The Role of Task Support in Spatial and Temporal Source Memory of Adults with ASD

Background: Bowler, Gardiner & Berthollier (2004) demonstrated intact recognition by adults with ASD of incidentally-encoded sources of studied words such as location on screen, voice of presentation or actions performed on the words. Recall of source, by contrast was diminished in comparison with that of a matched typical comparison group.  Source recall difficulties are also seen in multi-list recall paradigms such as the California Verbal Learning test where individuals with ASD make more intrusions from earler-studied lists into the recall of later lists (Bennetto et al, 1996), raising the question of whether support at retrieval would facilitate memory for which list a particular word belonged to at study.  We predicted that adults with ASD would benefit from support at test for memory for the location of studied words on the screen but that support for which block of the word belonged to (first, second or third) would be less effective.

Objectives: To compare the effects of task support at test on memory for spatially and temporally-defined source in adults with ASD.

Methods: 18 adults with ASD and 18 verbal ability matched typical adults took part. Participants studied a list of 27 words consisting of three temporally-distinct blocks of 9 words labelled ‘List 1’, ‘List 2’, ‘List 3’.  Within each block, three words were presented at the top, middle or bottom of the computer screen.  Words were presented at a rate of one every 4 seconds with a 6-second pause between each block.  The test consisted of a yes/no recognition procedure where studied words were presented randomly interspersed with lures.  If participants made a ‘yes’ response they were either asked to state where on the screen the word had been presented or in which list it appeared (unsupported trials), or to select from source cues (TOP, MIDDLE, BOTTOM  or LIST 1, LIST 2, LIST 3) presented on the screen (supported trials). Order of supported and unsupported test and spatial and temporal source were systematically varied.

Results: :  Overall recognition memory was similar for both groups, replicating the majority of existing findings on recognition.  The source memory data were analysed using a 2 (Group) x 2 (Temporal/Location) x 2 (Support/No Support) ANOVA. There was no group difference in overall source memory, but memory for temporal source was superior to that of spatial source.  No other main effect was significant.  The only significant interaction was for Group x Source x Support.  This showed a marginally superior effect for support for location source in the ASD group but the reverse in the typical group.  Presence of support had no effect on temporal source memory in either group.

Conclusions: The findings on the effect of support for memory of source location replicates those of Bowler et al. (2004). The lack of an effect of support for temporal source goes against our prediction and may simply reflect a difficulty in experimentally operationalising temporal cues.