KATHOLIEKE UNIVERSITEIT LEUVEN FACULTEIT PSYCHOLOGIE EN PEDAGOGISCHE WETENSCHAPPEN Laboratorium voor Experimentele Psychologie Local - Global Visual Processing in ASD: Multiple Object Tracking and Gaborized Visual Search Master thesis presented to obtain the degree of Master in Psychology By Nele Soors Astrid Van Der Most Promotor: Prof. Johan Wagemans Co-promotor: Dr. Bart Boets Daily advisor: Ruth Van der Hallen While typically developing individuals are known to perceive the world as a coherent and structured whole, research findings suggest this to be different for individuals with autism spectrum disorder (ASD). Two major, non-social cognitive frameworks, the Weak Central Coherence theory (WCC) and the Enhanced Perceptual Functioning theory (EPF), have been dominating this discussion for years. Both theories have provided a tremendous contribution to the study of visual processing in ASD but, despite their effort, research thus far has led to rather ambiguous results. Part of this ambiguity seems to be due to task characteristics and, more specifically, due to differences in task instruction. Therefore, we investigated the possible influence of implicit versus explicit task characteristics on local versus global visual processing in a group of 9 - 13 year old children with high functioning ASD (HFA; FSIQ > 70; n=26) and typically developing children (TD; n=27) matched for age, gender and FSIQ. The research protocol entailed a visual search task with Gabor elements and a multiple object tracking task. Both task instruction and stimulus complexity were manipulated in both tasks. We hypothesized that children with ASD would show intact local processing in the implicit and explicit task condition, but impaired global processing in the implicit task conditions. Results for the Gaborized visual search task in terms of accuracy, revealed a significant effect of stimulus complexity for both groups and a differential effect of task instruction for both participant groups: While no group difference was present with explicit task instructions, the ASD group performed significantly worse than the TD group with implicit task instructions. Results for the multiple object tracking task, however, revealed no such group difference, but showed a similar pattern of grouping interference for both participant groups. Our results suggest that task characteristics play an important role in differences in local-global processing between children with and without ASD. ACKNOWLEDGEMENTS This research article on local – global visual processing in ASD is established through the intensive collaboration of Nele Soors and Astrid Van Der Most, two master students in clinical psychology, who had not met before. Working together on a joined product seemed evident from the start of the project. The challenges encountered throughout the research process were overcome thanks to the shared interests in the autism spectrum and the belief that together one can achieve more. However, this thesis could also not have been accomplished without the help and support of several people and institutions. Therefore, we want to use this section to express our special appreciation to all those people. First of all we would like to say thank you to Prof. Dr. Johan Wagemans, our promotor, for providing us with essential background information and positive feedback and support during this two-year project. To Dr. Bart Boets, our co-promotor, for giving us a head start on our literature research and for his valuable contribution and support with the data analyses as well as the feedback provided during the final stage of our research. To Ruth Van der Hallen, our daily advisor, for the tremendous amount of continuous support during this research project. All your efforts in providing us with the necessary insights, feedback and suggestions, seemingly endless patience and motivation has helped us meet the challenges that came about throughout the process and inspired us in this research experience. A second thank you goes to all our professors and teachers of the Department of Clinical Psychology for all the knowledge provided to us during our education. A third and sincere thank you goes to all participants in our study for their efforts in taking part in this research. Without their participation this entire research would not have been possible. You have been a source of inspiration to us. Last, but not least, a special thank you goes out to our parents, husband and boyfriend, siblings and friends, for their never-ending support, patience and interest in everything we did during the past years. Thank You. CONTRIBUTION AND INVOLVEMENT Inspired by the research topic and the range of different elements this study offered, we started a two-year journey in November 2012 studying local-global visual information processing in ASD by means of Gaborized visual search and Multiple object tracking. The reading, researching and writing has been conducted as autonomously as possible, though we consider the end result to be a true team effort in which we both provided an equal contribution. Nonetheless, we could not have accomplished this result without the feedback and help of Prof. Dr. Johan Wagemans, Dr. Bart Boets and Ruth Van der Hallen. The first phase of our research involved a thorough reading of the literature on the autism spectrum and visual processing. Supplementary experimental and diagnostic materials were previously selected by the research team and provided a framework for our research. Some basic background literature on visual processing was provided to us and complemented with self-searched articles on this and other topics. The different topics were divided after which the separate parts of the introduction to visual processing in ASD were written by one and rewritten by the other. This way the individual work of one was complementary to the other with respect to content and written language. The written parts were send to Ruth Van der Hallen, providing us with substantial feedback after which the content was further refined by both of us. The second phase of our research consisted of data collection on the ASD target group. Both of us conducted numerous series of test sessions (i.e. short version of the WISC-III and both computer tasks: Gaborized visual search and Multiple object tracking), spread out over several months. Data collection was supervised by Ruth Van der Hallen. The third phase of our research entailed the data-analysis. Together with Dr. Bart Boets and Ruth Van der Hallen we decided on the particular focus of the data-analysis, which they conducted for us. The data-analysis was interpreted and elaborated by both of us and further refined after feedback of Ruth Van der Hallen. The end result in the form of this research article was established as a duo, constantly refining the content of this manuscript and with the indispensible feedback of Ruth Van der Hallen. We chose to write a master thesis in the form of an English article to expose ourselves to the challenge and in order to be able to present it to scientific journals such as Journal of Autism and Developmental Disorders or Journal of Child Psychology and Psychiatry. Table of Content Introduction to Visual Processing in ASD Autism Spectrum Disorder Visual Processing: Two Major Frameworks Research Goal and Research Questions Method Participants Wechsler Intelligence Scale Child Behavior Checklist Social Responsiveness Scale Materials Gaborized Visual Search Multiple Object Tracking Procedure Gaborized Visual Search Multiple Object Tracking Data Analyses and Results Gaborized Visual Search Gabor Task: Accuracy Gabor Task: Reaction Times Multiple Object Tracking Mixed Model Analysis Degree of Global Interference Discussion Gaborized Visual Search Multiple Object Tracking Conclusions and Recommendations References Appendices Appendix A. Results Data-Analyses: Gaborized Visual Search Appendix B. Results Data-Analyses: MOT Introduction to Visual Processing in ASD Autism Spectrum Disorder Autism Spectrum Disorder (ASD) refers to an early onset neurodevelopmental disorder characterized by deficits in social communication and interaction and restricted and repetitive behaviors, interests or activities (American Psychiatric Association, 2013). Symptoms in both domains have to be present from early childhood on, even if these symptoms are only recognized at later age. Whereas the previous ASD classification of ‘pervasive developmental disorders’ (PDD) by the former DSM-IV-TR distinguished four ASD sub-types, i.e. Autistic Disorder, Asperger’s Syndrome, Childhood Disintegrative Disorder, and Pervasive Developmental Disorder Not Otherwise Specified (PDD-NOS), no such subdivision is present in the current DSM-5 (DSM-IV-TR; American Psychiatric Association, 2000; American Psychiatric Association, 2013). The DSM-IV-TR’s categorical approach of the autism spectrum has been abandoned and is remodeled into a more generic, all-encompassing dimensional approach of ASD. Both core domains are conceptualized as continua, where symptom severity can vary from mild to severe. This way, individual variation at the level of symptomatology is taken into account. The renewed dimensional approach is suggested to be more equipped to grasp and encompass the heterogeneity present across the ASD-spectrum and to improve diagnostics, in particular in case of more complex psychopathology (APA, 2012 and 2013). Prevalence rates of the autism spectrum are estimated between 1/88 and 1/100, where boys outnumber girls by approximately 5:1. The precise gender ratio differs with the degree of mental disability, i.e. higher in individuals with low functioning ASD and lower in persons with high functioning ASD (Verhulst F. C., 2008; APA, 2013; Fombonne, 2012). Prevalence rates of autism spectrum disorder have risen, since the first mentioning in the early 1940s (Kanner, 1943; Fombonne, 2012). It has been suggested that the rise in prevalence is partly due to early detection and recent broadening of the ASD diagnostic criteria. As a result, ASD-awareness has increased and diagnostic protocols have become more attuned to ASD-symptomatology. However, questions have been raised concerning the specificity and sensitivity of the ASD conceptualization in the renewed DSM-5. Some studies suggest poor sensitivity (Mattila, et al., 2011; Wing, Gould, & Gillberg, 2011) or suggest room for improvement on specificity (Huerta, Bishop, Duncan, Hus, & Lord, 2012). However, other studies suggest that most children with former ASD subtype diagnoses would indeed continue to qualify for an ASD diagnoses when applying DSM-5 criteria, indicating good levels of sensitivity as well as specificity of the proposed criteria across age and ability level (APA, 2012; Kent, et al., 2013). Visual Processing: Two Major Frameworks The study of visual processing plays an important role in trying to define and understand the ASD pathology. Although atypical visual processing is not one of the core diagnostic criteria of ASD, the current DSM-5 discusses atypical sensory processing, in particular in the visual domain. Two non-social, cognitive frameworks have been dominating the discussion on atypical visual processing in ASD, namely the Weak Central Coherence (WCC) theory and the Enhanced Perceptual Functioning (EPF) hypothesis. Although both frameworks differ in a number of ways, both aim at grasping the how’s and what’s of atypical perceptual processing in ASD. The WCC framework has launched the idea that individuals with ASD suffer from weak central coherence or, in other words, experience an inability to integrate elements of information into coherent wholes (Frith, 1989; Happé & Frith, 2006). The concept of ‘coherent wholes’ originates from early Gestalt principles, that postulate the idea that objects and organized patterns are automatically perceived as a whole and such perception precedes the awareness of the parts (Koffka, 1935 and Wertheimer, 1928 in Gray, 2007). According to Gestalt psychologists, a whole affects the perception of the parts through unconscious inference. This process involves primarily top-down control, rather than bottom-up processing (Happé & Booth, 2008). With the WCC theory, Frith (1989) first advocated weak central coherence in ASD to stem from a core deficit in central processing and a failure to extract global form and meaning. This hypothesis, however, has been reformulated and WCC now claims that differences between individuals with and without ASD result from an atypical, detailed-focused cognitive processing style in ASD. This automatic processing style could be overcome in situations with explicit demands for global processing and is seen as one aspect of a cognitive profile alongside deficits in social cognition. Within this ‘detailed-focused processing style’, the WCC theory describes two separate dimensions or continua: a reduced tendency to integrate information and an increased tendency to feature processing (Happé & Booth, 2008). The EPF hypothesis is used as a framework to study superior low-level perceptual performance in ASD, without making assumptions about the ability to process global visual information. EPF suggests individuals with ASD: 1) to display enhanced low (e.g. discrimination) and mid-level (e.g. pattern detection) cognitive processing of information, 2) have a default locally oriented processing style, and 3) present greater activation of primary perceptual brain areas, as well as relative autonomy of perceptual processes from top-down influences (Mottron & Burack, 2001; Mottron, Dawson, Souli, Hubert, & Burack, 2006; Mottron, Dawson, & Soulières, 2009; Samson, Mottron, Soulières, & Zeffiro, 2012; Mottron, et al., 2013). Mottron et al. (2013) introduced veridical mapping as underlying perceptual mechanism of pattern detection for EPF, bridging ASD peak abilities and neural correlates involved in EPF. By including neuronal perspectives on cognitive differences between ASD and typically developing children (TD), and the underlying perceptual mechanism by means of veridical mapping, the EPF hypothesis has a more mandatory basis than the more “optional cognitive style” proposed by Happé & Frith (2006) and the WCC theory (Haesen, 2008; Simmons, Robertson, McKay, Toal, McAleer, & Pollick, 2009). Research Goal and Research Questions The research on visual information processing in ASD conducted so far has provided us with a broad range of information and insights. Yet, much of the conducted research has led to mixed results and leaves us with unresolved questions in the domain of global versus detailed-focused information processing in ASD. Questions on the nature of peak performances of ASD compared to TD and the heterogeneity within the spectrum on task performances in the domain of visual processing remain largely unanswered. While individuals with ASD or their relatives often report visual atypicalities in natural situations, it remains difficult to elucidate those symptoms in experimental settings. One element that might attribute to the mixed evidence in research on visual processing in ASD is the fact that researchers do not take into account to what extent, either implicitly or explicitly, participants are asked to process visual information. Real-life situations require implicit and automatic processing, rather than explicit processing; when driving a car, one needs to be able to evaluate traffic situations on the spot and as an integrated whole in order to react adequately. If there is less or slower integration of the different elements in traffic, one is not able to respond as adequately and ends up at risk for car accidents or traffic violations. While in real-life situations one is usually not explicitly informed as to how one should process information best, experimental situations often make this explicit, which might help individuals with ASD to adopt the right processing style but masks any evidence on possible habitual differences. One study that investigated the impact of spontaneous versus instructed processing is a study by Koldewyn, Jiang, Weigelt, & Kanwisher (2013). Children with ASD were less likely to report global information compared to typically developing children, when given the choice to report either one. When given an explicit task instruction, their ability to process global information seems unimpaired. With this research, Koldewyn et al. (2013) indicated that children with ASD have intact global information processing abilities, yet are less inclined to attend to and report global information. The current study aims at clarifying the existing ambiguity in current ASD research by investigating the nature of atypical local-global visual processing in individuals with ASD. More specifically, we investigated the possible influence of implicit versus explicit task characteristics and instruction on local versus global visual processing in ASD and typically developing.We administered a Gaborized visual search task (Gabor) in which the implicit or explicit nature of the local versus global task instruction was directly manipulated, and a multiple object tracking (MOT) task in which participants were instructed to track local objects in the presence of distracting global structures. For the Gabor task, we hypothesize that children with ASD will show intact local processing in the explicit and implicit task conditions, but impaired global processing in the implicit task conditions. For the MOT task, we hypothesize that children with ASD show less grouping interference in object tracking compared to TD. Method Participants Two groups (n1=27, n2=26) of 9 to 13 year old children participated in our study. All participants had normal or corrected-to-normal vision and were native Dutch-speakers. Our first group of participants, (n1=27) typically developing children, was recruited at local Belgium schools (17 boys, 10 girls, Mage = 123 months, age range= 9-13 years). Our second group of participants, (n2=26) children with ASD, was recruited through the Autism register of the department of Child Psychiatry of the University Hospital Gasthuisberg (19 boys, 7 girls, Mage = 126,8 months, age range = 9-13 years). All participants were previously diagnosed with ASD by a child psychiatrist or a multidisciplinary team according to the criteria of the DSM-IV-TR (DSM-IV-TR; American Psychiatric Association, 2000). ASD diagnoses were re-evaluated within the research protocol, using the Dutch translation of the Autism Diagnostic Observation Schedule (ADOS; De Bildt, De Jonge, Lord, Rutter, DiLavore, & Risi, 2008). To evaluate level of intelligence, emotional, attentional and behavioral problems or social impairment in both participant groups, several scoring instruments were administered. Demographic details of both the ASD group and the TD group are shown in Table 1. Table 1. Descriptive Statistics after Matching. ASD (n = 26) TD (n = 27) (19 boys, 7 girls) (17 boys, 10 girls) Mean (SD) Range Mean (SD) Range p-value of t-test TIQ 101.1 (12.75) 76.5 – 130 104 (11.45) 80.5 – 128 .4209 VIQ 98 (15) 70 – 129 104 (13) 78 – 133 .1264 PIQ 03 (13) 83 – 129 104 (14) 77 – 129 .9475 Age (in 126.8 2920 – 4659 123 (14.3) 3044 – 4385 .4094 months) (19.1) Gender .4401 Wechsler Intelligence Scale Intellectual ability was assessed by the shortened version of the WISC (WISC-III-R), which encompasses subtests Vocabulary, Similarities, Block Design and Picture Completion. Although not a complete intelligence test, the shortened version provides reliable measures of full scale IQ (FSIQ), verbal IQ (VIQ) and performance IQ (PIQ). Raw scores were transformed into scaled scores and to IQ scores for full scale IQ, verbal and performance IQ. Both participant groups were matched on VIQ and PIQ. Individuals with a VIQ, PIQ, or FSIQ ≤ 70 were excluded from the analysis. Child Behavior Checklist To assess emotional and behavioral problems, parents completed the Child Behavior Checklist (CBCL; Achenbach, 2001). This questionnaire uses a dimensional approach that assesses emotional and behavioral disorders in children aged 6–18 years. Both emotional and behavioral disorders are known to co-occur at a higher rate or with greater severity in individuals with ASD (Achenbach, 2001; Verhulst & van den Ende, 2012; Gau, et al., 2010). A clinically significant range of concern includes behaviors that are moderately to very deviant in comparison to the scores from a normative peer sample, i.e. a T-score of 63 or higher on these scales (Achenbach & Ruffle, 2000; Skokauskas & Gallagher, 2012). Children with comorbid emotional and/or behavioral disorders were excluded from the analysis. Social Responsiveness Scale To assess ASD severity, parents completed the Dutch version of the Social Responsiveness Scale (SRS; Roeyers, Thys, Druart, De Schryver, & Schittekatte, 2011). This is a 65-item rating scale, measuring the severity of impairments typical for ASD occurring in natural social settings (Roeyers, et al., 2011). The SRS quantifies social impairment in 4 to 18 year old children across a wide range of severity, i.e. social awareness, social information processing, capacity for reciprocal social communication, social anxiety or avoidance and autistic preoccupations and traits. Overall SRS T-scores are categorized according to four classes: high degree of social responsiveness (<40); normal social responsiveness (40-60); mild to moderate social responsiveness (61-75) and serious shortcomings in social responsiveness (≥ 76). Children with high scores on the social responsiveness scale were excluded from the TD group. Materials Gaborized Visual Search Participants were presented with an array of Gabor patches, in which they had to search for a local or global target as fast as possible. Targets were presented on the right or on the left side of the screen and the number of left-right presentations was equally divided. A Gabor patch is an element defined by the combination of a sine wave luminance grating and a two-dimensional Gaussian window (for more extensive details, see Sassi, Vancleef, Machilsen, Panis, & Wagemans, 2010). Three types of target stimuli were presented: local targets, open global targets and closed global targets. All target stimuli were embedded in a background of Gabor distractors, displayed at two types of complexity and in the presence or absence of orientation noise (Figure 1). Local targets consisted of one single Gabor patch, which orientation slightly deviated from the surrounding distractor patches. The degree of deviation was manipulated to create two types of complexity: a) Simple (50°) or b) Complex (35°)1. Background distractor patches were presented with or without orientation noise (noise between -12° and 12°). Open global targets were constructed from multiple Gabor patches that formed an open contour. The shape of the open contour was manipulated to create two types of complexity: a) Simple (straight line) or b) Complex (curved: partial radial frequency pattern2), each with or without orientation noise (noise between -30° and 30°). Background distractor patches were presented in random orientation. Closed global targets comprised of multiple Gabor patches that formed a randomly shaped closed contour. The shape of the closed contour was manipulated to create two types of complexity: a) Simple radial frequency patterns or b) Complex radial frequency patterns, each with or without orientation noise (noise between -30° and 30°). Background distractor patches were presented in random orientation. 1 Local stimuli are not considered to be simple or complex structures since they consist out of one single Gabor patch. However, the manipulation in their presentation (i.e. 50° or 35°) makes the task easier or more difficult. 2 A Radial Frequency Pattern (RFP) is a sinusoidal modulation of the radius of a basic circle forming a closed contour. Each sinusoidal modulation has a certain frequency (cycles/pixel) and amplitude (Wilkinson, Wilson, & Habak, 1998). It allows a systematic, parametric variation of shape, in terms of the deviations from a circle. Figure 1. Illustration of several types of target stimuli, displaying different types of targets, their type of complexity and background manipulation in terms of orientation noise: a) Local, simple target (50°), b) Local, complex target (35°), c) Open, simple target (straight line), d) Closed, simple target (simple RFP), e) Open, complex target (curved: partial RFP) in the presence of background orientation noise, f) Closed, complex target (complex RFP) in the presence of background orientation noise. Stimuli were created with the Grouping Elements Rendering Toolbox (GERT; Demeyer & Machilsen, 2012). To measure the influence of the implicit or explicit nature of the task instruction on local-global visual information processing, the same task was presented first with an explicit and later with an implicit task instruction and its impact was evaluated on both local and global targets. Participants performed three blocks of trials in which they were explicitly instructed to find one of the three specific targets. In each block, only one type of target was presented. The order of presentation of these three blocks was randomized across participants. Next, participants were presented with a larger block, where all target types were presented intermixed in a randomized order and no explicit instruction was provided. In sum, this 2x3x2x2x2 research design shows the following independent variables: two groups (ASD versus TD), three target conditions (local targets, open global targets and closed global targets), two types of task instruction (explicit or implicit), background noise (present or absent) and two types of complexity (Local: 50° or 35° or, Open: straight or curved, Closed: simple RFP or complex RFP). Performance was measured in terms of accuracy as well as reaction time. Multiple Object Tracking In the multiple object tracking (MOT) task participants were presented with a number of moving local targets and were asked to track certain targets amongst moving distracters. With this task, we aimed to investigate the effect of connection-based grouping in children with and without ASD. The MOT paradigm applied in this study, was developed by Pylyshyn and Storm (1988) to study object-based visual attention. Although it has been established that visual attention can select visual objects (Scholl, Pylyshyn, & Feldman, 2001), what counts as objects remains unclear. In a classical MOT paradigm, subjects are asked to track a number of independently and unpredictably moving identical items in a field of identical distracters (Scholl, et al., 2001). In contrast to visual search tasks, where focus remains unclear until target onset, in MOT participants are explicitly instructed to direct their attention to some of the targets rather than all moving objects (Keane, Mettler, Tsoi, & Kellman, 2011). MOT appeals to a primary form of visual information processing in the sense that object tracking does not depend on featural properties of the tracked objects (Keane, et al., 2011). For our research design, we adapted the Boxes, Necker Cubes and Necker Controls conditions from Scholl, et al. (2001). We will refer to them as Ungrouped, Grouped, and Connected. The Ungrouped condition served as our baseline condition. Items were constructed as simple squares (Figure 2a.). In this condition the squares, both targets and distracters, could only be perceived as individually moving objects. In the Grouped condition one target box and one distracter box were connected, with each vertex of the target box connected to a corresponding vertex of the distracter box through four long thin lines. Despite being connected, target and distracter continued to move completely independent of one another. By connecting the vertexes of the boxes, target and distracter visually merged into a bar shape (Figure 2b.). The connected target and distracter box could therefore be perceived as one global form. Such task-irrelevant object formation is suggested to interfere with object tracking. The Connected condition was set up as a control condition to verify that any differences in performance in conditions a and b were not simply due to the fact that four lines were added to the visual display and therefore the visual display was more cluttered. It was similar to the Grouped condition in a way that one target box and one distracter box were connected through four lines, but differed in that instead of connecting lines to the vertexes of the boxes, lines were attached to the middles of each side of the boxes (Figure 2c.). This alteration maintained a similar amount of visual information, but did not result in a bar shaped visual grouping. Similar to the Grouped condition, the target and distracter boxes that were connected continued to move independent of one another. Figure 2. The images are displayed as an illustration of the three conditions and are not drawn to scale. Each drawing consists of eight items, four of which were targets and four of which were distracters. Targets were indicated by means of four seconds of flashing dollar signs. Procedure The Gabor task and the MOT task were individually administered in a soundproof and slightly darkened room. Participants were seated on an adjustable chair, so that height and distance to the computer screen was identical for every participant (distance to the screen: 57 cm). Prior to starting the practice series, a brief and standardized task instruction was given to each participant. In addition, they were requested to concentrate and focus their attention onto the screen throughout the entire task. After completing the three explicit instruction blocks of the Gabor task and the MOT task, we administered the ADOS and the WISC intelligence scale. Lastly, we administered the final trial block of the Gabor task where trial types were presented in random order (implicit search task). To keep participants motivated a token economy system was used in which participants could collect stickers when completing subtasks. After completing the entire protocol, all participants received a small gift as a token of gratitude for their participation. In addition, all were given the opportunity to take regular breaks at specific moments throughout the entire research protocol. Gaborized Visual Search The three explicit subtasks were explained through a standardized explicit verbal and visual instruction, indicating the type of target stimulus prior to each subtask, followed by a practice block and the test block. Each block comprised a series of trials in which the specific target stimulus had to be identified. Participants were shown a one second fixation cross in the middle of the screen after which the specific target stimulus appeared on one side of the screen (either left or right), surrounded by distractor patches. Participants indicated as quickly and accurately as possible on which side (left or right) of the screen the target was present, by pressing one of two response buttons. Participants received feedback on their responses after each trial (i.e. accurate response was followed by a green screen; inaccurate response was followed by a red screen). If participants did not respond within the 8s time limit, the item was scored as inaccurate and a new trial was automatically presented. Multiple Object Tracking The MOT task was presented as a treasure hunt in which participants had to indicate target stimuli hiding treasures in the form of dollar signs. All participants received the same standardized task instruction prior to the test trials, explaining the objective of the task, followed by two blocks of practice trials and the actual test block. Participants used the computer mouse to indicate targets on the computer screen. The first block of practice trials contained three trials (one per condition, i.e. Ungrouped, Grouped, and Connected). Each trial presented six objects, of which three were the actual targets and three were distracters. Participants had to indicate which three objects were actual targets. The second block of practice trials again contained three trials (one per condition). This time each trial presented eight objects, of which four were actual targets and four were distracters. Participants had to indicate which four objects were the actual targets. The test block comprised a total of sixty trials (20 Ungrouped, 20 Grouped, 20 Connected). Similar to the second block of practice trials, each trial consisted of eight items, of which four were actual targets and four were distracters. Participants had to indicate which four objects were the actual targets. As each trial started, the outline of the target squares (whether Ungrouped, Grouped, or Connected) lit up and a dollar sign appeared within the squares, to indicate that these squares were the actual targets. After four seconds the outlines turned back to their initial color and the dollar signs disappeared. From that moment onwards, all squares started moving randomly across the screen at an average speed of 2,8 degrees per second, following an independently determined paths (using Bézier curves to create smooth movements). After eight seconds all squares stopped moving and participants were asked to indicate which squares were the targets. Subjects received immediate feedback: A golden dollar sign appeared and a sound (‘k-ching’) was played when providing a correct answer. The indicated squares turned grey when the answer was incorrect. Data Analyses and Results Assumptions of normality and homogeneity were checked by means of a histogram, qq-plot and Shapiro-Wilk and Kolmogorov-Smirnov test. Analyses were conducted with a repeated-measures mixed model analysis using the general statistical software package SAS (Version 9.3; SAS Institute Inc., 2011). Significance tests were conducted with a significance level of 5%. Post-hoc tests were Tukey-corrected. For the Gabor task, performance was measured in terms of accuracy and reaction time. In order to meet the assumption of normality, reaction times were logarithmically transformed. For the MOT task, performance was measured in terms of accuracy. Although there was a small ceiling effect for the Ungrouped condition, MOT data were normally distributed. Gaborized Visual Search This 2x3x2x2x2x2 research design includes ‘Sample’, ‘Target Condition’, ‘Instruction’, ‘Noise’ and ‘Complexity’ as independent variables. The variable ‘Target Condition’ refers to the three types of target stimuli, namely local, open and closed target stimuli. The variable ‘Instruction’ refers to the two types of task instruction: an explicit task instruction or an implicit task instruction. ‘Noise’ refers to the orientation noise in the background or on the contour (present or absent). The variable ‘Complexity’ refers to the complexity of the target stimulus, either a simple target stimulus (Local: 50°, Open: straight line, Closed: simple RFP) or a complex target stimulus (Local: 35°, Open: curved line, Closed: complex RFP). Given the scope of the present master thesis and given that ‘Noise’ and ‘Complexity’ have been differently operationalized across the various target conditions, the two-way interactions with ‘Noise’ and ‘Complexity’ will not be discussed here. Figure 3 and Figure 4 display mean accuracy and mean reaction times for both groups in the three conditions. Figure 3. Mean accuracy on the Gabor task for the ASD and TD group in the three target conditions. Figure 4. Mean of log reaction time on the Gabor task for the ASD and TD group in the three target conditions. Gabor Task: Accuracy A repeated-measures mixed model analysis including ‘Instruction’, ‘Condition’, ‘Noise’ and ‘Complexity’ as within-subject variables, ‘Sample’ as between-subject variable and accuracy as dependent variable reveals a main effect of ‘Target Condition’ (F (2, 1196) = 37.21, p < .0001) ‘Noise’ (F (1, 1196) = 319.38, p < .0001) and ‘Complexity’ (F (1, 1196) = 625.49, p < .0001) (Figure A1 – A3), as well as a two-way interaction effect between ‘Sample’ and ‘Instruction’ (F (1, 1196) = 7.70, p = .0056), and between ‘Instruction’ and ‘Condition’ (F (2, 1196) = 24.30, p < .0001), (Figure A4 – A5). The main effect of ‘Target Condition’ reveals that, for both ASD and TD groups, accuracy scores depended on the type of target stimulus that was presented. Participants were most accurate in finding open targets (M = .94, SD = .10), accurate for closed targets (M = .91, SD = .12) and least accurate for local targets (M = .89, SD = .14). Post hoc analysis (Tukey-corrected) reveals significant differences in mean accuracy between each of the three conditions (all p < .001, Figure A1). The main effect of ‘Noise’ reveals that, for both participants groups, accuracy scores depended on the presence or absence of orientation noise on the background or contour Gabor elements (F (1, 1196) = 319.38, p < .0001). On average, targets were found more accurately in absence of noise (M = .96, SD = .08) than when orientation noise was present (M = .87, SD = .14) (Figure A2). The main effect of ‘Complexity’ reveals that, for both participant groups, accuracy scores depended on the complexity of the stimulus type that was presented (F (1, 1196) = 625.49, p < .0001). On average, targets were found more accurately for simple targets (M = .97, SD = .05) compared to complex targets (M = .85, SD = .15) (Figure A3). Although no main effect is observed for ‘Instruction’, an interesting two-way interaction is present between ‘Sample’ and ‘Instruction’ (F (1, 1196) = 7.70, p = .0056). Post-hoc analysis reveals that, whereas the ASD sample performed as accurate in the implicit condition as in the explicit condition (t (1196) = 1.05, p = .7182), the typically developing group was more accurate in the implicit condition compared to the explicit condition (t (1197) = -2.89, p = .0207, Figure A4). The two-way interaction between ‘Instruction’ and ‘Target Condition’ (F (2, 1196) = 24.30, p < .0001) reveals that for both groups, the strength of learning from the explicit condition to the implicit condition differed given the type of target stimulus. While a significant increase in accuracy was present for the closed targets (t (1196) = 2.88, p = .0471) as well as the local targets (t (1196) = -6.36, p < .0001), no such increase was present for the open targets (t (1197) = 1.26, p = .8059). Note, however, that accuracy in general, was highest for open targets so the possibility of improving from one condition to the next, was reduced (Figure A5). Gabor Task: Reaction Times A repeated-measures mixed model analysis including ‘Instruction’, ‘Task Condition’, ‘Noise’ and ‘Complexity’ as within-subject variables, ‘Sample’ as between-subject variable and reaction time as dependent variable reveals a main effect of ‘Instruction’ (F (2, 1196) = 82.65, p < .0001), ‘Task Condition’ (F (2, 1196) = 356.62, p < .0001), ‘Noise’ (F (1, 1196) = 344.52, p < .0001) and ‘Complexity’ (F (1, 1196) = 1169.42, p < .0001), as well as a two-way interaction effect between ‘Sample’ and ‘Instruction’ (F (1, 1196) = 11.81, p = .0006) and between ‘Instruction’ and ‘Task Condition’ (F (2, 1196) = 7.89, p = .0004) as displayed in Figure A6 – A11. The main effect of ‘Instruction’ reveals that, for both ASD and TD groups, reaction times depended on the type of task instruction that was given (F (2, 1196) = 82.65, p < .0001). Participants were quicker on the second task with implicit task instruction (M = 1181, SD = 404.92), compared to the first task with explicit task instruction (M = 1311, SD = 476.17) (Figure A6). However, this main effect should be interpreted in view of the significant interaction between ‘Sample’ and ‘Instruction’ (F (1, 1196) = 11.81, p = .0006), revealing that the effect is much stronger for the TD group (t (1196) = 8.93, p < .0001) than for the ASD group (t (1196) = 3.97, p = .0004, Figure A7). The main effect of ‘Task Condition’ reveals that, for both ASD and TD groups, reaction times depended on the type of stimulus that was presented (F (2, 1196) = 356.62, p < .0001). Participants were faster finding open targets (M = 1078, SD = 352.42), than closed targets (M = 1179, SD = 421.19) than local targets (M = 1478, SD = 459.87) (All post hoc contrasts significant with Tukey-correction, Figure A8). The main effect of ‘Noise’ reveals that, for both participants groups, reaction times depended on the presence or absence of orientation noise on the background Gabor elements (F (1, 1196) = 344.52, p < .0001). On average, targets were found faster in absence of noise (M = 1129, SD = 392.95) than when orientation noise was present (M = 1362, SD = 467.06) (Figure A9). The main effect of ‘Complexity’ reveals that, for both participant groups, reaction times depended on the complexity of the stimulus type that was presented (F (1, 1196) = 1169.42, p < .0001). On average, simple targets were found faster (M = 1023, SD = 271.06) compared to complex targets (M = 1467, SD = 477.12) (Figure A10). The two-way interaction between ‘Instruction’ and ‘Task Condition’ (F (2, 1196) = 7.89, p = .0004) reveals that for both groups, in the speed of tracing a target stimuli depended on the particular combination of the type of target stimulus and task instruction. Although a decline in reaction time from the explicit condition to the implicit condition was present for both groups and all three targets, the strength of decline significantly differed: present for open targets (t (1197) = 3.48, p = .0069) and closed targets (t (1196) = 3.78, p = .0023), but especially apparent for local targets (t (1196) = 8.50, p < .0001, Figure A12). Multiple Object Tracking There are three conditions (Ungrouped, Grouped and Connected), in which participants (ASD and TD) could obtain a score from 0 to 4. Mean accuracy for each of the conditions is displayed in figure 5. Figure 5. Mean of accuracy on the MOT task for the ASD and TD group in the three conditions. Mixed Model Analysis A repeated-measures mixed model analysis with accuracy as dependent variable, ‘Sample’ as between-subject variable and ‘Condition’ as within-subject variable, reveals a main effect of Condition (F (2,102) = 84.72, p < .0001), no significant main effect of ‘Sample’ (F (1,51) = 1.98, p = .1656) and no significant ‘Sample’ by ‘Condition’ interaction (F (2,102) = .88, p = .42). The main effect of ‘Condition’ reveals that, for both ASD and TD groups, the mean accuracy differed significantly between conditions. Participants were most accurate in the Ungrouped trials (M = 3.68, SD = .07), followed by Connected (M = 3.04, SD = .07) and Grouped trials (M = 2.95, SD = .07). Post hoc analysis (Tukey-corrected) reveals that mean accuracy in the Ungrouped condition was significantly higher compared to both other conditions, i.e. Connected versus Ungrouped (t (102) = -10.45, p < .0001) and Grouped versus Ungrouped (t (102) = -11.94, p < .0001). Analysis reveals no significant difference in mean accuracy on the Grouped versus Connected condition (t (102) = -1.49, p = .30, Figure B1). The absence of a main effect for ‘Sample’ or interaction effect between ‘Sample’ and ‘Condition’ indicates that there is no difference in performance between the ASD group and the TD group on any of the conditions. Both groups performed equally accurate on the MOT task. Degree of Global Interference For both groups, the degree of global interference was calculated, by subtracting their average performance on their global trials (average of Grouped trials and Connected trials) from their baseline performance on the Ungrouped, local trials. The differences in the degree of global interference between ASD and TD, was not significant (t (51) = -.51, p = .61). This indicates that de degree of global interference experienced by the ASD group (M = .72, SD = .47) was similar to the degree of global interference experienced by the TD group (M = .65, SD = .44) and global interference did not lead to significant differences in accuracy scores between both groups. Discussion In the current study visual processing in children with ASD was investigated and compared to a typically developing group using a Gaborized visual search task and Multiple object tracking task. Participant groups were matched for age, gender and FSIQ. For the Gabor task, we hypothesized children with ASD to show intact local processing in the explicit and implicit task conditions, but impaired global processing in the implicit task conditions. For the MOT task, we hypothesized children with ASD to experience less global interference on object tracking. Gaborized Visual Search A first interesting result revealed in the Gabor task is the differential effect of task instruction for ASD compared to TD; while both groups performed equally well when given an explicit task instruction, the TD group outperformed the ASD group when given an implicit task instruction. This differential effect of task instruction might be interpreted in terms of a differential learning effect. As the explicit task condition was always presented before the implicit task condition, the presence of a learning effect from one condition to the next is likely. While such a learning effect seems present for the TD group, data suggest no such learning effect for the ASD group. Whereas the TD group becomes faster and more accurate in the implicit condition compared to explicit condition, the ASD group’s performance is similar to their performance in the explicit condition. This might raise questions pertaining to the learning capacities of children with ASD. However, several previous studies found learning to be intact in ASD (e.g. Brown, Aczel, Jiménez, Kaufmann, & Plaisted Grant, 2010; Nemeth et al., 2010). Therefore, it is valid to assume that children with ASD do not suffer from a general learning deficit and the lack of improvement is due to the implicit nature of the second condition. Accordingly, ASD participants did learn, but the results reflect an impact of the random presentation of the targets. The random presentation made it more difficult to deduct what target had to be found. This could have possibly hampered the performance of the ASD group and suppressed the learning effect. Impairments of behavioral and cognitive flexibility might have hindered the ASD group to quickly deduct the correct type of target and respond adequately to the randomized presentation. Behavioral and cognitive inflexibility is manifested by an insistance of sameness and resistance of change, making it more difficult to adapt and respond flexibly to new situations (Green, et al., 2007). In the current study, this relates to a situation in which task instruction is given implicitly and flexibility is required to adapt to a task in which target stimuli are presented randomely. Lastly, the fact that both participants groups performed equally under explicit task conditions strengthens the idea of the WCC theory that children with ASD show intact (local-global) processing when they are explicitly instructed to do so (Happé & Booth, 2008; Koldewyn et al., 2013). A second interesting and surprising result revealed in the Gabor task, is the main effect of target type and the absence of a group by target type interaction. We hypothesized children with ASD to show a more detail focused processing style, and therefore perform better when presented with local targets, compared to the typically developing. This hypothesis is in line with both EPF theory’s default setting of a locally oriented processing style and WCC’s detailed-focused processing style (Happé & Booth, 2008; Mottron & Burack, 2001; Mottron et al., 2006; Mottron et al., 2009; Mottron et al., 2013; Samson et al., 2012) and would suggest superiority in tracing local stimuli in both the explicit and implicit conditions for children with ASD. Results in the current research, however, do not reveal enhanced perfomance in ASD for local targets. Both the ASD and typically developing group were most accurate and fastest in tracing global targets compared to local targets, and performance for local trials did not differ between groups. A possible reason as to why the ASD group failed to show enhanced performance for local targets may be the imbalance of local versus global trials in our designs (1:2 ratio). This might have induced an object-based allocation for both the ASD and TD group, possibly diminishing the detailed focused processing style in the ASD group that was expected in the current research and postulated by WCC theory (Happé & Booth, 2008). Multiple Object Tracking The MOT task revealed similar performance for both participant groups in each of the three conditions, indicating that the performance was equally affected by object formation. While we expected children with ASD to experience less global interference, this was not revealed by our results and a default locally oriented processing style in ASD could not be confirmed. Both participant groups performed best in the Local condition compared to both Global conditions. In addition, both groups showed similar amounts of global interference when tracing object in Global conditions. Our findings contrast with the findings of Scholl et al., (2001), where the visual clutter that was added in the Connected trials caused less interference compared to the Grouped trials in typically developing adults. Whereas O’Hearn et al. (2013) found no differential effect of grouping interference on tracking ability, Evers et al. (2014) did show the ASD sample to be less susceptible to grouping interference. This last study is most similar to the current study, but applied a 1:1 ratio of grouped versus ungrouped trials rather than a 2:1 ratio. The 2:1 ratio of grouped versus ungrouped trials as applied here, may have induced an unwanted focus on global targets or object formation, or at least, allowed for more learning to occur for global trials compared to local trials. Interesting to note is that the participants in the current study were as old as the children in that of O’Hearn et al., (2013), but older than the children that participated in the Evers et al. (2014) study (8-14 years versus 6-10 years old). Although it is difficult to draw any conclusion with regard to developmental differences based on these three studies, this might also suggest that a subtle developmental effect is in place with regard to global interference on object tracking. Whereas O’Hearn et al., (2013) revealed that children aged 9-to-12years old show a similar amount of grouping interference as adolescents and adults, other studies suggest divergent developmental trajectories from adolescence on, with individuals with ASD failing to acquire global shape perceptions (Scherf, Luna, Kimchi, Minshew, & Behrmann, 2008). It has also been suggested that element grouping develops at a relative early age, whereas shape formation develops into late adolescence (Kimchi, Hadad, Behrmann, & Palmer, 2005; Scherf et al., 2008). Therefore, younger children might experience less global interference in tracking targets when targets are grouped with distracters. Conclusions and Recommendations In sum, the Gabor task and the MOT task revealed that children with ASD show normal local and global visual processing abilities. In addition, the Gabor task revealed intact visual processing for ASD when targets were presented in separate blocks (explicit task instruction), but diminished performance compared to TD with random presentation (implicit task instruction). One suggestion for future research is to look at this differential effect of task instruction in more detail. In future research it might be interesting to record performance on the Gabor task with different presentation orders, to disentangle the impact of learning effects from impact of task instruction. Our findings provide an interesting perspective on research on visual information processing conducted so far. Most of these studies only used explicit task instructions and often find no differences in performance between children with and without ASD. This study suggests that it is imperative to differentiate between explicit and implicit visual processing when trying to understand differences in perceptual organization between children with and without ASD. Concerning future research on Multiple object tracking it might be interesting to further eleborate the influence of grouped versus ungrouped ratios on the level of global interference. This way building on the current research findings and further exploring to what extent specific Gestalt grouping principles (e.g. common motion based grouping, connection-based grouping) influence tracking performance between children with and without ASD (Evers et al., 2014; O'Hearn et al., 2013). Additionally, it can be interesting to take different developmental stages further into account. This way broadening the perspective on possible developmental differences in visual processing in children with ASD throughout the lifespan. References Achenbach. (2001). Child Behavior Checklist (CBCL). Retrieved 05 14, 2013, from ASEBA: www.aseba.org Achenbach, T. M., & Ruffle, T. M. (2000). The Child Behavior Checklist and related forms for assessing behavioral/ emotional problems and competences. Pediatrics in review, Vol. 21, No. 8 , 265-271. American Psychiatric Association. (2013). American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition. Arlington, VA. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders. Washington, DC: Author. APA. (2012 йил 10-April). Autism Spectrum Disorder fact sheet. Retrieved 2013 йил April from DSM-5 Facts: http://dsmfacts.org/materials/autism-spectrum-disorder-fact-sheet/ APA. (2013 йил 8-March). Fact sheet: Autistic spectrum disorders. From American Psychological Association: http://www.apadivisions.org/division-54/evidence-based/autistic-spectrum-disorders.aspx?__utma=12968039.1508978625.1362644205.1362653943.1362733634.3&__ utmb=12968039.3.10.1362733634&__utmc=12968039&__utmx=-&__utmz=12968039.1362733634.3.3.utmcsr=google|utm Brown, J., Aczel, B., Jiménez, L., Kaufmann, S. B., & Plaisted Grant, K. (2010). Intact implicit learning in autism spectrum conditions. The Quarterly Journal of Experimental Psychology, 63: 9 , 1789-1812. De Bildt, A., De Jonge, M., Lord, C., Rutter, M., DiLavore, P., & Risi, S. (2008). Autisme Diagnostisch Observatie Schema (ADOS). Amsterdam: Hogrefe Uitgevers BV. Demeyer, M., & Machilsen, B. (2012). The construction of perceptual grouping displays using GERT. Behavioral Research Methods, 44(2) , 439-446. Evers, K., de-Wit, L., Van der Hallen, R., Haessen, B., Steyaert, J., Noens, I., et al. (2014). Brief report: Reduced grouping interference in children with ASD: Evidence from a Multiple objec tracking task. Journal of Autism and Developmental Disorders . Evers, K., Panis, S., Torfs, K., Steyaert, J., Noens, I., & Wagemans, J. (2014). Disturbed interplay between mid- and high-level vision in ASD? Evidence from a contour identification task with everyday objects. Journal of Developmental Disorders, 44 , 801-815. Fombonne, E. (2012, July). Encyclopedia on early childhood development [online]: Autism - Incidence of autism - Epidemiology of autism. Retrieved February 2014, from child-encyclopedia.com: http://www.child-encyclopedia.com/pages/PDF/FombonneANGxp1.pdf Frith, U. (1989). Autism: Explaining the enigma. Blackwell Publishing. Gau, S. S.-F., Chou, M.-C., Lee, J.-C., Wong, C.-C., Chou, W.-J., Chen, M.-F., et al. (2010). Behavioral problems and parenting style among Taiwanese children with autism and their siblings. Psychiatry and Clinical Neurosciences, Vol.64, No.1 , 70-78. Gray, P. (2007). The Psychology of Vision. In P. Gray, Psychology (pp. 263-301). New York: Worth Publishers. Green, V. A., Sigafoos, J., O'Reilly, M., Pituch, K. A., Didden, R., Lancioni, G. E., et al. (2007). Behavioral flexibility in individuals with autism: Theory, assassment, and intervention. In L. B. Zhao, Autism research advances (pp. 63-77). New York: Nova Science Publishers, Inc. Haesen, B. (2008 йил 1-October). Local and global visual processing in Autism Spectrum Disorders. Retrieved 2013 йил 15-February from http://www.gestaltrevision.be: http://www.gestaltrevision.be/en/what-we-do/individual-projects/51-phds/171-local-and-global-visual-processing-in-autism-spectrum-disorders Happé, F. G., & Booth, R. D. (2008). The power of the positive: Revisiting weak coherence in autism spectrum disorders. The Quarterly Journal of Experimental Psychology, Vol. 61, No. 1 , 50-63. Happé, F. G., & Frith, U. (2006). The weak coherence account: Detail-focused cognitive style in autism spectrum disorders. Journal of Autism and Developmental Disorders, Vol. 36, No. 1 , 5-25. Huerta, M., Bishop, S. L., Duncan, A., Hus, V., & Lord, C. (2012). Application of DSM-5 criteria for autism spectrum disorder to three samples of children with DSM-IV diagnoses of Pervasive Developmental Disorders. American Journal of Psychiatry, Vol. 169, No. 10 , 1056-1064. Kanner, L. (1943). Autistic disturbances of affective contact. Nervous Child 2 , 217-250. Keane, B. P., Mettler, E., Tsoi, V., & Kellman, P. J. (2011). Attentional signatures of perception: Multiple object tracking reveals the automaticity of contour interpolation. Journal of Experimental Psychology: Human Perception and Performance, Vol. 37, No. 3 , 685-698. Kent, R. G., Carrington, S. J., Le Couteur, A., Gould, J., Wing, L., Maljaars, J., et al. (2013). Diagnosing autism spectrum disorder: Who will get a DSM-5 diagnosis? Journal of Child Psychology and Psychiatry , 1-9. Kimchi, R. (1992). Primacy of wholistic processing and global/local paradigm: a critical review. Psychological Bulletin, Vol. 112, No. 1 , 24-38. Kimchi, R., Hadad, B., Behrmann, M., & Palmer, S. E. (2005). Microgenesis and ontogenesis of perceptual organization: Evidence from global and local processing of hierarchical patterns. Psychological Science , 282-290. Koldewyn, K., Jiang, Y. V., Weigelt, S., & Kanwisher, N. (2013). Global/local processing in autism: Not a disability, but a disinclination . Journal of Autism and Developmental Disorders 43 , 2329-2340. Mattila, M.-L., Kielinen, M., Linna, S.-L., Jussila, K., Ebeling, H., Bloigu, R., et al. (2011). Autism spectrum disorders according to DSM-IV-TR and comparison with DSM-5 draft criteria: An epidemiological study. Journal of the American Academy of Child & Adolescent Psychiatry, Volume 50, Issue 6 , 540-542. Mottron, L., & Burack, J. A. (2001). Enhanced perceptual functioning in the development of autism. In J. A. Burack, N. Charman, N. Yirmiya, & P. R. Zelazo, The development of autism: Perspectives from theory and research (pp. 131-148). Mahwah, NJ: Lawrence Erlbaum. Mottron, L., Bouvet, L., Bonnel, A., Samson, F., Burack, J. A., Dawson, M., et al. (2013). Veridical mapping in the development of exceptional autistic abilities. Neuroscience and Behavioral Reviews 37 , 209-228. Mottron, L., Dawson, M., & Soulières, I. (2009). Enhanced perception in savant syndrome: patterns, structure and creativity. Philosophical transactions of the Royal Society of London , B, 364 , 1385-1391. Mottron, L., Dawson, M., Souli, I., Hubert, B., & Burack, J. (2006). Enhanced perceptual functioning in autism: An update, and eight principles of autistic perception. Journal of Autism and Developmental Disorders, Vol. 36, No. 1 , 27-43. Nemeth, D., Janacsek, K., Balogh, V., Zsuzsa, L., Mingesz, R., Fazekaz, M., et al. (2010). Learning in Autism: Implicitly Superb. PLoS ONE 5(7) , e11731. O'Hearn, K., Franconeri, S., Wright, C., Minshew, N., & Luna, B. (2013). The development of individuation in autism. Journal of Experimental Psychology: Human Perception and Perfromance, 39(2) , 494-509. Pylyshyn, Z. W., & Storm, R. W. (1988). Tracking multiple independent targets: Evidence for a parallel tracking mechanism. Spatial Vision, 3 , 179-197. Roeyers, H., Thys, M., Druart, C., De Schryver, M., & Schittekatte, M. (2011). SRS Screeningslijst voor autismespectrum stoornissen. SRS . Amsterdam: Hogrefe Uitgevers BV. Samson, F., Mottron, L., Soulières, I., & Zeffiro, T. A. (2012). Enhanced visual functioning in autism: An ALE meta-analysis. Human Brain Mapping, 33 , 1553–1581. Sassi, M., Vancleef, K., Machilsen, B., Panis, S., & Wagemans, J. (2010). Identification of everyday objects on the basis of Gaborized outline versions. i-Perception, Vol. 1 , 121-142. Scherf, K. S., Luna, B., Kimchi, R., Minshew, N., & Behrmann, M. (2008). Missing the big picture: Impaired development of global shape processing in ASD. Autism Research , 114-129. Scholl, B. J., Pylyshyn, Z. W., & Feldman, J. (2001). What is a visual object? Evidence from target merging in Multiple Object Tracking. Cognition, 80 , 159-177. Simmons, D. R., Robertson, A. E., McKay, L. S., Toal, E., McAleer, P., & Pollick, F. E. (2009). Vision in autism spectrum disorders. Vision Research, 49 , 2705-2739. Skokauskas, N., & Gallagher, L. (2012). Mental health aspects of autistic spectrum disorders in children. Journal of Intellectual Disability Research, Vol.56, No.3 , 248-257. Verhulst, F. C. (2008). Leerboek Kinder- en Jeugdpsychiatrie . Assen: Van Gorcum. Verhulst, F., & van den Ende, J. (2012). Achenbach System of Emperically Based Assessment (ASEBA). Retrieved 05 14, 2013, from aseba.nl: http://www.aseba.nl Version 9.3; SAS Institute Inc. (2011). In Base SAS® 9.3 Procedures Guide. Cary, NC, USA: SAS Institute Inc. Wechsler, D. (1991). Wechsler Intelligens Scale for Children (3rd ed.). San Antonio, TX: The Psychological Corporation. Wilkinson, F., Wilson, H. R., & Habak, C. (1998). Detection and recognition of radial frequency patterns. Vision Research, 38 , 3555-3568. Wing, L., Gould, J., & Gillberg, C. (2011). Autism spectrum disorders in the DSM-V: Better or worse than the DSM-IV? Research in Developmetal Disabilities 32 , 768-773. Appendices Appendix A. Results Data-Analyses: Gaborized Visual Search Figure A1. Main effect of ‘Condition’ on mean accuracy in a model which included ‘Instruction’, ‘Condition’, ‘Noise’ and ‘Complexity’ as within-subject variables and ‘Sample’ as between-subject variable. Note. * p < .001, ** p < .0001. Error bars indicate +/- 1 SD of the mean. Figure A2. Main effect of ‘Noise’ on mean accuracy in a model which included ‘Instruction’, ‘Condition’, ‘Noise’ and ‘Complexity’ as within-subject variables and ‘Sample’ as between-subject variable. Note. ** p < .0001. Error bars indicate +/- 1 SD of the mean. Figure A3. Main effect of ‘Complexity’ on mean accuracy in a model which included ‘Instruction’, ‘Condition’, ‘Noise’ and ‘Complexity’ as within-subject variables and ‘Sample’ as between-subject variable. Note. ** p < .0001. Error bars indicate +/- 1 SD of the mean. Figure A4. Interaction effect of ‘Sample’ and ‘Instruction’ on mean accuracy in a model which included ‘Instruction’, ‘Condition’, ‘Noise’ and ‘Complexity’ as within-subject variables and ‘Sample’ as between-subject variable. Note. ** p < .0001. Error bars indicate +/- 1 SD of the mean. Figure A5. Interaction effect of ‘Instruction’ and ‘Condition’ on mean accuracy in a model which included ‘Instruction’, ‘Condition’, ‘Noise’ and ‘Complexity’ as within-subject variables and ‘Sample’ as between-subject variable. Note. ** p < .0001. Error bars indicate +/- 1 SD of the mean. Figure A6. Main effect of ‘Instruction’ on mean log reaction time in a model which included ‘Instruction’, ‘Condition’, ‘Noise’ and ‘Complexity’ as within-subject variables and ‘Sample’ as between-subject variable. Note. **p < .0001. Error bars indicate +/- 1 SD of the mean. Figure A7. Interaction effect of ‘Sample’ and ‘Instruction’ on mean log reaction time in a model which included ‘Instruction’, ‘Condition’, ‘Noise’ and ‘Complexity’ as within-subject variables and ‘Sample’ as between-subject variable. Note. **p < .0001. Error bars indicate +/- 1 SD of the mean. Figure A8. Main effect of ‘Condition’ per group on mean log reaction time in a model which included ‘Instruction’, ‘Condition’, ‘Noise’ and ‘Complexity’ as within-subject variables and ‘Sample’ as between-subject variable. Note. **p < .0001. Error bars indicate +/- 1 SD of the mean. Figure A9. Main effect of ‘Noise’ per group on mean log reaction time in a model which included ‘Instruction’, ‘Condition’, ‘Noise’ and ‘Complexity’ as within-subject variables and ‘Sample’ as between-subject variable. Note. **p < .0001. . Error bars indicate +/- 1 SD of the mean. Figure A10. Main effect of ‘Complexity’ on mean log reaction time in a model which included ‘Instruction’, ‘Condition’, ‘Noise’ and ‘Complexity’ as within-subject variables and ‘Sample’ as between-subject variable. Note. **p < .0001. Error bars indicate +/- 1 SD of the mean. Figure A12. Interaction effect of ‘Condition’ and ‘Instruction’ on mean log reaction time in a model which included ‘Instruction’, ‘Condition’, ‘Noise’ and ‘Complexity’ as within-subject variables and ‘Sample’ as between-subject variable. Note. **p < .0001. Error bars indicate +/- 1 SD of the mean. Appendix B. Results Data-Analyses: MOT Figure B1. Main effect of ‘Condition’ on mean accuracy in a model that included ‘Sample’ as between-subject variable and ‘Condition’ as within-subject variable. Note.**p < .0001. Error bars indicate +/- 1 SD of the mean. © 2014 – Laboratorium voor Experimentele Psychologie KU Leuven. All rights reserved. No part of the data and/or other content of this thesis may be reproduced or distributed by any means without explicit written permission by the Laboratorium voor Experimentele Psychologie KU Leuven. This thesis is part of a broader research project en therefore may not be considered as the primary publication.