Perception of mirror symmetry in autism spectrum disorders

Abstract

Gestalt grouping in autism spectrum disorders (ASD) is selectively impaired for certain organization principles but for not others. Symmetry is a fundamental Gestalt principle characterizing many biological shapes. Sensitivity to symmetry was tested using the Picture Symmetry Test, which requires finding symmetry lines on pictures. Individuals with ASD showed decreased sensitivity to symmetry and a correlation of test performance with performance IQ. Decreased sensitivity for symmetry in ASD is discussed in relation to reduced visual experience of faces in early development.

Keywords

autism symmetry Gestalt

Darwin suggested that ‘the eye prefers symmetry or figures with some regular recurrence’ (see Darwin on the sense of beauty, 1874: 93). This preference shows a special role of symmetry for visual organization. Symmetry is a global rather than local feature in visual organization (Pomerantz and Pristach, 1989). Although it can be detected preattentively and automatically, broad evidence suggests that symmetry mechanisms are not working by point-by-point comparisons (Wagemans, 1995). The Gestalt law of Praegnanz, or figural goodness, incorporates symmetry as a fundamental principle (Koffka, 1935). Gestalt processing has recently received attention in research on autism spectrum disorders (ASD). It has been shown that Gestalt processing is selectively impaired, depending on task specificities and the principle under consideration (Brosnan et al., 2004; Falter et al., 2010). In particular, processing of the Gestalt principle of proximity is spared in ASD, while processing of similarity is impaired (Falter et al., 2010). An unexpected effect in the study by Falter and colleagues was an interference of symmetry of the stimulus display with similarity grouping in the control group, which seemed to be absent in the ASD group.

The aim of the current study was to put the hypothesis of reduced symmetry processing in ASD to a test. For this purpose we tested the understanding of the principle of symmetry in a group of individuals with ASD and a typically-developing (TD) group using the Picture Symmetry Test. This test is adapted from Mirror Puzzles (e.g. Walter, 1988), which have been successfully used in mathematical education to teach the principles of bilateral (mirror) symmetry. As discussed above, previous findings predict that participants with ASD will be impaired on this task, indicating a decreased sensitivity to symmetry.

Methods

Participants

Sixteen individuals with ASD (25 ± 8 years) and 16 TD participants (25 ± 6 years) were recruited through their prior involvement in studies at the University of Oxford. Standardized clinical measures (Autism Diagnostic Interview-Revised (ADI-R), Lord et al., 1994; Autism Diagnostic Observation Schedule-Generic (ADOS-G), Lord et al., 2000) were performed. ASD diagnoses were regarded as confirmed if ADI-R scores did not fall more than one point below autism spectrum cut-off in more than one domain. Three participants had a diagnosis of high-functioning autism and the remaining participants of Asperger syndrome (Diagnostic and Statistical Manual of Mental Disorders IV Text Revision (DSM-IV-TR), American Psychiatric Association, 2000). Verbal and performance IQ scores were obtained using the Wechsler Abbreviated Scale of Intelligence (WASI; Wechsler, 1999). Exclusion criteria were comorbid psychiatric disorders (any psychiatric diagnosis in the TD group), medication and WASI scores below 70. None of the individuals selected for the study had to be excluded. ASD and TD groups were well-matched by chronological (t(30) = −0.14, p = .89) and mental age (verbal IQ: t(30) = −0.99, p = .33; performance IQ: t(30) = 0.67, p = .51). All participants had normal or corrected-to-normal vision. Ethics approval was formally obtained. Written informed consent was obtained from all participants (or parents) prior to testing. Demographic data is given in Table 1.

Table 1.

Demographic details of the study participants

	Autistic spectrum disorder (ASD) group (N = 16; 1 female)			Typically developing (TD) group (N = 16; 3 females)
	Mean	SD	Range	Mean	SD	Range
Age^a	25:9	8:5	16:9–42:11	25:4	6:6	14:10–37:6
VIQ	113.2	10.6	89–127	109.5	10.5	95–139
PIQ	113.5	12.5	92–136	115.9	7.5	104–125
ADI-A	16.3	5.0	9–26
ADI-B	14.2	3.9	9–21
ADI-C	6.1	3.2	2–12
ADOS-A	3.4	1.7	1–7
ADOS-B	6.1	2.8	1–11
ADOS-C	1.2	1.1	0–4

Ages are years:months

VIQ: verbal IQ, PIQ: performance IQ, ADI-A: Autism Diagnostic Interview-Revised (ADI-R) Social Interaction Domain, ADI-B: ADI-R Communication Domain, ADI-C: ADI-R Repetitive Behaviours Domain, ADOS-A: Autism Diagnostic Observation Schedule-Generic (ADOS-G) Communication Domain, ADOS-B: ADOS-G Social Interaction Domain, ADOS-C: ADOS-G Stereotyped Behaviours and Restricted Interests Domain.

The picture symmetry test

Stimuli (Figure 1) were provided courtesy of Marion Walter and were adapted from the Mirror Puzzles (Walter, 1988), a game developed to teach the principles of mirror symmetry in educational settings. In the original Mirror Puzzles children are provided with a mirror and a target picture (see inset, Figure 1) and have to try to re-create a series of trial pictures (e.g. Figure 1a–c) from the target picture using the mirror. In the adaptation for the Picture Symmetry Test no mirror was used but instead participants were provided with a transparent sheet with a black line (the ‘imaginary mirror’) and had to indicate how the trial pictures could be re-created from the target picture by placing the line onto the target picture. The purpose of using a symmetry line rather than a real mirror was to avoid participants solving the task by a mere trial and error strategy. In each trial a new trial picture was presented to them and they had to show where on the target picture they would place the line in order to re-create the respective trial picture. If a trial picture could not be re-created from a target picture, participants had to verbally respond accordingly (e.g. ‘the picture cannot be re-created’). The time to complete each trial was taken with a stopwatch from the start of a trial to correct placement of the symmetry line or a verbal response respectively. Participants had to complete one training series and two test series of eight trials each. The order of the pictures was fixed across participants. Feedback was given in the training series to ascertain that participants understood the instructions. No feedback was given in the test series. Four trial pictures without a symmetry line served as catch trials.

Figure 1.

Example stimuli depicting the target stimulus on the top left, two trial stimuli (a,c) with and without a bilateral symmetry line (b). Note that the figures were presented in color in the actual task.

Results

Between-group differences (Figure 2)

Medians (± quartile deviation) for accuracy were 78 ± 10% in the ASD group and 88 ± 9% in the TD group and medians for reaction times (RT) were 20 ± 7 s in the ASD group and 16 ± 3 s in the TD group. A significant negative correlation between accuracy and RT scores (r = –.658, p = .006) indicated a speed-accuracy trade-off in the ASD group. No trade-off was found in the TD group. Inverse efficiency is a standard composite measure of reaction times and accuracy scores (RT/accuracy for each individual), which corrects reaction times by accuracy scores and thereby provides a measure that is free from the bias of speed-accuracy trade-offs. Smaller inverse efficiency (IE) scores signify better performance (see Falter et al., 2010). A Mann–Whitney U test was performed to evaluate the hypothesis that the TD group (TD median = 0.18 ± 0.04) performed better on the task than the ASD group (ASD median = 0.24 ± 0.11). The results were in the expected direction and significant (z = −1.982, p = .047, r = –.35).

Figure 2.

Reaction times, accuracy and inverse efficiency scores (median ± quartile deviation).

Correlations between performance and levels of development

Spearman test was used to correlate IE with age and IQ scores. There was a significant correlation between IE and performance IQ (r = –.746, p = .001) in the ASD group (see Figure 3). There were no other correlations in the ASD or the TD group.

Figure 3.

Inverse efficiency over performance IQ for the ASD (left) and theTD groups (right).

Discussion

Sensitivity to symmetry was tested in ASD using the Picture Symmetry Test. Participants with ASD showed a decreased sensitivity to symmetry as previously predicted (Falter et al., 2010). Many biological shapes possess a bilateral symmetry, such as human faces and bodies as well as many fruits and vegetables (Wilson and Wilkinson, 2002). Could decreased lack of visual experience of human faces in ASD be related to decreased sensitivity for symmetry? Although symmetrical patterns are processed faster and are preferred over asymmetrical patterns in adults, infants do not show symmetry preference before 12 months of age (Bornstein et al., 1981). Bornstein and colleagues interpret this finding as evidence that symmetry results from visual experience in particular with faces. Gaze aversion and the resulting decreased viewing of faces in early development might well contribute to a decreased sensitivity for symmetry in ASD. Natural faces are never perfectly symmetrical (Wagemans, 1995) and therefore represent a particular challenge for many models of symmetry perception, in contrast to objects and artificial faces, which show fewer ambiguities (Poirier and Wilson, 2010). Hence, expertise with natural faces might lead to a more robust symmetry mechanism beyond that acquired from experience with object symmetries. It will need to be assessed though how difficulties with two-dimensional symmetry processing in ASD, as shown in the current study, translate to three-dimensional face perception. The current results are in line with previous studies (Falter et al., 2010), suggesting that Gestalt principles relying more on top-down processing are impaired in ASD. Symmetry can be regarded as a type-N relationship (Pomerantz, 1983), which is characterized by the nature of elements (i.e. elements cannot be exchanged without losing the global configuration). This reliance of symmetry on top-down processing is in keeping with the correlation of task performance with IQ in the ASD group, possibly reflecting increasing top-down processing abnormalities with decreasing levels of functioning. As a possible limitation the Picture Symmetry Test includes the requirement of a symmetry line placement. Although correct placement of a symmetry line relies on perception of symmetry in the picture, performance could be tested by a purer perceptual task.

In conclusion, bilateral symmetry processing was found to be atypical in ASD. A speculative relation of decreased sensitivity for symmetry and decreased experience with human faces will need to be tested by future research.

Footnotes

Acknowledgements

Thanks to all participants for their time and commitment and to The Baily Thomas Charitable Trust and the German Research Council (DFG) for financial support.

References

American Psychiatric Association (2000) Diagnostic and Statistical Manual of Mental Disorders, 4th edn. Washington, DC: American Psychiatric Press, Inc.

Bornstein

Ferdinandsen

Gross

(1981) Perception of symmetry in infancy. Developmental Psychology 17(1): 82–6.

Brosnan

Scott

Fox

Pye

(2004) Gestalt processing in autism: Failure to process perceptual relationships and the implications for contextual understanding. Journal of Child Psychology and Psychiatry 45: 459–69.

Darwin

(1874) The Descent of Man, and Selection in Relation to Sex, 2nd edn. London: John Murray.

Falter

Plaisted Grant

Davis

(2010) Object-based attention reveals selective abnormalities of visual integration in autism. Autism Research 3(3): 128–36.

Koffka

(1935) Principles of Gestalt Psychology. New York: Harcourt Brace Jovanovich.

Lord

Risi

Lambrecht

Cook

Leventhal

DiLavore

. (2000) The Autism Diagnostic Observation Schedule-Generic: A standard measure of social and communication deficits associated with the spectrum of autism. Journal of Autism and Developmental Disorders 30: 205–23.

Lord

Rutter

Le Couteur

(1994) Autism Diagnostic Interview – Revised: A revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. Journal of Autism and Developmental Disorders 24: 659–85.

Poirier

FJAM

Wilson

(2010) A biologically plausible model of human shape symmetry perception. Journal of Vision 10(1): 1–16.

10.

Pomerantz

(1983) Global and local precedence: Selective attention in form and motion perception. Journal of Experimental Psychology: General 112: 516–40.

11.

Pomerantz

Pristach

(1989) Emergent features, attention, and perceptual glue in visual form perception. Journal of Experimental Psychology 15(4): 635–49.

12.

Wagemans

(1995) Detection of visual symmetries. Spatial Vision 9(1): 9–32.

13.

Walter

(1988) The Mirror Puzzle Book. Norfolk: Tarquin Publications.

14.

Wechsler

(1999) Wechsler Abbreviated Scale of Intelligence. San Antonio, TX: The Psychological Corporation.

15.

Wilson

Wilkinson

(2002) Symmetry perception: a novel approach for biological shapes. Vision Research 42: 589–597.