Preserved Sensitivity to Higher-Order Conceptual Versus Lower-Level Perceptual Information During Explicit Verbal Memory Encoding In ASD: Limits to EPF and WCC?

Background: It is well established that individuals with Autism Spectrum Disorder (ASD) experience difficulties in utilising conceptual relations among stimuli to facilitate memory (Tager-Flusberg, 1991; Brit J Dev Psych, 9, 417-413). The reasons for this difficulty, however, remain unclear. According to the Enhanced Perceptual Functioning theory (EPF; Mottron et al., 2006; JADD, 36, 27-43), encoding of conceptual information in ASD is sometimes compromised by interference from enhanced processing of lower-level perceptual information. This view is supported by  studies employing levels of processing manipulations showing that cued recall following deep, conceptual encoding (e.g. ‘Is this a Fruit?) is preserved in ASD whilst cued recall following shallow, perceptual encoding (e.g. ‘Is this a 2-syllable word’?) is superior. The question remains, however, of whether individuals with ASD also spontaneously demonstrate an encoding superiority for low-level perceptual information. 

Objectives: To assess the degree to which individuals with ASD spontaneously encode perceptual and conceptual aspects of word stimuli (i.e. when no specific encoding instructions are provided).

Methods: 20 adolescents and adults with a confirmed diagnosis of ASD and 20 age and IQ matched typically developed participants studied a list of 42 words presented at 4 sec intervals. Unbeknownst to the participant, word lists were constructed to comprise 4 exemplars of six semantic categories, 2 exemplars of another six and 1 exemplar of another six categories. In words were chosen so that they could also be grouped on orthographic/phonological grounds (e.g. words starting with a silent ‘k’ etc.). Again, either 4, 2 or 1 exemplars of each of these orthographic/phonological categories was presented. The size of these groupings was independent of the size of the semantic categories.

After study, participants were presented with a list of semantic and orthographic/phonological category labels (e.g. Fruits, Words containing an x, etc.) and asked to estimate how many (if any) of the study words were part of that category. Responses between 0 and 9 (inclusive) were allowed. Following this estimation task the same category labels were re-presented with instructions to try to recall relevant words from the list.

Results: Both groups were significantly quicker (F(1,38) = 4.85; p <.05) and more accurate (F(3,114) = 3.73, p<.05) at estimating the number of exemplars from semantic categories in the study list as compared to the number of orthographic/phonological category exemplars. Both groups also demonstrated superior cued recall for the semantic as compared to the orthographic/phonological category exemplars (F(1,38) = 38.30; p<.001). There were no group differences or interactions between group and any of our experimental factors.

Conclusions: Our results suggest that during unguided encoding of verbal material individuals with ASD are no more likely than typically developing comparison individuals to encode low-level perceptual features of stimuli – or at least those features manipulated here. Nor were they any less likely to encode high-level, categorical information. It may therefore be the case that although individuals with ASD process perceptual information more effectively when instructed to do so, the are not spontaneously biased to processing perceptual over conceptual information.