Drawing links between the autism cognitive profile and imagination: Executive function and processing bias in imaginative drawings by children with and without autism

Abstract

Little is known about the relation between cognitive processes and imagination and whether this relation differs between neurotypically developing children and children with autism. To address this issue, we administered a cognitive task battery and Karmiloff-Smith’s drawing task, which requires children to draw imaginative people and houses. For children with autism, executive function significantly predicted imaginative drawing. In neurotypically developing controls, executive function and cognitive-perceptual processing style predicted imaginative drawing, but these associations were moderated by mental age. In younger (neurotypically developing) children, better executive function and a local processing bias were associated with imagination; in older children, only a global bias was associated with imagination. These findings suggest that (a) with development there are changes in the type of cognitive processes involved in imagination and (b) children with autism employ a unique cognitive strategy in imaginative drawing.

Keywords

autism spectrum disorder cognition drawing executive function imagination local processing bias

Introduction

Imagination is a uniquely human ability that allows us to transcend space and time (Carlson et al., 2013). It is manifest in a variety of symbolic activities that create new ideas and works, such as pretend play, mental imagery, drawing, and writing. Following Vygotsky (2004), imagination is a combinatorial activity that involves the reworking and synthesis of elements of past experience into novel ideas. It is the combinatorial aspect of imagination that we examined in this study.

A number of cognitive mechanisms have been suggested to be involved in imagination (see Singer and Singer, 2013), including executive functioning (higher-order thought processes; Carlson et al., 2013) and local processing bias (LPB; a bias for details rather than unified gestalts; Scott, 2013). One approach to study the relation between cognition and imagination is to turn to clinical samples that are characterized by both imagination and cognitive deficits, such as children with autism spectrum disorder (ASD; Jarrold et al., 1996; Scott, 2013). The goal of this study is to clarify the relations between imagination and cognition in neurotypically developing (NT) children and children with autism. Importantly, we are among the first to address systematically the link between cognition and imaginative drawing in children on the autism spectrum.

Imagination in children with autism

Children with ASD frequently display difficulties in imaginative activities. They are less likely to produce spontaneous pretend play (e.g. Jarrold et al., 1993), although performance is enhanced by external prompts and instruction (e.g. Charman and Baron-Cohen, 1997). In spontaneous drawings, children with ASD exhibit a restricted range of idiosyncratic ideas and topics in their spontaneous drawings (e.g. Burton, 2010). In a seminal study by Scott and Baron-Cohen (1996), 15 children with ASD (mean age, 13;0), 14 children with cognitive delay (mean age, 12;8), and 15 NT children (mean age, 4;10) were administered Karmiloff-Smith’s (1990) “draw an impossible person” task. In this task, children are instructed to draw a person or house that “does not exist, an impossible person (or house).” Children with ASD performed significantly worse than matched controls; 92% of children with ASD drew realistically looking, rather than impossible, people. This effect was replicated with different samples of children with ASD (Low et al., 2009) and higher-functioning children with ASD (Craig et al., 2001).

The ASD imagination deficit is still not well understood. Scott and Baron-Cohen (1996) and Craig and Baron-Cohen (1999) have suggested that children with ASD lack the ability to form or hold non-veridical representations. However, children with ASD are able to copy impossible geometric figures (Sheppard et al., 2009) and are successful at imaginative drawing tasks when given prompts and support (Leevers and Harris, 1998). This has led others to examine difficulties in executive function (EF) and visual-perceptual style as the root of the imagination deficit in children with ASD (e.g. Low et al., 2009).

In this study, we address three cognitive explanations of the imagination deficit in children with ASD: EF (Hill, 2004), generativity (Low et al., 2009), and the LPB (Mottron et al., 2006). Few studies have been conducted on the relation between cognitive processes and imaginative drawing in children with ASD.

Imagination and EF

According to executive dysfunction theory, impaired EF underlies the symptoms associated with ASD (Hill, 2004). Researchers have consistently observed EF impairments in individuals with ASD, across many ages and levels of functioning (see Ozonoff et al., 2007). Notably, individuals with ASD exhibit poor planning (Robinson et al., 2009) and failure to mentally shift sets (Ozonoff et al., 2007).

Following the executive dysfunction theory, EF deficits should also underlie imagination deficits in ASD. Indeed, empirical evidence suggests that in NT children, EF constrains imaginative drawing development (Morra, 1995). As described above, the imaginative drawing process involves the manipulation of representations to generate something new and original (Vygotsky, 2004). Morra (1995, 2005) suggests that a child’s ability to modify an existing representation depends on the amount of attentional resources (“M-capacity,” the maximum number of schemes a person can simultaneously activate) as well as the activation of executive schemes that allow a child to set appropriate goals and monitor performance. According to this theory, EF is important in imaginative drawing because children must initially construct visuospatial representations of the drawing ideas they generate, maintain the representation in working memory, and monitor the unfolding drawing (Leevers and Harris, 1998; Low et al., 2009), suggesting important roles for planning and working memory in the course of this process. Indeed, in NT children, visuospatial planning is important for novel picture production (Low and Hollis, 2003; Thomas and Silk, 1990), and working memory is important in the manipulation of stereotypical drawing schemes (e.g. modifying the canonical representation of a human figure into a person in movement, see Morra, 2005).

Imagination and generativity

Generativity refers to the ability to generate or initiate an appropriate and novel response and is a precursor to effective rule-governed behavior (Turner, 1999). Generativity impairments in ASD were first reported by Turner (1999), who found that individuals with ASD performed poorly on tasks that required participants to name new uses for objects, or to suggest interpretations of meaningless line drawings. However, other studies have found no ASD impairments on a range of generativity tasks (e.g. Barnard et al., 2008).

Low et al. (2009) directly examined whether generativity is linked to imagination in children, with ASD and NT matched controls. They administered the Karmiloff-Smith drawing task (Karmiloff-Smith, 1990) and 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. By contrast, for NT children, generativity was linked to imaginative drawing, but this relation was mediated by false belief reasoning, not EF. This suggests that generativity influences imaginative drawing in children with ASD and NT children, but in different ways.

Imagination and the LPB

The Weak Central Coherence account suggests that individuals with ASD process information in a piecemeal fashion and fail to integrate information into a meaningful whole or gestalt (Frith, 1989, 2003). Revised and modified accounts of visual perception in children with ASD, such as the Enhanced Perceptual Functioning model, posit that ASD processing enhances the representation of local details and features, despite the intact ability to integrate the local features into a coherent, global whole (Plaisted Grant and Davis, 2009). The model suggests that a local orientation is the “default setting” of individuals with autism, whereas higher-order processing is an optional strategy (Mottron et al., 2006).

People with ASD tend to exhibit an LPB on visual tasks such as embedded figures tasks (Shah and Frith, 1983), block design tasks (Shah and Frith, 1993), and tasks assessing susceptibility to visual illusions (Happé, 1996, 1999). Other researchers (e.g. Plaisted Grant and Davis, 2009) have found that children with ASD show no preference for local detail, suggesting children with ASD can process global features normally.

LPB has been tested in copying tasks and observed in drawing style (e.g. Allen and Chambers, 2011; Sheppard et al., 2009). Surprisingly, very few researchers have addressed the role of LPB in drawing tasks that required the synthesis of various schemes into a unified drawing, which is an important component of imaginative drawing. Low et al. (2009) suggest that LPB may contribute to “hierarchical levels of drawing production” (p. 440) in imaginative drawing, but they failed to report on the relation between LPB and imagination. A study by Craig et al. (2001) examined imagination in drawing tasks in children with ASD and verbal mental age (MA)-matched mild learning disability and NT controls. Children with ASD were impaired on a task that required children to produce novel real or unreal entities by mixing categories of images (e.g. children were asked to draw an animal that is half fish and half mouse). Children with ASD tended to draw two real, separate entities (e.g. a separate fish and mouse). This finding indicates that combinatorial activities are disrupted in imaginative drawing for children with ASD. However, the researchers also found that children with ASD could successfully join two other images (making a “J” and a rotated “D” into an umbrella), suggesting that children with ASD are capable of fusing two representations (at least in simpler scenarios).

This study

The primary goal of this study is to examine the relations between imagination and cognitive processes in children with and without autism. We gauged the relative contribution of EF, generativity, and LPB to imagination in children with ASD and NT children. An examination of imaginative drawing is a valuable tool in understanding cognitive development in children with autism because drawing removes much of the verbal and social emphasis associated with other measures of imagination. Because previous research has shown a difference in ASD performance on imaginative drawings of people and houses (Ten Eycke and Müller, 2015), we will examine both types of stimuli in order to distinguish the social and non-social nature of these two types of stimuli.

The “multiple deficit account” of ASD (Charman and Swettenham, 2001) suggests that children with ASD show several cognitive atypicalities, which together explain the autism symptomatology. Accordingly, imagination deficits may be due to a combination of impairments in multiple cognitive processes. This is plausible, given that there are many, diverse cognitive demands for imagining. For example, a child must transform, modify, and combine previous held schemas, likely requiring a strong EF foundation, while the capability to generate a wide array of representations may depend on high generativity. Alternatively, imagination deficits might be attributable to specific impairment (e.g. executive dysfunction).

The second goal of this study was to determine whether the same processes are linked to imagination in children with ASD as in NT children. People with ASD are widely reported to have unique cognitive mechanisms and processes, distinct from NT cognition (for reviews, see Rajendran and Mitchell, 2007; Vanegas and Davidson, 2015). People with ASD also display unique neurocorrelates of cognitive processing (Mottron et al., 2006). It is therefore likely that children with and without ASD exhibit different relations between EF, generativity, and LPB and imagination.

The final goal of this study was to determine whether the contribution of different cognitive processes changes with age. This goal was motivated by two considerations. First, research suggests that younger and older children may rely on different approaches to the imaginative drawing task. For example, Karmiloff-Smith (1990) found that, whereas 4- to 6-year olds tended to delete or change the shape of elements of the figure, often making changes toward the end of the drawing, 8- to 10-year olds would often change the shape of the whole image, change the position or orientation of the image, make insertions of new elements, or make cross-category insertions. Second, ASD is a neurodevelopmental disorder, characterized by developmental deficits that manifest early in development (American Psychological Association, 2013). Therefore, it is possible that children with ASD fail to reach NT imaginative developmental milestones or are slower in achieving them. Therefore, we assessed whether age-moderated the relation between EF, generativity, and LPB and imagination.

Method

Participants

A total of 65 school-age children (5–14 years old) were recruited via word of mouth and community advertisements. Children with ASD were independently diagnosed by a qualified specialist, using either ADOS (Autism Diagnostic Observation Schedule; Lord et al., 2000) or ADI-R (Autism Diagnostic Interview–Revised; Rutter et al., 2003) diagnostic tools (a current diagnostic standard for the province where this study was conducted). Of the children in the ASD group, 16 had autistic disorder, 2 had Pervasive Developmental Disorder-Not Otherwise Specified (PDD-NOS), and 4 had Asperger syndrome. Caregivers of participants completed the Gilliam Autism Rating Scale-2 (GARS-2; Gilliam, 2006) in order to screen the NT group for possible ASD diagnoses. The GARS-2 is a frequency-based rating scale that helps identify ASD diagnosis and estimate severity of stereotyped behaviors, and social and communication deficits. Several participants in the ASD group had GARS-2 autism indices below 85, suggesting that ASD diagnosis was unlikely (Gilliam, 2006). However, diagnosis based on ADOS (Lord et al., 2000) and ADI-R (Rutter, Le Couteur and Lord, 2003) are considered to be more comprehensive and valid assessment tools and therefore these participants were included in analyses.

Overall, a total of six children were dropped from the ASD group because they failed a screening task (see below) and three children with a composite intelligence quotient (IQ) < 70 were excluded. A total of five children were dropped from the NT group because (a) their caregiver reported a diagnosis of a developmental disorder other than ASD or (b) their GARS-2 scores were in the “Possibly” or “Very Likely” range for an ASD diagnosis. This left 22 children with ASD (19 boys) and 29 NT children (12 boys) as participants in this study.

Children from the ASD and NT groups were matched on non-verbal IQ; verbal MA, non-verbal MA, and overall MA; and chronological age. Importantly, despite a wide age range, mean chronological age did not differ between groups, t(49) = 0.969, p = 0.337, and, according to Levene’s test for equality of variances, there was no significant difference in variance between the two groups F(1,40) = 0.219, p = 0.642. According to the Kolmogorov–Smirnov test of normality, chronological age follows a normal distribution for both groups, ASD: D(22) = 0.100, p = 0.200, NT: D(29) = 0.098, p = 0.200. Table 1 displays participant characteristics. Informed consent/assent was obtained from all participants.

Table 1.

Description of participants with autism spectrum disorder and neurotypically developing children.

	Group
	ASD (n = 22)		NT (n = 29)
	Mean (SD)	Range	Mean (SD)	Range
Age (months)
Chronological age	111.6 (31.5)	60–176	103.3 (28.7)	61–167
Non-verbal mental age	107.4 (36.3)	62–188	102.5 (36.8)	61–186
Verbal mental age	111.8 (41.2)	49–201	113.7 (36.7)	61–190
Overall mental age	109.6 (38.7)	58–194	109.0 (37.2)	61–185
Intelligence
Non-verbal IQ	96.7 (16.8)	59–127	98.4 (16.7)	68–134
Verbal IQ*	99.0 (15.1)	75–129	109.3 (12.8)	90–151
Overall IQ*	97.7 (15.1)	71–122	104.6 (14.9)	80–143
Autism indices
Stereotyped behaviors**	8.0 (3.0)	3–14	3.1 (1.5)	1–7
Communication**	7.8 (2.8)	1–14	2.6 (1.7)	1–8
Social interaction**	8.4 (3.4)	3–16	2.5 (1.5)	1–6
Autism index**	87.4 (17.2)	55–117	52.6 (8.5)	41–68

ASD: autism spectrum disorder; NT: neurotypically developing; SD: standard deviation; IQ: intelligence quotient; KBIT-2: Kaufman Brief Intelligence Test-2.

IQ was measured with the KBIT-2 (Kaufman and Kaufman, 2004) and autism indices were measured with the Gilliam Autism Rating Scale 2 (Gilliam, 2006). Group differences based on independent-samples t-test are also indicated.

p < 0.05, **p < 0.001.

Procedure

All children were seen individually by an experimenter, and the child and experimenter were seated next to each other. All tasks were administered in a single, 1.5-h session. All children were administered the screening task and, if successful, an intelligence task, the imaginative drawing task, and cognitive tasks.

Control measures

Screening task

Children were asked to copy three shapes (square, circle, and rhombus) from a two-dimensional image in order to confirm their ability to follow verbal instruction and to confirm adequate fine motor skill. This task was scored as pass or fail.

Intelligence

Children were administered the Kaufman Brief Intelligence Test-2 (KBIT II; Kaufman and Kaufman, 2004). The KBIT-2 offers a screening estimate of verbal and non-verbal intelligence, plus a composite IQ score.

The Karmiloff-Smith drawing task

Imaginative drawing was measured with Karmiloff-Smith’s (1990) imaginative drawing task, with a few modifications as proposed by Hollis and Low (2005) and Low et al. (2009). With the aid of a short story, children were asked to draw a picture of a person, making it “look as funny and strange looking as you can.” Children were then asked to draw a “real, normal looking person” (to confirm that the content in the imaginative drawings was novel and imaginative). In addition to the imaginative person condition, we had children draw an imaginative/real house. Sample drawings of imaginative people and houses are presented in Figure 1. The order of administration of house and person drawings were counterbalanced; the order of administration did not have an effect on outcome measures. We also examined an aggregate measure of imaginative person and house drawings, despite the distinction between the social and non-social nature of the stimuli (respective). We included the aggregate to acknowledge the complexity of imagination and to attempt to simulate the demands of everyday imagining.

Figure 1.

Sample drawings of imaginative people and houses by children with autism spectrum disorder (ASD) and neurotypically developing children (NT).

To control for the possibility that children misunderstood task instructions, children were given a test in which they identified eight real and impossible images, as outlined by Leevers and Harris (1998).

Drawings were coded based on scoring criteria used by Marsh et al. (1996). In each drawing, the proportion of imaginative content was calculated (the number of features that were imaginative was divided by the total number of features). Superficial changes that are possible in reality (e.g. hairstyle) and drawings of completely different but real entities (e.g. animals at the zoo) were not considered imaginative. Higher scores indicate more imaginative drawings.

The experimenter and a second independent rater (who was blind to the child’s group membership and age) evaluated all drawings. Raters were trained using sample drawings from the current and previous studies. Average ratings from the two raters were used. Interclass correlation was performed to determine consistency among raters; interrater reliability was high (>0.9).

Cognitive/perceptual measures

EF

Two EF tasks were administered: the Tower of Hanoi (ToH) and the self-ordered pointing task (SOPT). In the ToH task (adapted from Welsh, 1991), children were given a wooden model with three equally sized pegs and three differently sized rings which were arranged on the pegs in various start positions (depending on trial). Children were instructed to move all three rings to the right-most peg with the largest ring on bottom and the smallest on top (in order to match a duplicate model placed in front of them). Using an instructional story involving a family of monkeys, children were informed that they could only move one ring at a time, that a larger ring could not be placed on a smaller ring, and that the rings must remain on pegs when they were not being moved.

Children were presented with six trials (two problems each trial) that could be solved in two to seven moves. A trial was discontinued when a child performed more than 20 moves without reaching a solution. The task was discontinued when a child failed to solve two consecutive trials. The highest level of problem solved by a child was assigned as a score (with a maximum score of 7).

As a second measure of EF, the verbal and non-verbal conditions of the SOPT (Petrides and Milner, 1982) were administered (order counterbalanced). Children viewed pictures of concrete, single-syllable, nameable objects (verbal condition), or pictures of abstract designs that were difficult to name or encode verbally (non-verbal condition). Each condition included a two-item practice trial followed by trials of increasing length (two of each item length; 3, 4, 6, 9, 12 items). In a four-item trial, for example, four sheets of paper were presented sequentially, with each sheet depicting the same four stimuli, but each time in different spatial arrangements. Children were instructed to touch a different picture on each presentation. SOPT performance score was based on the maximum SOPT span achieved (number of consecutive novel pointing responses prior to the first error). An overall EF score was computed by averaging z-scores for ToH, and an average of verbal (V-SOPT) and non-verbal SOPT (NV-SOPT) maximum spans (z-scores were calculated based on each group’s mean).

Generativity

Three generativity tasks were used: the uses of objects task, the category fluency task, and the design fluency task. In the uses of objects task (see Turner, 1999), participants were shown one object (e.g. a cup, a piece of fabric) at a time and asked to describe uses to which it could be put. For each conventional item, a typical use (e.g. the cup could be used to drink from) and a novel use (e.g. the cup could be used as a doll’s hat) were provided. For each non-conventional item, a novel use was suggested (e.g. a piece of fabric could be used to make a doll’s shirt). Then, participants were invited to: “tell me all the other ways you think a [object] could be useful.” Participants were given a maximum of 2.5 min for each object. Children were given a score for the mean number of correct (appropriate and novel) uses they provided.

In the semantic fluency task, children were asked to (a) “tell me as many animals you know” and (b) “tell me as many foods you know.” Children were given 60 s (for each category) to generate as many words as possible belonging to each category. Children were given a score for the mean number of correct responses provided.

In the design fluency task (adapted from Baldo et al., 2001), participants were presented with a page of 35 squares, each of which contained an identical array of dots. Participants were instructed to make a different design in each square by connecting dots, always using four straight lines. Children were given three practice trials and then were told that they would have 60 s to draw as many different designs as they could. During the testing phase, the experimenter prompted the participant once (i.e. explained the error) if she or he drew three consecutive, incorrect designs (e.g. perseverated on the same design or used only three lines). A score was given for total number of correct/novel designs produced.

Z-scores, based on group means, for “total correct” scores for each of the three generativity tasks were averaged to create an overall generativity score.

LPB

Two tasks were administered to assess processing style bias: the Children’s Embedded Figure Task (CEFT; Witkin et al., 1971) and the Titchener Circles Illusion (Happé, 1993). For the CEFT, children were shown a cardboard cut-out of a target shape and asked to find the same shape hidden within a larger complex picture as quickly as they could. A maximum of 30 s was allowed to successfully complete each trial and 24 trials were administered. The outcome variable was the total number of correctly identified shapes.

A version of the Tichener Circles illusion was used as a second measure of processing bias (adapted from Happé, 1993). In this task, the experimenter presented six examples of the Titchener Circles Illusion with varying relative dimensions of the inner circles (making the illusion more or less strong), as well as six control items (only the inner circles were presented). The 12 figures were presented in random order, and children were asked to judge whether the two target items (inner circles) were the same size or different sizes. The outcome variable was totally correct for the illusion items.

For both processing bias tasks, higher scores indicate LPB and lower scores indicate a global processing bias. The overall processing bias score was calculated as an average of z-scores for Titchener Illusion and CEFT scores (based on group means). This study was approved by the University of Victoria Human Research Ethics Board (Ethics protocol number: 13-071).

Results

First, we examine performance on control tasks, followed by a brief analysis of group differences on imagination and cognitive tasks. We then examine the relative contribution of cognitive predictors to imaginative drawing by presenting (a) full and partial correlations and (b) regression models containing composite measures of EF, generativity, and LPB. Finally, we add autism indices, MA, and MA by cognitive predictor interactions to determine the influence of autism symptomatology and development on the cognition–imagination relationship.

Preliminary analyses

Children from the ASD group and the NT group were equally proficient at identifying which imagination control stimuli were real and which were “funny and strange looking” (ASD: M = 7.73, standard deviation (SD) = 0.63; NT: M = 7.93, SD = 0.26, t[26]) = 1.427, p = 0.165). Levene’s test indicated unequal variances (F = 11.543, p = 0.001), so degrees of freedom were adjusted from 49 to 26. There was also no significant group difference in discriminating the Titchener Circle illusion control stimuli (ASD: M = 5.36, SD = 0.73; NT: M = 5.45, SD = 0.95, t(49) = 0.348, p = 0.729).

Performance of children from the ASD group did not significantly differ from that of NT controls on the EF, generativity, and LPB tasks, with the exception that children with ASD reached a lower maximum NV-SOPT score (M = 6.36, SD = 1.79) than NT controls (M = 7.45, SD = 1.88), t(49) = 2.083, p = 0.043. The ASD group produced drawings of houses with an equal proportion of imaginative features as NT controls (ASD: M = 5.36, SD = 0.73; NT: M = 5.45, SD = 0.95, t(49) = 0.348, p = 0.729), but produced drawings of people with significantly lower proportions of imaginative features (ASD: M = 5.36, SD = 0.73; NT: M = 5.45, SD = 0.95, t(49) = 0.348, p = 0.729).

Cognitive predictors of imaginative drawing

The remaining analyses were conducted separately for children in the ASD and NT groups.

ASD group

Associations between cognitive predictors and imaginative drawing outcomes (full and partial) for the ASD group are reported in Table 2. Regression analyses were conducted to examine whether composite measures of EF, generativity, and LPB were uniquely associated with imaginative drawing. Results of regression analyses for the ASD group are reported in Table 3. Controlling for verbal intelligence and chronological age, only EF significantly predicted house imaginative drawing, over and above variance explained by generativity and LPB.

Table 2.

Correlations (bivariate and full partial with chronological age and verbal IQ controlled) between outcome imaginative drawing measures (person, house, and averaged drawing task conditions) and cognitive predictors for the ASD group.

Measure	Task condition
	Person		House		Aggregate
	Bivariate	Partial	Bivariate	Partial	Bivariate	Partial
CA	0.505*	–	0.125	–	0.390	–
IQ	0.251	–	0.459*	–	0.432*	–
MA	0.528*	–	0.314	–	0.518*	–
GARS-2
Stereotyped	0.124	0.122	0.360	0.407	0.293	0.333
Communication	0.152	0.242	0.202	0.257	0.216	0.312
Social	0.093	0.201	0.305	0.232	0.241	0.334
EF
NV-SOPT	0.492*	0.375	0.308	0.496*	0.492*	0.547*
V-SOPT	0.422	0.291	0.502*	0.582*	0.564**	0.546*
TOH	0.263	−0.009	0.376	−0.004	0.390	−0.008
Composite	0.527*	0.325	0.574**	0.559*	0.673**	0.553*
Generativity
Uses of Objects	0.260	−0.005	0.350	0.276	0.381	0.166
Category	0.552**	0.395	0.179	0.044	0.452*	0.283
Design	0.107	−0.112	0.061	−0.122	0.103	−0.147
Composite	0.389	0.148	0.247	0.120	0.391	0.170
LPB
Titchener	0.252	−0.100	−0.035	−0.494*	0.137	−0.368
CEFT	0.384	0.133	0.337	0.246	0.441*	0.237
Composite	0.411	0.017	0.181	−0.287	0.365	−0.165

ASD: autism spectrum disorder; CA: chronological age; IQ: intelligence quotient; MA: mental age; GARS-2: Gilliam Autism Rating Scale-2; EF: executive function; NV-SOPT: non-verbal self-ordered pointing task; V-SOPT: verbal self-ordered pointing task; TOH: Tower of Hanoi; Category: category fluency task; Design: design fluency task; LPB: local processing bias; Titchener: Titchener Illusion task; CEFT: Children’s Embedded Figure Task; KBIT-2: Kaufman Brief Intelligence Test-2.

IQ was measured with the KBIT-2 (Kaufman and Kaufman, 2004) and autism indices were measured using the GARS-2 (Gilliam, 2006).

p < 0.05, **p < 0.01.

Table 3.

Standardized regression coefficients for two control variables (verbal IQ and chronological age) and three cognitive predictors (executive function, generativity, and local processing bias) for the two conditions of the imaginative drawing task (person and house) and the aggregate imaginative drawing score.

	Task condition
	Person	House	Aggregate
Control variables
CA	0.336	−0.276	0.045
VIQ	0.105	0.312	0.252
Cognitive predictors
EF	0.347	0.891**	0.750*
Generativity	−0.057	−0.242	−0.180
LPB	−0.074	−0.169	−0.147
R ²	0.330	0.551*	0.503*

IQ: intelligence quotient; ASD: autism spectrum disorder; CA: chronological age; VIQ: verbal intelligence quotient; EF: executive function; LPB: local processing bias; KBIT-2: Kaufman Brief Intelligence Test-2.

Results are displayed for children in the ASD group. IQ was measured with the KBIT-2 (Kaufman and Kaufman, 2004). EF, Generativity, and LPB are composite measures.

p < 0.05, **p < 0.01.

NT group

Associations between cognitive predictors and imaginative drawing outcomes (full and partial) for the NT group are displayed in Table 4. Regression analyses were conducted to examine whether composite measures of EF, generativity, and LPB were uniquely associated with imaginative drawing. Results from these analyses can be found in Table 5. Controlling for verbal intelligence and chronological age, only generativity significantly predicted person imaginative drawing, over and above variance explained by EF and LPB.

Table 4.

Measure	Task condition
	Person		House		Aggregate
	Bivariate	Partial	Bivariate	Partial	Bivariate	Partial
CA	0.287	–	0.179	–	0.280	–
IQ	0.214	–	0.121	–	0.180	–
MA	0.328	–	0.210	–	0.184	–
GARS-2
Stereotyped	−0.103	−0.057	0.266	0.333	0.130	0.211
Communication	−0.134	0.196	0.172	0.401*	0.048	0.394*
Social	−0.501**	−0.377	0.172	0.370	−0.150	0.069
EF
NV-SOPT	0.202	0.128	0.189	0.129	0.241	0.161
V-SOPT	0.296	0.122	0.303	0.255	0.372*	0.249
TOH	0.078	0.012	0.234	0.199	0.206	0.150
Composite	0.229	0.096	0.324	0.274	0.350	0.249
Generativity
Uses of objects	0.176	0.028	0.276	0.217	0.288	0.172
Category	0.547**	0.472*	0.250	0.185	0.470*	0.384*
Design	0.340	0.216	0.027	−0.110	0.202	0.035
Composite	0.449*	0.347	0.231	0.148	0.404*	0.291
LPB
Titchener	−0.313	−0.220	0.040	0.087	−0.140	−0.053
CEFT	0.249	0.028	0.249	0.178	0.309	0.144
Composite	−0.031	−0.153	0.219	0.162	0.137	0.036

IQ: intelligence quotient; NT: neurotypically developing; CA: chronological age; MA: mental age; GARS-2: Gilliam Autism Rating Scale-2; EF: executive function; NV-SOPT: non-verbal self-ordered pointing task; V-SOPT: verbal self-ordered pointing task; TOH: Tower of Hanoi; Category: category fluency task; Design: design fluency task; LPB: local processing bias; Titchener: Titchener Illusion task; CEFT: Children’s Embedded Figure Task; KBIT-2: Kaufman Brief Intelligence Test-2.

IQ was measured with the KBIT-2 (Kaufman and Kaufman, 2004).

p < 0.05, **p < 0.01.

Table 5.

Regression coefficients for two control variables (verbal IQ and chronological age) and three cognitive predictors (executive function, generativity, and local processing bias) for the two conditions of the imaginative drawing task (person and house) and the aggregate imaginative drawing score.

Control variables	Task condition
	Person	House	Aggregate
CA	−0.151	−0.115	−0.162
VIQ	0.057	−0.022	0.016
Cognitive predictors
EF	0.294	0.307	0.374
Generativity	0.615*	0.194	0.468
LPB	−0.428	−0.017	−0.242
R ²	0.328	0.119	0.244

IQ: intelligence quotient; NT: neurotypically developing; CA: chronological age; VIQ: verbal intelligence quotient; EF: executive function; LPB: local processing bias; KBIT-2: Kaufman Brief Intelligence Test-2.

Results are displayed for children in the NT group. IQ was measured with the KBIT-2 (Kaufman and Kaufman, 2004). EF, Generativity, and LPB are composite measures.

p < 0.05.

Autism indices and development

Next, we examined autism indices and MA in greater detail. This allowed us to better understand whether differences in underlying cognitive processes in imaginative drawing were a function of autism symptomatology or of general developmental delay.

First, for both groups, we tested a model containing cognitive measures (EF, generativity, LPB) and three autism indices drawn from the GARS-2 (social interaction, communication, stereotyped behaviors). We did not include chronological age or verbal intelligence (to reduce collinearity). Only one predictor uniquely significantly predicted imaginative drawing: for the NT group, social interaction score significantly predicted the proportion of imaginative features in the person drawings (β = −0.608, p = 0.017), over and above cognitive predictors and communication/stereotyped behavior scores. None of the predictors for ASD analyses were significant.

Differences in cognitive predictors of imagination between groups may also have been due to the general developmental delay associated with ASD. In other words, it was possible that cognitive predictors only influence imaginative drawing at a particular point in development. Therefore, we examined whether relations between cognitive predictors and imagination outcomes were dependent on MA. We tested three models for each imagination outcome (person condition, house condition, and aggregate), each containing one cognitive predictor (EF, generativity, LPB), MA, and an interaction term obtained by multiplying the cognitive predictor by MA. The interaction between MA and LPB (above MA or LPB alone) was significant in predicting mean imaginative drawing score for the NT group (β = −1.722, p = 0.021), and the interaction between MA and EF (above MA or EF alone) was marginally significant in predicting aggregate imaginative drawing score for the NT group (β = −1.199, p = 0.058).

To further investigate these two interactions, we conducted post hoc probing using methods described by Holmbeck (2002). EF, LPB, and MA were centered on the NT group means, and two conditional moderator variables (high and low MA) were calculated by centering MA on 1 SD above and below the mean (±37.21, respective), as per the recommendation of Cohen et al. (2003). Next, two regression models were run for LPB, each one containing either high or low MA conditional moderator variables, an interaction between the moderator variable and LPB, and LPB (predicting aggregate imaginative drawing). The same process was repeated for EF.

For younger NT children, LPB was positively associated with imaginative drawing score (b = 0.208, p = 0.064), and for older NT children, LPB was negatively associated with imaginative drawing score (b = −0.222, p = 0.069), although both predictors were only marginally significant. This indicates that imagination was associated with LPB in younger children, and a more global processing bias in older children. For younger NT children, EF was significantly positively associated with imaginative drawing (b = 0.174, p = 0.032); however, EF and imagination were not significantly related in the older NT group (b = −0.073, p = 0.455).

Discussion

This study investigated the links between EF, generativity, LPB, and imaginative drawing. Results showed that EF was associated with imagination in children with ASD and young NT children. Generativity and the social interaction score predicted imagination in the person condition of the task for NT children. LPB was associated with better imagination in younger NT children, but a global processing bias was associated with better imagination in older NT children. These findings suggest that the type of cognitive processes recruited for imagination changes with age and that children with autism employ a cognitive strategy primarily based on EF in imaginative drawing.

Imagination and EF

We hypothesized that better EF would be associated with imaginative drawing because higher-order control processes are important in creating, managing, and maintaining novel mental representations (Leevers and Harris, 1998; Low et al., 2009). Although a relation between imagination and EF was not present in older NT children and the NT group as a whole, better EF was associated with more imaginative house drawings in children with ASD and young NT children, which is consistent with previous research (e.g. Low et al., 2009; Low and Hollis, 2003; Thomas and Silk, 1990). Specifically, the fact that measures of EF (SOPT and ToH) we used in this study involve visuospatial planning and working memory suggests that these components play an important role in imagination (Booth et al., 2003; Leevers and Harris, 1998). Future research should clarify the developmental relation between imagination and EF (see Carlson et al., 2013) and the contribution of other EF components to imagination.

In this study, EF and imagination were not associated in older NT children, suggesting that the approach to an imaginative drawing is qualitatively different in younger and older NT children. This is consistent with the finding by Karmiloff-Smith (1990) that younger children use an element-by-element approach and older children a more holistic approach to imaginative drawing. Performance-based measures of EF, such as the measures used in this study, are highly structured and are thought to rely on an algorithmic level of analysis, a type of processing that would be crucial for an element-by-element approach to an imaginative drawing (Toplak et al., 2013). However, this type of processing may not be helpful to the holistic approach of older children, which may rely on more integrative and reflective processes.

A second explanation draws on the work of Ribot (1906 [1901]), who suggested that imagination and rational thought are antagonistically related. Initially, imagination is “purely subjective and without any rational element” (p. 168), but with the gradual growth of reasoning imagination is increasingly suppressed, in creative individuals, into a new alliance with rational thought. Because reasoning is strongly associated with EF (Markovits, 2000), it is possible that EF may suppress imagination in later development. Investigation of the developmental trajectories of cognitive systems and how they interact with imagination will be an interesting avenue for future research.

Imagination and generativity

Contrary to previous suggestions (e.g. Turner, 1997), generativity did not play an important role in imaginative drawing in children with ASD. This finding might reflect an autism-specific cognitive approach to producing imaginative drawings that relies predominantly on EF. For example, instead of recalling many visual schemes to achieve high imaginative content, children with ASD may apply more elaborate transformations to a small set of existing representations, a process with a heavy EF (but not generativity) component.

Generativity was associated with imagination in the NT group, but only when drawing imaginative people. This is consistent with previous research suggesting that generativity is an important precursor for imagination (e.g. Turner, 1999). In this study, we also found the social interaction score to predict imagination in person drawings, perhaps partially mediating the effect of generativity on imagination.

Imagination and LPB

A bias for local details was predicted to impair imaginative drawing, yet no evidence for this association was found for children with ASD. If imaginative drawing is a process of combining internal visual representations (Vygotsky, 2004), it should involve the construction of global coherence. However, a bias toward processing local features is not synonymous with a deficit in integrating elements into a whole (Mottron et al., 2006; Plaisted et al., 2003), so it is possible that most children with ASD are proficient at integrating internal representations into novel, imaginative products. Correspondingly, Craig et al. (2001) suggest that children with ASD do not have a difficulty bolstering internal representations, as their sample of children with ASD were able to successfully combine real visual elements.

Processing bias was associated with imaginative drawing in NT children, and the association was age-dependent. In younger children, a bias toward local features was associated with high imaginative content, whereas in older children, a bias toward global features was associated with high imaginative content. It is possible that this pattern of associations reflects a developmental shift from a local, element-by-element processing to a more global (e.g. combining internal representations prior to drawing execution) processing approach to imaginative drawing, as discussed above.

Group differences

It is important to note that we found virtually no group differences in performance on the main measures of EF, generativity, and LPB. One reason for this is that our sample of children with ASD was high functioning. Although it was necessary to exclude children with low intellectual ability in order to have children comprehend and complete the cognitive tasks, this limits the generalizability of our findings.

Although performance on cognitive tasks was comparable, our results suggest that there are different cognitive processes involved in imaginative drawing in children with and without autism. Given the age-moderated effects of EF and LPB on imagination in NT children, it is possible that these differences are explained by the developmental delay associated with ASD. Consequently, we should observe similar shifts in the cognitive processes involved in imagination in an older sample of children with ASD. Differences may also be attributable to the unique cognitive profile of people with ASD. The cognitive profile of individuals with autism often includes not only delays or lifelong impairments in intelligence and EF but also enhanced processing in other domains (e.g. Hill, 2004). Researchers frequently report that some people with autism “think in pictures,” that is, use visual rather than verbal representations (see Kunda and Goel, 2008). Accordingly, it is likely that children with ASD use unique cognitive strategies in imagination, EF, and LPB tasks.

Limitations

Due to the limited sample size and the wide age range of this study, future research confirming our findings and exploring developmental changes would benefit from large-scale and longitudinal projects. As with most EF research, there are several psychometric weaknesses associated with lab-based tasks. For example, the ToH has poor test re-test reliability (e.g. Bishop et al., 2001), and the SOPT span score is considered unstable (Ross et al., 2007). Future studies should employ several indicators of each EF component and use confirmatory factor analysis to control for measurement error and nonexecutive task demands (Müller and Kerns, 2015) to add to our knowledge about the association between EF and imagination.

Conclusion

We found that elements of the unique ASD cognitive profile (EF, generativity, and LPB) were associated with imaginative drawing in children with and without ASD; however, the relations were often complex and unexpected. Our findings have implications for encouraging the development of imagination and control processes in children with and without ASD. Future research into the causal relation between EF, LPB, and imagination should further clarify avenues of treatment and intervention.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This project was funded by the Sarah Spencer Foundation.

References

Allen

Chambers

(2011) Implicit and explicit understanding of ambiguous figures by adolescents with autism spectrum disorder. Autism 15(4): 457–472.

American Psychological Association (2013) Diagnostic and Statistical Manual of Mental Disorders-DSM-5. Washington, DC: American Psychological Association.

Baldo

Shimamura

Delis

. (2001) Verbal and design fluency in patients with frontal lobe lesions. Journal of the International Neuropsychological Society 7(5): 586–596.

Barnard

Muldoon

Hasan

. (2008) Profiling executive dysfunction in adults with autism and comorbid learning disability. Autism 12(2): 125–141.

Bishop

DVM

Aamodt-Leeper

Creswell

. (2001) Individual differences in cognitive planning on the tower of hanoi task: neuropsychological maturity or measurement error? Journal of Child Psychology and Psychiatry 42(4): 551–556.

Booth

Charlton

Hughes

. (2003) Disentangling weak coherence and executive dysfunction: planning drawing in autism and attention-deficit/hyperactivity disorder. Philosophical Transactions of the Royal Society B: Biological Sciences 358(1430): 387–392.

Burton

(2010) Lassoing the world: drawing along the spectrum. In: Levett Gerber

Kellman

(eds) Understanding Students with Autism through Art. Reston, VA: National Art Education Association, pp.102–111.

Carlson

White

Taylor

(2013) Executive function, pretend play, and imagination. In: Carlson

White

(eds) The Oxford Handbook of the Development of Imagination. New York: Oxford University Press, pp.161–174.

Charman

Baron-Cohen

(1997) Brief report: prompted pretend play in autism. Journal of Autism and Developmental Disorders 27(3): 325–332.

10.

Charman

Swettenham

(2001) Repetitive behaviors and social-communicative impairments in autism: implications for developmental theory and diagnosis. In: Burack

Charman

Yirmiya

. (eds) The Development of Autism: Perspectives from Theory and Research. Hillsdale, NJ: Lawrence Erlbaum, pp.325–345.

11.

Cohen

West

. (2003) Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.

12.

Craig

Baron-Cohen

(1999) Creativity and imagination in autism and asperger syndrome. Journal of Autism and Developmental Disorders 29(4): 319–326.

13.

Craig

Baron-Cohen

Scott

(2001) Drawing ability in autism: a window into the imagination. The Israel Journal of Psychiatry and Related Sciences 38(3–4): 242–253.

14.

Frith

(2003) What do imaging studies tell us about the neural basis of autism? Novartis Foundation Symposium 251: 149–166; discussion 166–176, 281–297.

15.

Frith

(1989) Autism: Explaining the Enigma. Oxford: Wiley-Blackwell.

16.

Gilliam

(2006) GARS-2: Gilliam Autism Rating Scale. 2nd ed. Austin, TX: Pro-Ed.

17.

Happé

(1999) Autism: cognitive deficit or cognitive style? Trends in Cognitive Sciences 3(6): 216–222.

18.

Happé

(1993) Communicative competence and theory of mind in autism: a test of relevance theory. Cognition 48(2): 101–119.

19.

Happé

FGE

(1996) Studying weak central coherence at low levels: children with autism do not succumb to visual illusions: a research note. Journal of Child Psychology and Psychiatry 37(7): 873–877.

20.

Hill

(2004) Executive dysfunction in autism. Trends in Cognitive Sciences 8(1): 26–32.

21.

Hollis

Low

(2005) Karmiloff-Smith’s RRM distinction between adjunctions and redescriptions: it’s about time (and children’s drawings). The British Journal of Developmental Psychology 23(4): 623–644.

22.

Holmbeck

(2002) Post-hoc probing of significant moderational and mediational effects in studies of pediatric populations. Journal of Pediatric Psychology 27(1): 87–96.

23.

Jarrold

Boucher

Smith

(1993) Symbolic play in autism: a review. Journal of Autism and Developmental Disorders 23(2): 281–307.

24.

Jarrold

Boucher

Smith

(1996) Generativity deficits in pretend play in autism. The British Journal of Developmental Psychology 14(3): 275–300.

25.

Karmiloff-Smith

(1990) Constraints on representational change: evidence from children’s drawing. Cognition 34(1): 57–83.

26.

Kaufman

(2004) Kaufman Brief Intelligence Test. 2nd ed. Bloomington, MN: Pearson, Inc.

27.

Kunda

Goel

(2008) How thinking in pictures can explain many characteristic behaviors of autism. In: Proceedings of 7th IEEE international conference on development and learning: ICDL 2008, Monterey, CA, 9–12 August, pp.304–309.

28.

Leevers

Harris

(1998) Drawing impossible entities: a measure of the imagination in children with autism, children with learning disabilities, and normal 4-year-olds. Journal of Child Psychology and Psychiatry, and Allied Disciplines 39(3): 399–410.

29.

Lord

Rutter

DiLavore

. (2000) Autism Diagnostic Observation Schedule. Los Angeles, CA: Western Psychological Services.

30.

Low

Hollis

(2003) The eyes have it: development of children’s generative thinking. International Journal of Behavioral Development 27(2): 97–108.

31.

Low

Goddard

Melser

(2009) Generativity and imagination in autism spectrum disorder: evidence from individual differences in children’s impossible entity drawings. British Journal of Developmental Psychology 27(2): 425–444.

32.

Markovits

(2000) A mental model analysis of young children’s conditional reasoning with meaningful premises. Thinking & Reasoning 6(4): 335–347.

33.

Marsh

Landau

Hicks

(1996) How examples may (and may not) constrain creativity. Memory & Cognition 24(5): 669–680.

34.

Morra

(1995) A neo-piagetian approach to children’s drawings. In: Lange-Kuttner

Glyn

(eds) Drawing and Looking: Theoretical Approaches to Pictorial Representation in Children: The Developing Body and Mind. London: Harvester Wheatsheaf.

35.

Morra

(2005) Cognitive aspects of change in drawings: a neo-piagetian theoretical account. British Journal of Developmental Psychology 23(3): 317–341.

36.

Mottron

Dawson

Soulières

. (2006) Enhanced perceptual functioning in autism: an update, and eight principles of autistic perception. Journal of Autism and Developmental Disorders 36(1): 27–43.

37.

Müller

Kerns

(2015) The development of executive function. In: Liben

Müller

Lerner

(eds) Handbook of Child of Psychology and Developmental Science (7th ed. vol. 2): Cognitive Processes. Hoboken, NJ: Wiley, pp.571–623.

38.

Ozonoff

South

Provencal

(2007) Executive functions in autism. In: Perez

Gonzalez

Llorente Comi

. (eds) New Developments in Autism: The Future is Today. Philadelphia, PA: Jessica Kingsley Publishers, pp.185–213.

39.

Petrides

Milner

(1982) Deficits on subject-ordered tasks after frontal-and temporal-lobe lesions in man. Neuropsychologia 20(3): 249–262.

40.

Plaisted Grant

Davis

(2009) Perception and apperception in autism: rejecting the inverse assumption. Philosophical Transactions of the Royal Society of London: Series B: Biological Sciences 364(1522): 1393–1398.

41.

Plaisted

Saksida

Alcántara

. (2003) Towards an understanding of the mechanisms of weak central coherence effects: experiments in visual configural learning and auditory perception. Philosophical Transactions of the Royal Society of London: Series B: Biological Sciences 358(1430): 375–386.

42.

Rajendran

Mitchell

(2007) Cognitive theories of autism. Developmental Review 27: 224–260.

43.

Ribot (1906 [1901]) Essay on the Creative Imagination. Chicago, IL: The Open Court Publishing Company.

44.

Robinson

Goddard

Dritschel

. (2009) Executive functions in children with autism spectrum disorders. Brain and Cognition 71: 362–368.

45.

Ross

Hanouskova

Giarla

. (2007) The reliability and validity of the self-ordered pointing task. Archives of Clinical Neuropsychology 22(4): 449–458.

46.

Rutter

Le Couteur

Lord

(2003) Autism Diagnostic Interview-Revised. Los Angeles, CA: Western Psychological Services.

47.

Scott

(2013) The development of imagination in children with autism. In: Taylor

(ed.) The Oxford Handbook of the Development of Imagination. New York: Oxford University Press.

48.

Scott

Baron-Cohen

(1996) Imagining real and unreal things: evidence of a dissociation in autism. Journal of Cognitive Neuroscience 8(4): 371–382.

49.

Shah

Frith

(1983) An islet of ability in autistic children: a research note. Journal of Child Psychology and Psychiatry, and Allied Disciplines 24(4): 613–620.

50.

Shah

Frith

(1993) Why do autistic individuals show superior performance on the block design task? Journal of Child Psychology and Psychiatry, and Allied Disciplines 34(8): 1351–1364.

51.

Sheppard

Ropar

Mitchell

(2009) Perceiving the impossible: how individuals with autism copy paradoxical figures. Autism 13(4): 435–452.

52.

Singer

(2013) Historical overview of research on imagination in children. In: Taylor

(ed.) The Oxford Handbook of the Development of Imagination. New York: Oxford University Press, pp.11–27.

53.

Ten Eycke

Müller

(2015) Brief report: new evidence for a social-specific imagination deficit in children with autism spectrum disorder. Journal of Autism and Developmental Disorders 45(1): 213–220.

54.

Thomas

Silk

AMJ

(1990) An Introduction to the Psychology of Children’s Drawings. New York: New York University Press.

55.

Toplak

West

Stanovich

(2013) Practitioner review: do performance-based measures and ratings of executive function assess the same construct? Journal of Child Psychology and Psychiatry, and Allied Disciplines 54(2): 131–143.

56.

Turner

(1997) Toward an executive dysfunction account of repetitive behaviour in autism. In: Russell

(ed.) Autism as an Executive Disorder. New York: Oxford University Press, pp.57–99.

57.

Turner

(1999) Generating novel ideas: fluency performance in high-functioning and learning disabled individuals with autism. Journal of Child Psychology and Psychiatry, and Allied Disciplines 40(2): 189–201.

58.

Vanegas

Davidson

(2015) Investigating distinct and related contributions of weak central coherence, executive dysfunction, and systemizing theories to the cognitive profiles of children with autism spectrum disorders and typically developing children. Research in Autism Spectrum Disorders 11: 77–92.

59.

Vygotsky

(2004) Imagination and creativity in childhood. Journal of Russian and East European Psychology 42(1): 7–97.

60.

Wechsler

(1992) WISC-III: Wechsler Intelligence Scale for Children: Third Edition: Manual. San Antonio, TX: Psychological Corporation.

61.

Welsh

(1991) Rule-guided behavior and self-monitoring on the Tower of Hanoi disk-transfer task. Cognitive Development 6(1): 59–76.

62.

Witkin

Oltman

Raskin

. (1971) Manual for the Embedded Figures Test, Children’s Embedded Figures Test, and Group Embedded Figures Test. Palo Alto, CA: Consulting Psychologists Press.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB