Lack of Local Bias during Selective Attention Global-Local Processing in Autism Spectrum Disorders

Shannon A. Johnson, Ph.D.1, Leslie M. Blaha, B.Sc.2, Robin R. Murphy, Ph.D.2, James T. Townsend, Ph.D.2, Vanessa A. Bruce, B.A.3, and Julie C. Stout, Ph.D.4. (1) Departments of Psychology, Pediatrics, & Psychiatry, Dalhousie University, 1355 Oxford Street, Department of Psychology, Halifax, NS B3H 4J1, Canada, (2) Department of Psychological and Brain Sciences, Indiana University, 1101 East 10th Street, Bloomington, IN 47405, (3) Department of Psychology, University of Windsor, 173 Chrysler Hall South, 401 Sunset Avenue, Windsor, ON N9B 3P4, Canada, (4) School of Psychology, Psychiatry and Psychological Medicine, Monash University, Faculty of Medicine, Nursing and Health Sciences, Melbourne, 3800, Australia

Background: Previous studies of global-local (GL) processing in autism spectrum disorders (ASD; Rinehart et al., 2000; Plaisted et al., 1999) suggest a lack of both global advantage and global interference. However, their methods differed fundamentally (including unlimited stimulus presentation time, lack of spatial uncertainty, few trials) from traditional designs of GL studies. Studies of controls have indicated that such changes in GL task parameters significantly alter performance (Kimchi, 1992).

Objectives: In the current study, we characterized GL processing in ASD during selective attention tasks using traditional cognitive science methods.

Methods: Sixteen high-functioning individuals with an ASD and 16 age- and IQ-matched comparison subjects completed eight selective attention GL tasks. Four tasks used hierarchical letter stimuli and four used number stimuli. For each type of stimulus, we employed two methodologies: 1) replication of methods used in previous ASD GL studies, but with a greater number of trials and 2) methods consistent with traditional designs of GL studies, incorporating brief presentation times, spatial uncertainty, and backward masking.

Results: Repeated measures ANOVAs revealed main effects for attended dimension with faster response times for the global versus local dimension for both groups on all tasks. Contrary to findings from previous ASD studies, global interference effects were demonstrated by both groups for both replication and traditional GL tasks, and neither group demonstrated local interference. Analyses at the individual level indicated that few participants responded faster when attending to local versus global information. When present, this atypical result occurred at the same rate in both groups. Similarly, only a few control and ASD participants showed local interference.

Conclusions: We found typical GL processing in high-functioning ASD participants. Both control and ASD groups demonstrated global advantage, global interference, and lack of local interference. Findings are discussed in light of cognitive theories of ASD and methodological considerations.


See more of Cognition 1
See more of Oral Presentations

See more of The International Meeting for Autism Research (London, May 15-17, 2008)