The Local Bias in ASC: Weak Central Coherence or a Deficit in Executive Control?

Friday, May 15, 2015: 11:30 AM-1:30 PM
Imperial Ballroom (Grand America Hotel)
R. Booth1, A. W. Paciorek2 and F. Happe3, (1)Institute of Child Health, University College London, London, United Kingdom, (2)Institute of Modern Languages, Pedagogical University of Krakow, Krakow, Poland, (3)Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, United Kingdom

“Weak central coherence” refers to a processing bias in autism spectrum conditions (ASC), involving enhanced local processing and reduced global integration of information. In language this has been demonstrated in sentence completion tasks: people with ASC make more local completions then controls, even when globally inappropriate. Since autism is known to involve a deficit in executive function, the question arises whether the local processing bias can be encompassed within executive dysfunction.


To examine whether the local processing bias as evidenced on sentence completion tasks can be explained by a deficit in executive function or whether it represents a separate cognitive style. 


The study was part of a larger project which tested boys with ADHD (n=30), ASC (n=31) and a matched control group (n=31), aged 8-16, on a battery of tasks including sentence completion, IQ and tests of verbal fluency. The sentence completion tasks involved two variants: generation and suppression. In generation participants were asked to provide a coherent completion, e.g., “light” after hearing “When you go to bed, turn off the…”. In suppression participants were asked to provide a completion that did not make sense, e.g. “swimming” to the above example. There were two types of sentence: those whose completion was determined globally or locally. In the above example “turn off the…” provides most of the information needed, so the completion is ‘local’. Conversely, in “Most cats see very well at…” information must be incorporated from earlier in the sentence, so this item was classified as ‘global’. We hypothesized that if weak central coherence in autism is separate from executive dysfunction, ASC children will find it easier to inhibit meaningful completions if they are globally meaningful, while those with ADHD will experience similar difficulty inhibiting both types of completion.


All groups produced more meaningful completions to local than global items in the generation task, and more failures to suppress local than global completions in the suppression task with no difference between the ADHD and ASC groups (F < 1). However, group differences were evident in response times (RTs) to produce correct (i.e., non-meaningful) completions in the suppression task: ADHD correct RTs were slower for global than local items, whereas ASC correct RTs were slower for local than global items (as predicted by a local processing bias). An ANCOVA controlling for WISC vocabulary scale, revealed a significant interaction between ADHD/ASC and global/local, F (1, 52) = 4,68, p < 0.5. Also, ASC was the only group in which overall accuracy on the generation and suppression tasks correlated with verbal fluency (r = 0.61, p < 0.001) and age (r = 0.54 , p < 0.05).


The findings suggest weak central coherence explains cognitive bias on the present task better than executive dysfunction, given the difference in performance between ASC and ADHD groups. Differences in the reaction time distribution, as well as the correlation with verbal fluency and age, suggest that people with ASC may be using a different strategy to complete the meaning suppression task.