Development and test of a method of discriminating between the contributions of recollection and familiarity to declarative memory in young or learning disabled individuals with ASDs

Background:

We have hypothesised that whereas recollection is selectively impaired in individuals with HFA/AS, both recollection and familiarity are impaired in individuals with LFA, and that this pervasive impairment of declarative memory contributes to impaired language and intellectual disability in LFA.  To test this hypothesis the contributions to declarative memory of recollection and familiarity must be measured separately. However no method exists suitable for use with young children or older learning-disabled individuals.

Objectives:

Our objective was to develop and test such a method.

Methods:

Two groups of 15 young children took part: an HFA/AS group and an age- and language ability-matched TD group. Children were seen individually for a single session.

A forced-choice recognition test was administered first, using 16 non-meaningful 2-dimensional shape stimuli, each paired with 3 similar foils at test. This specific test is known to provide a relatively pure measure of familiarity in amnesic adults.

A cued recall test was administered next, using as cues the last 10 shapes successfully recognised by each child in the previous test, each paired with a stimulus-directed action (e.g., turning the stimulus over; placing a fist on it).  At test, shape-recognition was checked in a forced-choice recognition pre-test using a single novel foil, correct recognition being immediately followed by a test of shape-cued action recall. This constitutes a relatively pure measure of recollection. 

Results:

The groups performed similarly on the difficult shape-recognition test, with no floor or ceiling effects. All but one child performed at ceiling on the easy shape-recognition pre-test. The HFA/AS group was significantly impaired on the action-recall task.

Conclusions:

Children with HFA/AS have a selective impairment of recollection, as predicted. The predominantly nonverbal method we developed is suitable for use with young children and should transfer to the study of memory in older individuals with LFA.