• No results found

Cognition and Drawing in Children with and without Autism Spectrum Disorder

N/A
N/A
Protected

Academic year: 2021

Share "Cognition and Drawing in Children with and without Autism Spectrum Disorder"

Copied!
201
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Cognition and Drawing in Children with and without Autism Spectrum Disorder by

Kayla D. Ten Eycke MSc, Laurentian University, 2010

BSc, Carleton University, 2008

A Dissertation Submitted in Partial Fulfillment of the Requirements for the Degree of

DOCTOR OF PHILOSOPHY in the Department of Psychology

 Kayla Ten Eycke, 2013 University of Victoria

All rights reserved. This dissertation may not be reproduced in whole or in part, by photocopy or other means, without the permission of the author.

(2)

Supervisory Committee

Cognition and Drawing in Children with and without Autism Spectrum Disorder by

Kayla D. Ten Eycke MSc, Laurentian University, 2010

BSc, Carleton University, 2008

Supervisory Committee

Dr. Ulrich Müller, Department of Psychology Supervisor

Dr. Jim Tanaka, Department of Psychology Departmental Member

Dr. Sarah Macoun, Department of Educational Psychology & Leadership Studies Outside Member

(3)

Abstract

Supervisory Committee

Dr. Ulrich Müller, Department of Psychology Supervisor

Dr. Jim Tanaka, Department of Psychology Departmental Member

Dr. Sarah Macoun, Department of Educational Psychology & Leadership Studies Outside Member

Children with autism spectrum disorder show imaginative and representational drawing deficits, despite reports of a “visual thinking style”. I examined whether these two drawing characteristics could be explained by the unique cognitive style of children with autism (specifically, executive dysfunction and a local processing bias). I

administered a cognitive/drawing task battery to a group of 24 school-age children with autism and 29 mental age-matched neurotypically developing controls. I expected that better executive function ability would be associated with better imaginative and

representational drawing, and that a local processing bias (weak central coherence) would be associated with better representational drawing but worse imaginative drawing. In children with autism, better executive function was associated with better imaginative drawing. Greater central coherence was associated with better representational drawing, but executive function was associated with worse representational drawing. Underlying cognitive components of imaginative and representational drawing were different for the neurotypically developing children. Overall, findings were unexpected, leading to novel theoretical suggestions for the field of autism cognition and drawing research.

(4)

Table of Contents

Supervisory Committee ... ii!

Abstract ... iii

Table of Contents ... iv

List of Tables ... vii!

List of Figures ... x!

Acknowledgements ... xi!

Dedication ... xiii!

Chapter 1: Introduction ... 1!

Background Information ... 2!

Cognition and ASD. ... 2!

Executive dysfunction. ... 2!

Weak central coherence. ... 4!

Linking executive function and weak central coherence. ... 6!

Drawing and autism. ... 7!

Visual realism and representation of visual information. ... 8!

Imagination and creativity. ... 11!

Summary. ... 13!

Overview of the Current Study ... 13!

Study Hypotheses... 14!

Hypothesis 1: Executive dysfunction impairs imaginative drawing. ... 15!

Hypothesis 2: Weak central coherence impairs imaginative drawing. ... 17!

Hypothesis 3: Executive dysfunction impairs representational drawing. ... 18!

Hypothesis 4: Weak central coherence promotes representational drawing. ... 20!

Hypothesis 5: A stronger role for EF and WCC in CWA. ... 22!

Summary. ... 23! Chapter 2: Method ... 24! Participants ... 24! Recruitment. ... 26! Procedure ... 26! Parent Questionnaire ... 27! Study Tasks ... 28! Control measures. ... 28! Screening task. ... 29!

Fine motor functioning: The motor supplement of the Beery Buktenica Developmental Test of Visual Motor Integration (Beery VMI) (Beery, Buktencia, Beery, 2010). ... 29!

Intelligence: Kaufman Brief Intelligence Test II (KBIT II) (Kaufman & Kaufman, 1990). ... 29!

Drawing Tasks. ... 29!

Representational drawing, copying a 2D figure: Rey Osterrieth Complex Figure (ROCF)– (Kuschner et al., 2009; Osterrieth, 1944). ... 30!

Representational drawing, drawings from 3D models. ... 30!

Imaginative drawing: The Karmiloff-Smith Draw an impossible person and house. ... 32!

(5)

Drawing evaluation. ... 33!

Cognitive tasks. ... 34!

Executive function (EF) tasks. ... 34!

Tower of Hanoi (ToH). ... 35!

Self-Ordered Pointing Task (SOPT) (Petrides & Milner, 1982). ... 36!

Overall executive function score. ... 37!

Central coherence tasks. ... 37!

Children’s Embedded Figure Task.(CEFT ;Witkin et al., 1971). ... 37!

Visual illusion. ... 37!

Overall central coherence score. ... 38!

Generativity tasks. ... 38!

Uses of objects. ... 38!

Semantic fluency. ... 39!

Design fluency. ... 39!

Overall generativity score. ... 40!

Ethical Approval ... 40!

Chapter 3: Results ... 41!

Overview of Analyses ... 41!

Interrater Reliability ... 42!

Group Differences ... 42!

Group differences in cognitive measures. ... 42!

Group differences in drawing measures. ... 44!

Zero-Order Correlational Analyses ... 49!

Cognition. ... 49!

Drawing. ... 52!

Cognition, drawing, and age. ... 55!

Cognitive Predictors of Drawing ... 60!

Autism spectrum disorder group. ... 60!

Imaginative drawing. ... 61!

Representational drawing. ... 61!

Rey Osterrieth Complex Figure. ... 61!

Spontaneous drawing. ... 63!

Neurotypically developing group. ... 65!

Imaginative drawing. ... 66!

Representational drawing. ... 67!

Rey Osterrieth Complex Figure. ... 68!

Spontaneous drawing. ... 69!

Summary of basic models of cognitive predictors of drawing. ... 71!

ASD-NT group differences in cognitive predictors of drawing. ... 72!

Detailed Analyses of Cognitive Predictors of Drawing ... 74!

Fine motor control and cognitive predictors of drawing. ... 74!

Generativity and cognitive predictors of drawing. ... 78!

Components of executive function and drawing. ... 82!

Components of weak central coherence and drawing. ... 84!

(6)

Neurotypically developing group. ... 91!

Overall Summary of Cognitive Predictors of Drawing Outcomes ... 94!

Autism spectrum disorder group. ... 94!

Neurotypically developing group. ... 96!

Parent Survey ... 97!

Chapter 4: Discussion ... 99!

Summary of Findings ... 99!

Hypothesis 1: Executive dysfunction impairs imaginative drawing. ... 100!

Hypothesis 2: Weak central coherence impairs imaginative drawing. ... 100!

Hypothesis 3: Executive dysfunction impairs representational drawing. ... 100!

Hypothesis 4: Weak central coherence promotes representational drawing. ... 101!

Hypothesis 5: A stronger role for EF and WCC in CWA. ... 101!

Group Differences in Cognition ... 102!

Executive function and generativity. ... 102!

Central coherence. ... 105!

Cognitive Predictors of Drawing ... 107!

Imaginative drawing. ... 107!

Imaginative drawing in children with autism spectrum disorder. ... 107!

Executive function and representational drawing. ... 108!

Central coherence and representational drawing. ... 111!

Additional influences on imaginative drawing. ... 115!

Group differences in house and person conditions of the imaginative drawing task. ... 122!

Imaginative drawing in the neurotypically developing group. ... 126!

Executive function and imaginative drawing. ... 126!

Representational drawing. ... 131!

Representational drawing in children with autism. ... 131!

Three-dimensional representational drawing ability. ... 132!

Two-dimensional representational drawing ability. ... 139!

Representational drawing in the neurotypically developing group. ... 146!

Three-dimensional representational drawing. ... 147!

Two-dimensional representational drawing. ... 148!

Spontaneous representational drawing. ... 150!

Explaining ASD-NT Differences and Similarities ... 151!

Summary ... 153!

Chapter 5: Conclusion ... 157!

Study Evaluation ... 157!

Directions for Future Research ... 162!

Implications of Findings ... 165!

Conclusion ... 167!

References ... 169!

Appendices ... 184!

Appendix A:Testing Session Schedule ... 184

Appendix B: Parent Questionnaire ... 185!

(7)

List of Tables

Table 1. Description of participants with autism spectrum disorder (ASD, N=25) and neurotypically developing children (NT, N=29). Group differences based on

independent-samples t-test are also indicated. ... 25! Table 2. Group differences in cognitive measures for children with autism spectrum

disorder (ASD) and neurotypically developing children (NT). ... 43! Table 3. Group differences in drawing measures for children with autism spectrum

disorder (ASD, N=25) and neurotypically developing children (NT, N=29). ... 48! Table 4. Bivariate Pearson correlation coefficients for three generativity tasks (uses of

objects, category fluency, and design fluency) for children with autism spectrum disorder (ASD, N=25). ... 49! Table 5. Bivariate Pearson correlation coefficients for three generativity tasks (uses of

objects, category fluency, and design fluency) for neurotypically developing children (NT, N=29). ... 50! Table 6. Bivariate Pearson correlation coefficients for two central coherence tasks

[children’s’ embedded figures task (CEFT) and visual illusion (illusion and

control)] for children with autism spectrum disorder (ASD, N=25). ... 50! Table 7. Bivariate Pearson correlation coefficients for two central coherence tasks

[children’s’ embedded figures task (CEFT) and visual illusion (illusion and

control)] for neurotypically developing children (NT, N=29). ... 51! Table 8. Bivariate Pearson correlation coefficients for two executive function tasks

[self-ordered-pointing task (SOPT, verbal and non-verbal), and Tower of Hanoi (ToH)] for children with autism spectrum disorder (ASD, N=25). ... 51! Table 9. Bivariate Pearson correlation coefficients for two executive function tasks

[self-ordered-pointing task (SOPT, verbal and non-verbal), and Tower of Hanoi (ToH)] for neurotypically developing children (NT, N=29). ... 52! Table 10. Bivariate Pearson correlation coefficients for representational drawing tasks

(meaningful and non-meaningful stimuli) for children with autism spectrum disorder (ASD, N=25). ... 53! Table 11. Bivariate Pearson correlation coefficients for representational drawing tasks

(meaningful and non-meaningful stimuli) for neurotypically developing children (NT, N=29). ... 53! Table 12. Bivariate Pearson correlation coefficients for Rey Osterrieth Complex Figure

(copy and recall conditions) for children with autism spectrum disorder (ASD, N=25). ... 54! Table 13. Bivariate Pearson correlation coefficients for Rey Osterrieth Complex Figure

(copy and recall conditions) for children with autism spectrum disorder (NT, N=29). ... 55! Table 14. Bivariate Pearson correlation coefficients for chronological age, drawing

composite scores (imaginative, representational, Rey Osterrieth Complex Figure, and spontaneous) and cognitive composite scores (generativity, executive function, and weak central coherence) for children with autism spectrum disorder (ASD, N=25). ... 57! Table 15. Bivariate Pearson correlation coefficients for chronological age, drawing

(8)

and weak central coherence) for children with autism spectrum disorder (NT,

N=29). ... 59! Table 16. Results for linear regression models containing mental age (MA), executive

function (EF), and central coherence (CC) as predictors of each drawing outcome for children with autism spectrum disorder (ASD, N=25). ... 64! Table 17. Results for linear regression models containing mental age (MA), executive

function (EF), and weak central coherence (WCC) as predictors of each drawing outcome for children with Autism Spectrum Disorder (ASD, N=25). ... 65! Table 18. Results for linear regression models containing mental age (MA), executive

function (EF), and central coherence (CC) as predictors of each drawing outcome for neurotypically developing children (NT, N=29). ... 70! Table 19. Results for linear regression models containing mental age (MA), executive

function (EF), and central coherence (WCC) as predictors of each drawing outcome for neurotypically developing children (NT, N=29). ... 71! Table 20. Results for linear regression models containing mental age (MA), executive

function (EF), weak central coherence (WCC) and fine motor control (motor control) as predictors of each drawing outcome for children with Autism Spectrum Disorder (ASD, N=25). ... 75! Table 21. Results for linear regression models containing mental age (MA), executive

function (EF), weak central coherence (WCC) and fine motor control (motor control) as predictors of each drawing outcome for as predictors of each drawing outcome for neurotypically developing children (NT, N=29). ... 77! Table 22. Results for linear regression models containing mental age (MA), executive

function (EF), weak central coherence (WCC) and generativity as predictors of each drawing outcome for children with Autism Spectrum Disorder (ASD, N=25). ... 79! Table 23. Results for linear regression models containing mental age (MA), executive

function (EF), weak central coherence (WCC) and generativity as predictors of each drawing outcome for neurotypically developing children (NT, N=29). ... 81! Table 24. Results for linear regression models containing mental age (MA), Tower of

Hanoi maximum level achieved (TOH), non-verbal self-ordered pointing task

maximum span achieved (NV SOPT) and verbal self-ordered pointing task maximum span achieved as predictors of each drawing outcome for neurotypically developing children (ASD, N=25). ... 83! Table 25. Results for linear regression models containing mental age (MA), Tower of

Hanoi maximum level achieved (TOH), non-verbal self-ordered pointing task

maximum span achieved (NV SOPT) and verbal self-ordered pointing task maximum span achieved as predictors of each drawing outcome for neurotypically developing children (NT, N=29). ... 84! Table 26. Results for linear regression models containing mental age (MA), Titchener

Circles Illusion score (Illusion) and Childhood Embedded Figures Task (CEFT) as predictors of each drawing outcome for children with Autism Spectrum Disorder (ASD, N=25). ... 85! Table 27. Results for linear regression models containing mental age (MA), Titchener

Circles Illusion score (Illusion) and Childhood Embedded Figures Task (CEFT) as predictors of each drawing outcome for neurotypically developing children (NT, N=29). ... 86!

(9)

Table 28. Results for linear regression models containing, executive function (EF), and central coherence (WCC) as predictors of each drawing outcome for children with autism spectrum disorder (ASD, N=25). ... 88! Table 29. Results for linear regression models containing non-verbal intelligence

(NV-IQ), executive function (EF), and central coherence (WCC) as predictors of each drawing outcome for children with autism spectrum disorder (ASD, N=25). ... 89! Table 30. Results for linear regression models containing verbal intelligence (V-IQ),

executive function (EF), and central coherence (WCC) as predictors of each

drawing outcome for children with autism spectrum disorder (ASD, N=25). ... 90! Table 31. Results for linear regression models containing chronological age (CA),

executive function (EF), and central coherence (WCC) as predictors of each

drawing outcome for children with autism spectrum disorder (ASD, N=25). ... 90! Table 32. Results for linear regression models containing, executive function (EF), and

central coherence (WCC) as predictors of each drawing outcome for neurotypically developing children (NT, N=29). ... 92! Table 33. Results for linear regression models non-verbal intelligence (NV-IQ), executive function (EF), and central coherence (WCC) as predictors of each drawing outcome for neurotypically developing children (NT, N=29). ... 93! Table 34. Results for linear regression models non-verbal intelligence (V-IQ), executive

function (EF), and central coherence (WCC) as predictors of each drawing outcome for neurotypically developing children (NT, N=29). ... 93! Table 35. Results for linear regression models containing chronological age (CA),

executive function (EF), and central coherence (WCC) as predictors of each

drawing outcome for neurotypically developing children (NT, N=29). ... 94! Table 36. Group differences in parent survey measures for children with autism spectrum disorder (ASD) and neurotypically developing children (NT). ... 97!

Table C1. Table of interrater reliability coefficients, based on Cohen’s Kappa for nominal data and interclass correlation coefficients (average measures) for

(10)

List of Figures

Figure 1. Titchener circles (Happé, 1999) ... 6! Figure 2. Study hypotheses summary: cognitive predictors of drawing outcomes. ... 23! Figure 4. Photographs of representational drawing task stimuli. A toy elephant (A) was

used as a meaningful stimulus, and a geometric foam figure (B) was used as a non-meaningful stimulus. ... 32! Figure 5. Mean proportion of imaginative content for impossible person and house

drawing tasks for children with autism spectrum disorder (ASD) and neurotypically developing children (NT). Error bars represent 2 standard errors. ... 45! Figure 6. Graph indicating moderating effect of mental age on the prediction of copy

accuracy score for the Rey Figure, by weak central coherence, for children with autism spectrum disorder. ... 63! Figure 7. Graph indicating moderating effect of mental age on the prediction of copy

accuracy score for the Rey Figure, by executive function, for children with autism spectrum disorder. ... 63! Figure 8. Graph indicating moderating effect of mental age on the prediction of

spontaneous drawing imagination score by weak central coherence, for children with autism spectrum disorder. ... 64! Figure 9. Graph indicating moderating effect of mental age on the prediction of overall

imaginative drawing score by executive function, for neurotypically developing children. ... 67! Figure 10. Graph indicating moderating effect of mental age on the prediction of overall

imaginative drawing score by weak central coherence, for neurotypically developing children. ... 67! Figure 11. Graph indicating a significant interaction between executive function and

weak central coherence in predicting spontaneous art realism scores for

neurotypically developing children. ... 70! Figure 12. Graph indicating a significant interaction between weak central coherence and

fine motor control in predicting organization scores of the recall condition of the Rey Osterrieth Complex Figure, for neurotypically developing children. ... 76! Figure 13. Graph indicating a significant moderating effect of generativity on the relation

between executive function and spontaneous art realism scores, for children with autism spectrum disorder. ... 79! Figure 14. Graph indicating a significant interaction between weak central coherence and

generativity in predicting non-meaningful representational drawing, for

neurotypically developing children. ... 81! Figure 15. Most commonly drawn topics reported by parents for children with autism

(11)

Acknowledgements

I am extremely grateful for the expertise, encouragement and guidance provided by my supervisor, Dr. Ulrich Müller and my committee, Dr. James Tanaka, and Dr. Sarah Macoun. I am very fortunate to have these people enrich and challenge my ideas.

I would like to acknowledge Dr. Kim Kerns, Dr. James Tanaka, and Dr. Sarah Macoun, and Dr. Ulrich Müller for their assistance with participant recruitment and for the supervision and use of their neuropsychological assessment tools. I would also like to thank Dr. Andrea Piccinin for her statistical expertise, and all of the teachers who have provided me with the skill and intellectual inspiration to complete this project.

This dissertation would not be possible without the research assistants who aided in coding art and video data: Cortney Mills, Eric Jourdan, and Janyn Mercado.

I would also like to thank Dr. Grace Iarocci, Sarah Hutchison, Kristina Andrew, Carolyn Sheehan, Lara Fieldman, Dan and Lisa LeBlanc, Stephanie Gilchrist, Kathe Gerber, Julia Toller, Darlene DeMerchant, Maxine Fisher, the Centre for Autism

Research, Technology, and Education; Child Care Resources, Mosaic Learning Society, Apples to Oranges Consulting, BC Association of Speech Language Pathologists and Audiologists, Little Steps Therapy, Autism Society of BC, Autism Ontario, Victoria School District 61, Victoria Society for Children with Autism, Pivot Point Family Growth Centre Inc., The Cridge Centre for the Family, Victoria Group Perspectives Therapy, Katherine Paxton Counselling Services, West Shore Family Naturopathic Ltd., The Making Tomorrow Conference, READ Society, Oak and Orca School, Queen Alexandra Centre for Children's Health, Chatterblock, Kids in Victoria, and the YAM Autism Walk, for their help in participant recruitment.

(12)

Last, I would like to thank the participants in this study and their families for their time, patience, and dedication to autism research.

Funding for this study was provided by the Sara Spencer Foundation Research Award.

(13)

Dedication

This dissertation is dedicated to my always encouraging Grandfather, Milton Ten Eycke; my ever-faithful parents, Vivian and Michael Ten Eycke; my brilliant and loving fiancé, Eric Jourdan; and the reliable and witty Erin and Clayton Kennedy. I will never forget their support throughout the process.

I also dedicate this work to young artists, who have brought passion and excitement to my research.

(14)

Chapter 1: Introduction

Artists with autism spectrum disorder such as Stephen Wiltshire, Jessy Park, Ping Lian Yeak, Richard Wawro, and Jonathan Lerman have made major contributions to art publications and documentaries, and have been displayed in mainstream galleries around the world (Mullin, 2009). These artists have savantism, a rare condition that combines an exceptional ability in one field (in this case, visual art) with severe intellectual disabilities and difficulty in almost all other areas (Happé & Frith, 2010; Treffert, 1990). The

interesting cases of famous artists with autism spectrum disorder (ASD, a set of disorders characterized by qualitative impairments in communication and social skills and by restricted and repetitive behaviours and interests; APA, 2000) have sparked interest in developmental disability research, the fine arts, and cognitive psychology. While art and artistic processes are important for the artist and the observer, art research can also shed light on the underlying thought processes involved in a neurodevelopmental disorder as complex as ASD. In this study, I explore how the cognitive characteristics of ASD relate to the art that children with this disorder produce. More specifically, I will explain how two aspects of ASD cognition (executive dysfunction and weak central coherence) relate to drawing outcomes (imaginative and representational drawing) in children with autism (CWA) and children without ASD (neurotypically developing, NT). In this chapter, I will first provide background information on the topics of cognition and drawing in ASD. This will provide the bases for deriving the hypotheses that will be investigated in this study. Then, I outline the current study.

(15)

Background Information

Cognition and ASD. Autism Spectrum Disorder (ASD) is a pervasive developmental disorder, characterized by the DSM-IV and ICD-10 as involving a qualitative impairment in social interaction; restricted, repetitive behaviour and/or interests; and abnormal/impaired development in one of three areas: social interaction, language as used in social communication, or symbolic or imaginative play (APA, 2000; World Health Organization, 1993). Many researchers from genetic, neurological,

cognitive, and social fields of research have attempted to explain why and how ASD symptoms arise. In this study, I examine two cognitive theories that have received a great deal of attention: weak central coherence theory (WCC; Happé & Firth, 2006), and executive dysfunction (Hill, 2004). These two theories are described below.

Executive dysfunction. According to executive dysfunction theory, inadequate executive functioning (“EF”) underlies the symptoms associated with ASD (Hill, 2004). Executive function, an umbrella term, refers to the cognitive processes that activate, organize, and integrate controlled cognitive processes (Lezak, 1993). More specifically, executive functions guide adaptive behaviours and problem-solving skills for attainment of a future goal (Garon, Bryson, & Smith, 2008). EF is commonly divided into

components (e.g. Miyake & Friedman, 2012). The components that appear to be important in linking ASD and drawing are planning, working memory, and mental flexibility.

Planning involves the ability to identify or organize a procedure for reaching an end goal (Jurado & Rosselli, 2007). Planning requires an individual not only to make an initial plan, but also to monitor, re-evaluate, and update her or his planned actions (Hill,

(16)

2004). Often, planning is measured using the Tower of Hanoi (or its variants), where participants move disks from a pre-arranged sequence to a goal sequence in as few moves as possible, while following a set of rules (Hill, 2004). Researchers have consistently observed impairments in planning in samples of children, adolescents, and adults with ASD, relative to a variety of control groups (including children with dyslexia, attention deficit hyperactivity disorder [ADHD], Tourette syndrome, and neurotypically

developing age- and IQ-matched controls) (Booth, Charlton, Hughes, & Happé, 2003; Happé et al., 2006; Hughes, 1996; Hughes et al., 1994; Milner, 1965; Ozonoff & Jensen, 1999; Ozonoff & McEvoy, 1994; Ozonoff et al., 1991; Pellicano, 2007; Prior &

Hoffman, 1990; Robinson et al., 2009; Sinzig et al., 2008).

Working memory is the capacity to actively hold and manipulate information in mind (Baddeley, 1983). While most research supports preserved working memory function among those with ASD (Lopez et al., 2005; Mottron et al., 1996; Ozonoff & Strayer, 2001; Russell et al., 1996), there are also some studies that find impairment in individuals with autism (Bennetto et al., 1996). Thus, findings on working memory deficits in children with ASD are inconsistent. This may be due to the impurity of working memory tasks; that is, most working memory tasks also demand mental flexibility and inhibition (Jurado & Rosselli, 2007).

Mental flexibility (also referred to as shifting-set) is the ability to switch rapidly between different response sets (Garon, Bryson, & Smith, 2008; Jurado & Rosselli, 2007). Based on research using card-sorting tasks, high-functioning children with ASD seem to have poor mental flexibility, relative to both neurotypically developing

(17)

disorders, Tourette’s syndrome, and dyslexia) (Bennetto et al., 1996; Corbett et al., 2009; Heaton et al., 1993; Hughes et al., 1994; Nelson, 1976; Ozonoff& Jensen, 1999; Ozonoff et al., 1991; Ozonoff et al., 2004; Ozonoff, 1997; Pellicano, 2007). However, a number of researchers have also found weak or unclear relations between mental flexibility and ASD (Geurts et al., 2009; Liss et al., 2001; Minshew, Golstein, Muenz, & Payton, 1992; Robinson et al., 2009; Yerys et al., 2009). Such inconsistency in mental flexibility research may be due to different underlying components of mental flexibility. Ozonoff and colleagues (2007) suggest that mental flexibility be divided into perceptual flexibility and conceptual flexibility, and individuals with autism are thought to struggle with conceptual shifts rather than the perceptual shifts (Hughes et al., 1994; Ozonoff et al., 2007). Individuals with autism exhibit perseveration (a failure to shift between sets), known as “stuck-in-set” inflexibility (Hill, 2006).

Note that planning, mental flexibility, and working memory are important in a number of cognitive processes. For example, these executive functions influence rule and concept formation, shifting when formulating problem-solving strategies, and utilizing models to plan and devise alternative ways to achieve solutions to problems (Ameli, Courchesne, Lincoln, Kaufman, & Grillon, 1988; Minshew et al., 1992; Prior & Hoffman, 1990l Rumsey, 1985; Schneider & Asarnow, 1987; Szatamari et al., 1990).

Weak central coherence. The weak central coherence account of ASD consists of a set of theories that propose that individuals with ASD employ specific

cognitive-perceptual strategies and styles (note that I refer to the collective set of theories as “weak central coherence”, although this is not the term adopted by all proponents described below). The original WCC account of ASD was based on the notion that NT

(18)

(neurotypically developing) individuals process (perceptually and cognitively)

information and access higher-level meaning by extracting an overall meaning or gist; by contrast, individuals with ASD were proposed to have a weak or absent drive to extract “global coherence” (Frith, 1989, 2003; Frith & Happé, 1994; Happé, 1999). According to WCC, this weak central coherence results in processing information in a detail-focused or local manner while neglecting global and contextualized meaning (Frith, 1989; Frith & Happé, 1994). People with ASD are thought to see wholes as decomposed into parts rather than as “unified Gestalts” (Pring, Hermelin, & Heavy, 1995). More contemporary accounts suggest that WCC may not be a deficit in central or global coherence, but instead a distinct cognitive style whereby CWA show a preference for local processing (Happé 1999). Although the superior ability to process details may occur at the expense of context and global features, WCC theory now posits that ASD processing enhances the representation of local details and features despite an ability to properly integrate the local features into a coherent, global item (Plaisted et al., 2003, 2009). According to the Enhanced Perceptual Functioning model, children with autism perform well on “low-level” cognitive tasks due to an enhanced perceptual ability (including feature detection, pattern recognition, enhanced visual memory, etc.). The model suggests that a local orientation is the “default setting” of individuals with autism, while higher-order

processing is an optional strategy (Mottron, Dawson, Soulieres, Hubert, & Burack, 2006). This bias is reflected by the tendency for people with ASD to succeed on some tasks used to measure central coherence and fail on others.

(19)

People with ASD tend to exhibit WCC on visual tasks such as embedded figures tasks (Shah & Frith, 1983), block design tasks (Shah &Frith, 1993), and tasks assessing susceptibility to visual illusions (Happé, 1996, 1999). Shah and Frith (1989) found that CWA were faster than matched controls when asked to identify a figure or shape hidden

within a more complex picture (e.g., a triangle within a pram; the embedded figures task), suggesting that CWA attend to details better than NT children. Furthermore, Happé (1996, 1999) found that CWA are less likely to succumb to visual illusions such as the Titchener circles (see Figure 1), in which the presence of surrounding circles affects a NT child’s ability to judge whether the inner circles are really the same size. Happé (1996) found that CWA were more likely to indicate that that the inner circles were in fact the same size, presumably because CWA perceive and understand in a piecemeal way and neglect the global coherence of an image.

On the other hand, other researchers (e.g. Mottron et al., 1999; Ozonoff et al., 1994; Plaisted et al., 2003, 2009) have found children with ASD to show no preference for local detail, suggesting children with ASD can process global features normally.

Linking executive function and weak central coherence. It is important to note that difficulty with central coherence in CWA may rely on underlying EF skills (Mottron, Belleville, & Menard, 1999; Rinehart, Bradshaw, Moss, Brerton, & Tonge, 2000). For example, shifting between local and global components may be due to a difficulty with mental flexibility (Rinehart et al., 2000; Mann & Walker, 2003). Others suggest that WCC may cause poor EF, presumably because EF requires the ability to integrate

Figure 1. Titchener circles (Happé, 1999)

(20)

information from multiple sources (e.g., contextual cues) (Bennetto, Pennington, & Rogers, 1996; Pennington et al., 1997). Cross-sectional research has provided mixed findings on the link between EF and WCC (Booth, Charlton, Hughes, & Happé, 2003; Pellicano et al, 2006; Teunisse, Cools, van Spaendonck, Aerts, & Berger, 2001). Notably, in a longitudinal study by Pellicano (2010), no developmental relation between central coherence (measured with three visuo-spatial central coherence tasks) and EF (including planning, mental flexibility, and inhibition) were found in CWA.

Drawing and autism. I chose to analyze art based on two global aesthetic

features: visual realism or representational drawing and imagination/creativity. These two categories were chosen based on the idea that the aesthetic experience of a piece of art relies on the proper balance between how realistically the art represents aspects of the real world, and how novel and creative the art is (Cupchik, 2002). On one end of this spectrum, an artist is very close and consumed in her or his art (close aesthetic distance), and on the opposite end, an artist is very distantly removed from her or his art (distal aesthetic distance). While a high degree of visual realism reflects a proximal aesthetic distance, distance can be forced between an artist and his or her art by increasing the degree of creativity and imagination in the art (Cupchik, 2002).

It is important to note that defining representational and imaginative drawing varies between studies and fields of research. In this study, “representational drawing” refers to the process of depicting a specific object that the child has seen, either as a model in front of him or herself, or something specific from the child’s memory. On the other hand, “imaginative drawing” refers to drawing something novel that the child has created, and that is not available in the child’s immediate sight. In this study,

(21)

representational drawing and imaginative drawing were measured using drawing tasks in a lab setting. However, “spontaneous drawings” refers to drawings produced by the child without prompting or suggestion. Spontaneous drawings were completed at the child’s home or school, and submitted by parents.

Visual realism and representation of visual information. An important aspect of realistic drawing is the ability to represent and alter visual information in mind.

Researchers generally agree that the mental images that people generate are depictive, in that we have a picture in mind that describes locations in space and relations between objects, and that these representations are interpreted and changed by other cognitive processes (Kosslyn, 1994). The representation acts as a symbol for visual information and as a tool to categorize similarities in visual percepts. It is this type of representation that a young NT child uses when he or she draws a picture (Arnheim, 1956, 1970, 1974;

Luquet, 1927, 2001; Marr, 1982; Piaget & Inhelder, 1956; Willats, 2005). The use of graphic schemes and internal visual representations has been

extensively studied by analyzing the way children draw. To take a popular example, if a child is shown a teacup with the handle occluded and asked to draw it “exactly as it appears to you”, young children (under 8 years) will include the handle in their drawing. This is an example of intellectual realism; the child uses his or her internal representation of a teacup in the production of his or her drawing. This representation contains

conceptual knowledge and salient features of a teacup (e.g., the handle). Older children override this conceptual knowledge and draw the teacup as it actually appears (with the handle occluded) (Freeman, 1972). In general, the accuracy of a drawing decreases as people consider the conceptual properties of the object (Goodnow, 1978; Freeman, 1972;

(22)

Lee, 1989; Moore, 1986; Sheppard, Mitchell, & Ropar, 2008; Mitchell et al., 2005; Phillips et al., 1978).

Luquet (1927, 2001) best articulated the application of graphic representations to the development of realistic drawing in children in his four stages of realism. Luquet suggested that a child first achieves realism when he or she notices a similarity between his or her scribble and an actual object (“fortuitous realism”), but over time, the child draws something real with intention, albeit with many imperfections (“failed realism”). In the third stage, “intellectual realism”, a child begins to use symbolic activity in his or her drawing. Intellectually realistic drawings are based on “canonical representations” (which are similar to invariants of structure, object-centred representations, etc.), which are stereotypical, and represent visual information that is derived from multiple viewpoints and prior intellectual knowledge of the object over time (Freeman, 1980; Luquet, 1927, 2001). Children younger than 6 or 7 usually produce intellectually realistic drawings, and they prefer to use an internal model of a drawing, rarely looking at an actual model for visual information (Ford & Lord Rees, 2008; Luquet, 2001). Children then shift into the fourth stage, “visual realism”. Drawings with a high degree of visual realism are based on single fixed viewpoints that illustrate objects as they actually appear (Luquet 1927, 2001). With the emergence of visual realism, children spend more time observing a model before drawing (Willats, 2005).

Based on difficulties with conceptualization that have been reported in CWA (Frith, 1989, 2003; Frith & Happé, 1994; Happé, 1999), many researchers have hypothesized that the drawings of individuals with ASD would exhibit superior visual realism (Charman & Baron-Cohen, 1993; Ropar & Mitchell, 2002; Selfe 1983, 2011;

(23)

Sheppard et al., 2007; Snyder & Thomas, 1997). According to this idea, a lack of expectations that are based on conceptual knowledge allows CWA to draw what they actually see. In other words, CWA would be immune to the influence of prior knowledge and conceptual analysis on their perception and internal representations of the object, and instead draw exactly what they see with a great deal of visual accuracy (Selfe, 2011; Snyder & Thomas, 1997). Indeed, some studies suggest that CWA show superior visual realism (Charman & Baron-Cohen, 1993; Sheppard, Ropar, and Mitchell, 2007; Ropar & Mitchell, 2002; Sheppard 2007, 2009). However, several other studies using a number of different visual realism tasks do not clearly support the hypothesis of superior visual realism in CWA. For example, Ford and Lord Rees (2008) tested representational drawing in CWA, children with Down syndrome, and NT children. They had children draw a teacup and teapot that were scored for intellectual realism based on the omission of visible decorative attributes, or the commission of hidden categorical attributes (i.e., portraying a teacup handle which would suggest intellectual realism). There were no differences in commission errors of the occluded teacup handle or omission errors of the decorative detail between CWA and NT children. Further studies suggest that CWA draw with a similar degree of visual realism as NT controls (and in some cases, inferior use of visual realism) (Charman & Baron-Cohen, 1993; Eames & Cox, 1994; Hodgson & McGonigle-Chalmers, 2011).

Although there is consensus that CWA do maintain internal visual

representations, these studies have provided mixed findings regarding superior visual realism in CWA. In some cases, CWA exhibit superior visual realism, particularly involving structural and dimensional information (Sheppard, 2007, 2009), while in other

(24)

cases, CWA’s drawings do not differ from those of NT children (Charman & Baron-Cohen, 1993). The degree of conceptualization might distinguish contradicting findings. For example, only when conceptualization is most difficult do CWA prevail in visual realism, owing to a floor effect when conceptualization demands are low. Or perhaps inconsistency in results is due to the aspect of visual realism that is assessed; that is, researchers find different results when considering occlusion versus inclusion of detail, etc. The type of visual information might also explain inconsistent findings; internal representations might include structural information such as dimensionality, semantic and categorical information such as a teacup handle; or episodic information obtained from the current experience with the model. This might explain why ASD art is often focused on structural and dimensional elements (Burton, 2010; Kellman, 2010). Selfe (2011), whose research is based on savant artists with ASD, suggests that incongruent and null findings on visual realism drawing tasks in CWA are to be expected given the high degree of heterogeneity in cognitive profiles of CWA.

Imagination and creativity. According to Weisberg (1986), creativity involves problem solving (producing a novel response that solves a problem). Imagination involves the visualization (in the case of visual imagination) of a solution to a problem, which is a component of creative thinking (along with retrieving stored knowledge about the problem at hand) (Weisberg, 1986). Although imagination is a necessary component of creativity, imagination also occurs independently of solving a problem (Weisberg, 1986). Imagination could refer to the creation of ideas and images that are not available to the senses (Harris, 2000).

(25)

According to Vygotsky (2004), creative behaviour is a combinatorial activity, involving the creation of new images or actions by combining or reworking elements of past experiences (hence “combinatorial”). Thus, everything the imagination creates is a combination of different aspects of reality, and thus largely depends on previous experience. Importantly, Vygotsky suggests that the imagination is responsible for generating both novel real and unreal ideas.

CWA are thought to have an imagination deficit, a claim based on research on pretend play; CWA produce pretend play acts at slower rates than controls (Jarrold, Boucher, & Smith, 1996) and are less likely to produce spontaneous pretend play (Baron-Cohen, 1987; Jarrold et al., 1994; Ungerar & Sigmen, 1981; Wing & Gould, 1979). However, with external prompts and instruction, CWA can engage in pretense (Charman& Baron-Cohen, 1997; Jarrold et al., 1996; Jarrold et al., 1993; Lewis & Boucher, 1988).

Difficulties with imaginative drawing are also widely reported in the literature (e.g. Craig et al., 2001). For example, successive drawings of everyday objects produced by CWA have a significantly higher degree of thematic relatedness than those produced by NT controls, and CWA exhibit a restricted range of idiosyncratic ideas and topics in their drawings (Craig et al., 2001; Burton, 2010; Lewis & Boucher, 1991; Selfe, 2011). Numerous experimental studies also support creativity impairments in CWA (Craig et al., 2001; Scott & Baron-Cohen, 1996). For example, Scott and Baron-Cohen (1996)

administered the Karmiloff-Smith (1990) “draw an impossible person” task to 15 CWA (mean age 13;0), 14 children with cognitive delay (mean age 12;8), and 15 NT children (mean age 4;10). CWA preformed significantly worse than matched controls. Older NT

(26)

children added elements from other conceptual categories, and younger NT children deleted elements or changed the size or shape of real elements, whereas 92 % of CWA drew real, rather than impossible, people. Moreover, this effect was replicated with different samples of CWA (Low et al., 2009) and higher-functioning CWA (Craig et al., 2001).

Summary. So far I have described some of the common underlying cognitive features in individuals with ASD and the art that they produce. It is common for people with ASD to have poor executive functioning and a bias toward local rather than global features. While it remains unclear whether CWA have superior visual realism compared to NT controls, there are likely difficulties with imaginative drawing. In the hypotheses section of this report, I suggest how these cognitive atypicalities could explain drawing outcomes. I organize these links based on five study hypotheses. First, however, I give a brief overview of the study.

Overview of the Current Study

The topics of drawing and cognitive development in ASD are frequently studied. However, to the best of my knowledge, researchers have not yet conducted research linking the various aspects of cognition and drawing in children with autism. This leaves a gap in what we know about the disorder, and prompted me to ask the question, Can the unique cognitive style of children with ASD, specifically executive dysfunction and weak central coherence, explain the drawings these children produce in terms of visual realism and imagination? I attempted to answer this question using a cross-sectional mixed design that allowed me to compare children with ASD and neurotypically developing controls based on a battery of tasks including measures of executive function, central

(27)

coherence, visual realism, and imaginative drawing. To measure executive function, I used two global executive function tasks, the Self-Ordered Pointing Task (SOPT, Petrides & Milner, 1982), and the Tower of Hanoi (adapted from Welsh, 1991). I also measured generativity, using the the uses for objects task, the category fluency task (Turner, 1999), and the design fluency task (Baldo et al., 2001). I measured central coherence using a visual illusion task (the Titchener Illusion, Happé, 1993) and the Children’s Embedded Figures Task (CEFT ;Witkin, Oltman, Raskin, & Kay, 1971). Visual realism was assessed based on children’s drawings of two, three-dimensional figures (a toy elephant and a non-meaningful clay figure), and of the Rey Osterrieth Complex Figure (Kuschner et al., 2009; Osterrieth, 1944). Imaginative drawing was assessed using the Karmiloff-Smith Impossible Person and House task (Karmiloff-Smith (1990). Moreover, I examined the spontaneous drawings produced by children to gain a better understanding of the global aesthetic impression of children’s drawings. School-aged children (4-14 years of age) were included in the study to determine developmental trends in the link between cognition and drawing. I measured verbal and non-verbal intelligence (using the Kaufman Brief Intelligence Test II; Kaufman & Kaufman, 1990) and motor skill (using the motor supplement of the Beery Buktenica Developmental Test of Visual Motor Integration; Beery, Buktencia, Beery, 2010) to control for variables that likely contribute to drawing ability in addition to the specific cognitive factors that are a focus of this study.

Study Hypotheses. This study outlines four main hypotheses that link two predictors, cognitive theories of ASD (executive dysfunction and weak central

(28)

coherence), with two outcomes, representational drawing and imaginative drawing. Additionally, I consider one main expected difference between CWA and NT controls.

Hypothesis 1: Executive dysfunction impairs imaginative drawing. I predict that executive dysfunction will impair imaginative drawing in CWA. Executive functions are thought to be important in imaginative drawing because in order to create a novel

representation or drawing, children must initially construct visuospatial representations of the generated drawing ideas, maintain the representation in working memory, and

monitor the unfolding drawing (Leevers & Harris, 1998; Low et al., 2009). For example, some researchers believe that difficulties in imaginative drawing are due to difficulty carrying out visuospatial plans to produce a novel drawing. According to this notion, imaginative drawing requires the use of new and complex visual plans for drawings, but CWA instead re-execute easy and familiar graphic schemes or internal representations due to planning impairments (Leevers& Harris, 1998; Low et al., 2009; Prior & Hoffman, 1990). As expected, in NT children, visuospatial planning (tested using shape contrast and axis rotation tasks) is important for novel picture production (Freeman, 1987; Low & Hollis, 2003; Thomas & Silk, 1990).

Other researchers have hypothesized that imaginative drawings are based on generativity (Craig & Baron-Cohen, 1999). Generativity is sometimes considered a component of executive function, and is defined as the generation of novel ideas or behaviours (Bishop & Norbury, 2005; Hill, 2004; Jarrold et al., 1993, 1996; Turner, 1997, 1999). Turner (1997) suggests that CWA have difficulty creating hypothetical schemes that are necessary for flexible thought and subsequently for imaginative play and drawing.

(29)

Another possibility is that poor mental flexibility explains creativity impairments in CWA. Mental flexibility would be important for over-riding routine or reality-based responses in order for novel ideas and drawings to emerge (Craig & Baron-Cohen, 1999; Craig et al., 2010; Hughes et al., 1994; Ozonoff et al., 1994). Moreover, NT children learn a repertoire of different patterns and configurations by generating numerous and flexible attempts and trials in drawing (Luquet, 2001). The repetition and rituals due to poor mental flexibility observed in CWA diminishes experimentation, resulting in regular and predictable internal representations and drawings (Burton, 2010). From another perspective, Burton (2010) suggests that mental flexibility is part of the NT drawing process. She argues that the narrative of a drawing is fluctuating and changeable in NT children, but such an approach to drawing is not preferred by CWA due to poor mental flexibility. Stories, meaning, and intention of drawings would be stable and

circumscribed, and ultimately, less creative.

Low and colleagues (2009) put the hypothesis that executive dysfunction explains poor creativity to the test, using a sample of 27 children; including CWA and NT

matched controls. They administered the Karmiloff-Smith drawing task (1990) and cognitive measures of generativity (new uses of regular objects, such as a brick; and a pattern meaning task) and visuospatial planning (completing mazes, taken from the Wechsler Intelligence Scale for Children; Wechsler, 1992). The drawings by individuals with ASD exhibited deficits in imaginative content, and this deficit was predicted by generativity, and further mediated by visuospatial planning ability. Interestingly, they did not find the same relation in the NT group. For NT children, generativity was linked to imaginative drawing, but this relation was mediated by false belief reasoning, not EF.

(30)

This suggests that generativity influences imaginative drawing in CWA and NT children, but in different ways. Low and colleagues (2009), among others (e.g. Kana et al., 2006; Barsalou, 1999; Williams, Bowler, & Jarrold, 2012) suggest a “visual thinking style” in CWA that involves a problem solving approach using image schemas.

To summarize, executive dysfunction is predicted to limit imaginative drawing, but likely only in CWA.

Hypothesis 2: Weak central coherence impairs imaginative drawing.

Surprisingly, very few researchers have proposed a role for WCC in creative drawing in ASD. Nevertheless, based on Vygotsky’s (2004) theory that imagination is a process of combining elements, I expected that a difficulty with (or bias away from) global

coherence would influence imaginative drawing. Moreover, a difficulty making connections and integrating levels of meaning would likely impede creativity (Burton, 2010). Chiefly, Low and colleagues (2009) suggest that WCC may contribute to “hierarchical levels of drawing production” (p. 440) in imaginative drawing, but they only reported that central coherence (measured using the embedded figures task) related to drawing style, failing to report whether a relation between WCC and imagination existed. A study by Craig and colleagues (2001) involved the examination of imagination in drawing tasks in CWA (and a subgroup of children with Asperger’s disorder), and verbal mental age-matched mild learning disability and NT controls. In addition to the task of drawing an impossible man (on which CWA were impaired only relative to children with learning disabilities), they administered a more challenging task that required children to produce novel real or unreal entities by mixing categories of images. CWA exhibited a specific deficit; they tended to draw two real, separate entities. This

(31)

might suggest an influence of WCC, however the researchers found that CWA could successfully join two other images together (making a “J” and a rotated “D” into an umbrella).

Taken together, I predict that weak central coherence will have negative effects on imaginative drawing.

Hypothesis 3: Executive dysfunction impairs representational drawing. The third hypothesis is that EF impairments in CWA are associated with poor visual realism. This is based on the idea that executive functions are employed in the creation and management of internal representations of visual information in NT (Guerin, Ska, & Belleville, 1999; Karmiloff-Smith, 1990, 2008; Morra, 1995, 2005; Mottron et al., 1999; Selfe, 2011; Spensley & Taylor, 1999). Morra (1995, 2005) suggests that three factors allow children to modify their representation. The first factor is the amount of attentional resources (“M-capacity”, the maximum number of schemes a person can simultaneously activate) that can be used to activate schemes. The second factor involves the automatic activation of figurative schemes based on perception of a model. The last factor is the activation of executive schemes that allow a child to set appropriate goals and monitor performance (Morra, 1995). These capacities and abilities refer to executive functions.

Working memory is implicated in representational drawing by limiting the resources that activate internal representations or schemes [M Capacity, Morra (2005)]. Modifying a stereotype (graphic representation) requires many active schemes, and working memory limits how many schemes a child can simultaneously activate or manipulate while drawing. It follows that poor working memory ability would result in an inability to modify graphic stereotypes and thus result in highly intellectually realistic

(32)

drawings (because stereotypical schemes are used rather than novel, more appropriate schemes) (Morra, 1995, 2005).

As mentioned above, Morra (2005) outlines the importance of executive schemes in representational drawing, which are a type of operative scheme that monitor and regulate performance to ensure that the modification of a stereotypical representation meets task demands (previously set goals), which is analogous to the executive planning function. In particular, when a drawing requires the integration of many schemes,

advanced planning would be required (because changing one element of a drawing would result in changes to another) (Morra, 2005).

Last, poor mental flexibility could lead to poor visual realism because representational drawing development is contingent upon exploratory scribbling and drawing (Luquet, 2001), which CWA find difficult (Burton, 2010), likely owing to poor mental flexibility. Moreover, perseverative inflexibility might result in fixation on a small set of internal representations (Burton, 2010). CWA may fail to explore, practice, or add novel graphic elements to their established set of schemes or representations (Burton, 2010). This would diminish visual realism ability because CWA would have a narrower range of representations that would consequently be more stereotypical than

comprehensive.

Previous research supports a role for executive function in representational drawing. For example, Morra (2005) examined the relations between children’s ability to draw people in movement (counter to the stereotypical representation of a static person), their ability to modify a scheme of a human into a drawing of a kangaroo, and their working memory capacity (a sample of 645 5-9 year-olds). He found that a

(33)

developmental increase in the ability to modify schemes could be accounted for by increased working memory capacity. As a second example, Booth and colleagues (2003) tested high-functioning 8-16 year old boys with ASD, who were matched on age and intelligence to children with ADHD and NT children, on a drawing task requiring planning. Children were asked to include a new element to a figure (e.g., including teeth when they copied a snowman). They found drawing planning impairments in CWA and children with ADHD. Further, planning problems corresponded to some of the

characteristics of spontaneous ASD art, such as a tendency to draw heads or arms that are disengaged from the body (Kellman, 2010).

Developmentally, age-related changes in executive function may explain improvements in representational drawing over time. For example, older children have increased working memory ability, allowing them to expand their collection of

representations and to incorporate direct perceptual information (Karmiloff-Smith, 1990, 2008).

Thus, poor executive function is expected to impair representational drawing in CWA. It is expected that older CWA might improve in executive abilities, leading to improvement in representational drawing.

Hypothesis 4: Weak central coherence promotes representational drawing. I hypothesized that weak central coherence would facilitate representational drawing during drawing tasks, but hinder the realism expressed in spontaneous drawing.

It is intuitively plausible that the local processing bias observed in CWA relates to greater representational drawing ability because it allows for acute and enhanced

(34)

tendency not to categorize, or hierarchically organize local features allows children to avoid the influence of underlying logic and knowledge, and to use step-by-step

progressive analysis of visual information, shutting out distortions due to conceptual or Gestalt based perception (Bryson, 2005; Mottron et al., 1999; Mottron & Belleville, 1993; Sheppard et al., 2008, 2009; Snyder & Thomas, 1997). In other words, WCC would relate to greater visual realism due to a bias for detailed visual information and detachment from conceptual knowledge of the model.

This hypothesis might also relate to the way CWA store their internal

representations of visual information. According to Morra’s (2005) and Willats’ (1995) ideas of graphic representations, graphic schemes have a hierarchical organization such that a scheme of a cat is composed of schemes of circles and of triangles, for example. At the expense of categorization, CWA might store representations more precisely and separately (in memory) rather than in conceptually or semantically related units. Although this might result in more precise drawings, CWA could also have a difficult time integrating those schemes when faced with a drawing task (Kellman, 2010; Willats, 1995; Bryson, 2005; Sacks, 1995).

Local processing biases and poor conceptualization, and/or a difficulty switching between or integrating local versus global features can be seen in spontaneous ASD art. For example, CWA often respond to only one of many possible different stimulus elements, resulting in black and white drawings (overlooking colour), or structure-focused art (overlooking shade and hue) (Bryson, 2005; Kellman, 2010; Rincover & Ducharme, 1987). CWA also often fail to properly integrate and hierarchize aspects of their drawings (for example, by giving equal salience to foreground and background),

(35)

resulting in a flat appearance that lacks depth (Burton, 2010; Selfe, 2011). Sacks (1995) suggests that WCC accounts for unconnected artwork, “devoid of continuity and

development”. This is likely why we observe few contextual cues and instead see drifting features (and disembodied limbs) (Kellman, 2010).

Drake and colleagues (2010) examined drawings created by 27 6-12 year-old NT children. They found that a drawing score (based on the level of realism) predicted a local processing bias in their sample. They also found that the drawing score predicted the frequency of repetitive behaviours in ASD (assessed by the Childhood Asperger Syndrome Test).

To summarize, it is expected that a local processing bias in CWA will promote drawing with a high degree of visual realism on drawing tasks. However, when

spontaneous drawings are considered, WCC will manifest as difficulty with overall composition of drawings and inclusion of fine detail.

Hypothesis 5: A stronger role for EF and WCC in CWA. Because I expect that drawing processes in NT developing children rely more heavily on social processes such as ToM and language development, my final hypothesis was that the links between EF and WCC and drawing would be stronger in CWA than in NT controls. I did not predict an absent connection between these cognitive functions/styles and drawing outcomes in NT children. The intent of the current study was to better understand drawing and cognition in CWA, and so few predictions were made about the NT control group. They were included to determine whether the above model, linking cognition and drawing, is specific to the disorder.

(36)

Summary. The diagram below (Figure 2) summarizes the four main study hypotheses. Notice that I predicted visual realism to be diminished by executive dysfunction although promoted by weak central coherence. On the other hand, both executive dysfunction and weak central coherence result in poor imaginative drawing.

(37)

Chapter 2: Method

Participants

71 school-aged children (4-14 years) were recruited to participate in this study. Children with ASD (autism spectrum disorder) and NT (neurotypically developing) children were invited to participate. Six children (with ASD) were excluded because they failed the drawing screening task (copying three simple shapes). Three children in the NT group were excluded because their parents reported a diagnosis of a developmental disorder other than ASD (ADHD, Sensory Processing Disorder). Thirteen children with ASD whose parents reported a comorbid diagnosis of an additional developmental disorder were included. Among these children, one child had Disassociation Disorder, eight children had ADHD, four children had Sensory Processing Disorder, three children had Obsessive Compulsive Disorder, one child had Developmental Coordination

Disorder, and four children had learning disorders (including Dyslexia).

Two children in the NT group were excluded because their Autism Index Scores on the Gilliam Autism Rating Scale (GARS-2, Gilliam, 2006) fell within the “possibly” or “very likely” range for ASD (a score of 70 or higher), and one child in the NT group was excluded because his twin sister was in the ASD group. Parents of children with ASD were asked to provide proof of ASD diagnosis from a qualified specialist. This form indicated that the child has been diagnosed using ADOS (Autism Diagnostic Observation Schedule; Lord, Rutter, DiLavore, & Risi, 2001) and ADIR (Autism Diagnostic

Interview, Revised; Rutter, le Couteur, & Lord, 2003) diagnostic tools (a current diagnostic standard for diagnosis in British Columbia). This ensured that all children in the ASD group had been formally diagnosed with autism spectrum disorder. Of the

(38)

children in the ASD group, 19 had autistic disorder, 2 had PDD-NOS, and 4 had Asperger syndrome.

Children from the ASD and NT group were matched on Non-verbal IQ; Verbal mental age, Non-verbal mental age, overall mental age; and chronological age, by

dropping the 5 youngest participants in the NT group. All five children that were dropped were younger than 5 years old, and no children in the ASD group were younger than 5. Intelligence was assessed with the Kaufman Brief Intelligence Test II (KBIT; Kaufman & Kaufman, 1990). Based on exclusion criteria and matching, a total of six children were dropped from the ASD group, and a total of 11 children were dropped from the NT group, leaving 25 children with ASD and 29 neurotypically developing children. After matching, children with ASD ranged in mental age from 58-194 months old, and NT children ranged in mental age from 61-185 months old.

All children were tested in British Columbia or Ontario. Refer to Table 1 for a description of participant gender, age, IQ, fine motor control, and autism index for the ASD and NT groups.

Table 1. Description of participants with autism spectrum disorder (ASD, N=25) and neurotypically developing children (NT, N=29). Group differences based on independent-samples t-test are also indicated.

ASD NT

M SD M SD

Age (in months)

Chronological Age 115.3 31.5 103.34 28.7

Non-verbal Mental Age 100.85 39.4 102.47 36.8

Verbal Mental Age 107.5 44.2 113.7 36.7

Overall Mental Age 107.4 37.1 109.0 37.2

Intelligence

Non-verbal IQ 88.6 25.48 98.41 16.7

Verbal IQ* 92.7 24.0 109.3 12.8

(39)

Recruitment. Participants were recruited via word of mouth, advertisements on parent websites (e.g. Kids In Victoria .com), social media (e.g. Facebook), posters in the community, and advertising in local school boards.

Upon contact with the experimenter, parents received a copy of the study consent form, which outlined the research study. Parents then had the opportunity to confirm their participation by scheduling a testing session. Parents were offered a copy of the visual schedule used during the testing session.

Procedure

The testing session began with a familiarization period (approximately 5

minutes), during which the experimenter conversed with the child and parent in order to reduce anxiety surrounding the testing session. The parent was asked to sign the consent form, and fill out a parent questionnaire, the Gilliam Autism Rating Scale-2 (GARS-2, Gilliam, 2006), and the Fine Motor Control Portion of the Vineland Adaptive Behaviour Scale (VABS, Sparrow, Cicchetti, & Balla, 2005).

Next, the experimenter explained the visual schedule to the child, and had him or her check off tasks as they were completed. At the beginning of the testing session, children were asked for their verbal assent. All children were then administered the

Autism Indices

Stereotyped Behaviours Index** 7.8 3.1 3.1 1.5

Communication Index** 7.7 3.2 2.6 1.7

Social Interaction Index** 8.3 3.6 2.5 1.5

Autism Index** 85.1 22.8 52.6 8.5

Fine Motor Control

VABS Fine Motor Control Score* 14.1 2.8 17.9 8.5 Beery Fine Motor Control Score** 76.8 15.5 94.5 13.8 *p<.05

(40)

screening task and, if successful, the control, cognitive, and drawing tasks described below. Children were administered control measures [fine motor control (Beery VMI; Beery, Buktencia, Beery, 2010) and intelligence screening (KBIT, Kaufman & Kaufman, 1990), see below], followed by drawing tasks, then cognitive tasks. A schedule of task administration can be found in Appendix A. Between the three task segments, children were given the option to take a break, and were encouraged to stand, use the washroom, have a snack, or talk with a parent or the experimenter. Additional breaks were given if needed. At the end of the testing session, children were thanked for their time, and invited to choose a small gift for participating. Testing sessions lasted approximately 1.5 hours for both groups.

All drawings were completed with a standard #2HB pencil (without an eraser) on white 216cm x 279cm copy paper, and were later transformed into electronic files for ease of rating. In all cases, children used their dominant hand to draw. All task instructions were given orally. All sessions were videotaped using a standard video camera. Testing sessions were held in the child’s home or in the University of Victoria Child Development Lab, depending on parent and child preference.

Parent Questionnaire

In addition to the tasks administered to participants, parents (or caregivers) were given a brief questionnaire about their child’s at-home drawing behaviour (see Appendix B). For example, parents were asked how much time their child spends drawing each week, their child’s enjoyment of drawing, and the topics their child usually draws.

To assess functional motor skills, parents completed the Fine Motor items of the Vineland Adaptive Behavior Scale (VABS). This scale consists of 36 items that require

(41)

parents to rate the frequency of particular fine-motor behaviours (e.g. “Cuts out complex shapes” or “Uses eraser without tearing paper”) on a 3-point scale.

To confirm ASD diagnosis and screen for children who may not have received an ASD diagnosis but still demonstrate the autism phenotype, all parents completed the Gilliam Autism Rating Scale-2 (GARS-2, Gilliam, 2006). The GARS-2 indicates the severity of autism symptoms in individuals between the ages of 3 and 22 years. Items on the GARS-2 are based on the definitions of autism adopted by the Autism Society of America and the Diagnostic and Statistical Manual of Mental Disorders: Fourth Edition-Text Revision (DSM-IV-TR, American Psychiatric Association, 2000).

Data collected from parent questionnaires was important for examining the contribution of behavioural and motor characteristics on drawing development, and for controlling for potential effects beyond cognitive differences.

Study Tasks

Study tasks were broken into three categories: control measures (a screening task, an intelligence test, and a fine motor control task), drawing tasks (including tasks of imaginative and representational drawing), and cognitive tasks (including executive function and central coherence tasks). Tasks were selected based on whether they were appropriate to administer to a wide age range, for reasons of practicality, and based on previous research.

Control measures. To control for task comprehension, fine motor control, and intelligence, I administered three control tasks: a screening task, the fine motor control supplement of the Beery Buktenica Developmental Test of Visual Motor Integration

Referenties

GERELATEERDE DOCUMENTEN

In case we failed to find slips of action in our behavioural data, we intended to use the automaticity score of not taking the pill during the beginning (i.e. the first three days)

This paper attempts to study how entrepreneurs’ stable psychological attributes such as thinking style influence entrepreneurial decision-making behaviors associated with the

Als tweede zal een inventarisatie worden uitgevoerd van vormen van communicatie waarmee ondernemers met kennis kunnen worden bereikt en waarbij er ruimte komt

Dit argument verschilt van het vorige argument (§ 3.1.1) in de zin dat de reden voor de kwalificatie als bijzondere persoonsgegevens niet is dat gegevens omtrent

However, where children, a husband and family form important reasons for these Nicaraguan women to leave Nicaragua, come to Costa Rica and stay there for a while, there are also

By experimental design, Grosshans and Zeisberger show price paths have the potential to influence investor satisfaction and risk tolerance, also evidence for the presence of

Each one of the research methods that were used as mentioned above contributed to respond to the main research question, namely: To what extent can a single Education

The saturated semicrystalline polymer (P1-H) is water-insoluble but undergoes rapid backbone hydrolysis under neutral, basic, or acidic conditions when polymer films were immersed