's Take it Personal! trial is studied, but this time with a different lens. The microscope is zoomed in a bit more. Figure 3 shows what this lens may hypothetically reveal: different
25 General Introduction 1 change slopes of subgroups. In a separate panel for the intervention group (left panel) and control group (right panel) you can each see two colored lines in addition to the same black average lines from Figure 2. These colored lines show different change slopes: the average of the men (blue) and women (pink). Looking at the average gender-subgroup slopes illustrates that conclusions solely based on the black-lined group average (Figure 2) scarcely tell the full story. Whereas, on average, those who received the intervention decreased their problem behaviors over the course of two months, Figure 3 shows that the conclusion holds in particular for the men. The right panel shows that, although the average of the control condition did not change, zooming in on women and men separately does reveal change: men decreased their problem behaviors whereas it increased for the girls. Men in both intervention and control group thus decreased their problems, albeit more in the intervention group than the control group. The problems of the women in the intervention group, however, remained unchanged and even increased in the control group. Zooming in on specific subgroups may thus reveal who profited most from the intervention, which can be extremely useful information for practice. In this example, it would allow practice to further investigate why the intervention does not achieve as good of a result for the women with a mild intellectual disability. Perhaps the women would be in need of different treatment methods. Chapter 3 zooms in on various subgroups besides gender, assessing whether Take it Personal! is more or less effective in reducing substance use for groups of individuals that score high on certain other specific behavioral problems. In short, we assess various potentially moderating effects with the aim to get closer to an answer who (do not) change over time. Pre-posttest designs provide insight into whether or not people changed, but not really into how they changed over time. With only two measurement occasions, the process of the individuals and the group can only be visualized as a linear, straight line from baseline until follow-up (see Figure 2 and Figure 3). A different data collection method that is often referred to as the daily diary method or Ecological Momentary Assessments (EMA; Shiffman et al., 2008) can map out the change process in more detail. Daily self-monitoring with EMA entails that participants rate their own experiences once per day for several weeks or months for example by using an app on their mobile phone (or else by pen-and-paper). During the past two decades, researchers have often applied this in various clinical populations (e.g., schizophrenic, depressed, addicted people), but only recently has this method found its way to research in people with mild intellectual disability (Bakkum et al., 2024; Gosens et al., 2023; Wilson et al., 2020). Due to the novelty of this approach
26 Chapter 1 in this target group, important questions are: how many of their diaries do they complete, and what do they think of doing so? Figure 4 A hypothetical example of Chapter 4's lens for studying change Note. A hypothetical example of what change processes of the individuals from Figure 3 may look like when data is collected with ecological momentary assessments between baseline and follow-up. In Chapter 4 we explore the feasibility of collecting daily self-ratings for people with a mild intellectual disability. If people with mild intellectual disability can adhere to such a protocol, it would allow researchers to collect timeseries of behavioral problems. Research in clinical settings that zoomed in on individuals' change processes using timeseries has shown that change very rarely resembles a straight line but is most often discontinuous and characterized by sudden changes (Hayes et al., 2007; Kazdin, 2019; Tang & DeRubeis, 1999; Topolinsky & Reber, 2010). Figure 4 illustrates what one is thus likely to find. The colored lines now differ from the colored ones in Figure 3 – they do not reflect averages of subgroups of people, but individual people's change trajectories. Imagine the purple line in Figure 3 is one of the women who followed the treatment. The average of women (Figure 3) would suggest no change occurred between two time points two months apart, but at the individual level, one can expect variability from day to day. Whereas average lines of Figure 2 and Figure 3 gave the impression that those people had barely changed, the timeseries of these same individuals in Figure 4 illustrate that things most definitely did change – and scarcely ever in ways that resemble the average group-level change. In the remaining two chapters we studied change at the individual level. That is, instead of analyzing a change process that is aggregated
27 General Introduction 1 across individuals (e.g., the average lines in Figure 2), the change process of each individual is modeled separately. This dissertation started out with case descriptions of Kevin, John, Naomi, and Kyla – how they share a mild intellectual disability diagnosis but still differ in terms of their care-, educational- and familial histories, their (dis)abilities, personalities, and severity and nature behavioral problems (Segeren et al., 2018; Soenen et al., 2012; Nouwens et al., 2017). An unanswered question remains: how (dis) similar are their behavioral change processes? Figure 5 A hypothetical example of Chapter 5's lens for studying change Note. Illustration of assessing each individual's change process separately one by one. In Chapter 5 the microscope's lens zooms in on one individual, after which it is moved around to focus on the next individual, and so on. Figure 5 visualizes what it means to analyze each individual's process one by one. To this end, we aim to model the personality structure of each individual. Personality research goes hand in hand with the idiographic approach because it has always been about conceptualizing what makes a person unique (Allport, 1937). Personality is also highly relevant to this dissertation because treatment protocols in care facilities for people with mild intellectual disability might be tailored to the individuals’ specific personality traits. In particular, Take it Personal! (Gosens et al., 2022; Schijven et al., 2020a), adapts the treatment protocol to someone’s personality traits. In other words, personality assessments aid the identification of which treatment
28 Chapter 1 form John will get and which one Kyla may receive. But how does one then model someone’s unique personality structure from EMA data? In Chapter 5 we apply an increasingly popular analytical technique called idiographic network modeling. This approach enables studying processes bottom-up: first modeling each individual separately to then later evaluate (dis)similarities in personality-related change processes between people with a mild intellectual disability. Crucially, idiographic modeling (or any other type of model) should be grounded on a theory about how one expects individual-level processes to unfold over time. The main focus of Chapter 5 is on (in)congruence between idiographic network modeling and principles of complex systems theory that originally inspired these models. I will preserve details on the theory, models, and their incongruences for this chapter, but will give one spoiler here: the microscope should – and can – be turned up even further. This is what we have done in Chapter 6, in which we have used the lens with the narrowest field of view: only focusing on one individual. In care facilities, professionals target and adapt to accommodate the needs of each individual client. This support requires understanding the individual and their problems. Caretaker Eric, for example, wanted to understand why Kevin started using weed again on that particular day. An understanding of when and why such behaviors occur would enable Eric to implement timely preventive efforts in the future, which is evidently preferable over managing problem behaviors after they already happened. As with Kevin, care professionals all ask themselves when and why behaviors occur. For example, when and why John has aggressive outbursts, when and why Naomi self-harms, and when and why Kyla plays truant. An answer requires information about person- and time-specific context. Problem behaviors are highly contextualized: at any point in time they emerge from an interaction between the person and specific situational factors, like a quarrel with a loved one, a stressful exam coming up, peer pressure, or a traumatic event. By examining the context, one can assess whether there are temporal associations between patterns of specific situational factors and patterns of problem behaviors. The focus of the microscope’s next lens is therefore to qualitatively examine time- and person-specific contexts and pinpoint those on an idiographic timeline of problem behaviors. Figure 6 visualizes what such information may add to the interpretation of a client’s timeline. The last chapter describes a case study of a person with a mild intellectual disability with chronic aggressive and self-injurious behavior. The chapter details a timeline that shows how those behavioral patterns changed from day to day over the course of 560
29 General Introduction 1 days. Complex systems theory guides an exploration of why the pattern may (or may not) have changed at certain points in time. Figure 6 A hypothetical example of Chapter 6's lens for studying change Note. A hypothetical example of one person's timeline with qualitative information that provides some insight into why change may have occurred.
30 Chapter 1 Outline of the dissertation Chapter 1: General Introduction Chapter 2: The effectiveness of an indicated prevention programme for substance use in individuals with mild intellectual disabilities and borderline intellectual functioning: Results of a quasi-experimental study Chapter 3: Exploring the role of emotional and behavioral problems in a personality-targeted prevention program for substance use in adolescents and young adults with intellectual disability Chapter 4: The feasibility of daily monitoring in adolescents and young adults with mild intellectual disability or borderline intellectual functioning Chapter 5: Idiographic personality networks: Stability, variability and when they become problematic Chapter 6: A complex systems perspective on chronic aggression and self-injury: Case study of a woman with mild intellectual disability and borderline personality disorder Chapter 7: General Discussion My own change towards this PhD The chapters you are about to read also testify how my own perspective on how to study change, changed. Between 2015 and 2019 I studied at Radboud University, first a two-year Pre-Master’s program Educational Sciences and then the two-year Research Master’s program in Behavioural Science. What I learned in the curriculum was almost exclusively nomothetic. The scientific literature, the methods that were taught, and the required analyses were all nomothetic. Of course, nobody explicitly called these methods nomothetic. I realize now that the distinctions between nomothetic and idiographic principles play little to no part in the everyday lives of most social scientists because doing nomothetic research is so self-evident to the majority of scientists that it does not even need this label. Generally speaking, most scholars share the opinion that good science requires large sample sizes, assessed with standardized surveys that have proper psychometric qualities. Case study research was barely mentioned in my studies, and if at all, was deemed the lowest form of scientific evidence because generalization is not possible. I suppose that case study research was lacking in the curriculum because one cannot generalize from a case study to the population. This may be true, but is that a reason to only consider nomothetic designs? I think not. This general introduction thus can leave no doubt as to one thing: people (with a mild intellectual disability) differ from each other and change over time. I do not mean to insinuate that the existence of within- and
31 General Introduction 1 between-person differences in a dataset would actually be news to anyone at the university. Every statistics course I ever took (and there were many: descriptive statistics, inferential statistics, psychometrics and data analysis in the pre-masters curriculum and multivariate analysis, advanced statistics in R, mixed-effects models, and structural equation modeling during the Research Master) taught me various ways to account for such differences in the data. In regression analysis, for example, large variability around the group average automatically implies a large standard error, which reduces the significance of average effects. A statistical technique that is a bit more advanced is mixedeffects modeling, which can estimate average effects while controlling for individual variability by including random intercepts and/or random slopes. The implicit message from statistics – at least to me – was that inter- and intra-individual variance was something of a nuisance that statistically can and should be accounted for. When individual variation is filtered out, the path to making causal inferences between variables of interest (or "true effects", should they exist) is more open. Back then (and now still) I considered the pursuit of causal laws an important aim because it can produce knowledge to eventually help individuals. For example, it was nomothetic between-person research that taught the world that smoking causes lung cancer (Hill, 1965; cf. Borsboom et al., 2022) forming the basis for large-scale prevention and cessation programs which, eventually, benefit the individuals. It was with this nomothetic background that I arrived at the office of Roy Otten to discuss possibilities for a major research project. This was in 2018, at the start of my last academic study year. In preparation for that meeting, I had thought of some clever study design with which I thought I could impress. I can't really remember its details, but it of course required recruiting and analyzing a large clientele, to find out whether some kind of true effect existed. Roy grinned and told me that not only was setting up such a study unfeasible in one year within a residential care setting, but it was also not necessary. "Do a smaller number of extensive case studies, that's much more useful for clinical practice", he said. That was a stance I had not heard before, so I was surprised. It made much more sense a few weeks later when I started my major research project as an intern at the research department of the healthcare organization Pluryn. There I was co-supervised by Evelien Poelen, whose work showed that – to mental healthcare professionals and practiceoriented researchers alike – it was glaringly obvious that one size would never fit all. In practice, everyone particularly valued personalization. Each client, after all, is unique. A somewhat different take than at the university. I was lucky to be an intern at two projects on substance use, one nomothetic project
32 Chapter 1 by Esmée Schijven and one idiographic research project by Lotte Gosens. The former forms the basis for the first two chapters of this dissertation, while the latter project adopted a repeated single case study design by means of collecting EMA data. At that time, researchers at Pluryn (a.o. Roy Otten and Evelien Poelen) wondered how to best analyze EMA data. And so did I. Traditional nomothetical analyses felt somewhat underequipped to unravel the richness of these data. That year, I took the elective analytical course dynamics of complex systems, taught by Fred Hasselman, Maarten Wijnants, and Merlijn Olthof. They spoke about non-linear timeseries analyses to study the unicity of a single individual's behavioral dynamics. I probably understood not even 5% of what they talked about in the course (terms like periodicity, order parameters, fractal dimensions, bifurcations, dynamic complexity, entropy, spectral slopes, synergetics, and perturbations are enough to make anyone's head spin). Despite the avalanche of difficult terminology, one thing did stand out: this was the study of individuals, instead of averages of distributions (cf. McManus et al., 2023), which made it both fascinating and highly relevant. Moreover, it seemed an underexplored field of research. I had to learn more, which is why I joined the Complex Systems Group (which is now called CiBS group). I thought then that youngsters at Pluryn stood to gain considerable advantages if we could incorporate innovative insights from complexity science in daily care. Conversely, insights from real-world experiences in specialized care could ensure that complex systems theoretical principles remain grounded and applicable. Right before my PhD, I found myself somewhere in between practice and complexity science. In hindsight, it was inevitable that Anna Lichtwarck-Aschoff, whose research focus is as practiceoriented as it is complexity, would join Roy and Evelien in supervising the dissertation you are about to read. Acknowledgments I would like to thank Roy Otten, Evelien Poelen, Anna Lichtwarck-Aschoff, Anna Bosman, and Angela Luteijn for their feedback on earlier versions of this general introduction. Another special thanks goes to Emre Sezen, who made the four wonderful drawings you see at the beginning of this chapter.
www.ridderprint.nlRkJQdWJsaXNoZXIy MTk4NDMw