بافر اپیزودیک در کودکان دارای معلولیت ذهنی: مطالعه اکتشافی

Performance on three verbal measures (story recall, paired associated learning, category fluency) designed to assess the integration of long-term semantic and linguistic knowledge, phonological working memory and executive resources within the proposed ‘episodic buffer’ of working memory (Baddeley, 2007) was assessed in children with intellectual disabilities (ID). It was hypothesised that children with ID would show equivalent performance to typically developing children of the same mental age. This prediction was based on the hypothesis that, despite poorer phonological short-term memory than mental age matched peers, those with ID may benefit from more elaborate long-term memory representations, because of greater life experience. Children with ID were as able as mental age matched peers to remember stories, associate pairs of words together and generate appropriate items in a category fluency task. Performance did not, however, reach chronological age level on any of the tasks. The results suggest children with ID perform at mental age level on verbal ‘episodic buffer’ tasks, which require integration of information from difference sources, supporting a ‘delayed’ rather than ‘different’ view of their development.

The episodic buffer of working memory (Baddeley, 2000 and Baddeley, 2007) is proposed as a limited capacity storage system responsible for integrating information from several sources to create a unified memory, sometimes referred to as a single ‘episode’. The episodic buffer does this by “binding” information from the various systems of working memory (e.g. phonological loop, visuospatial sketch pad) and relevant activated long-term semantic and linguistic knowledge, into a coherent whole. The episodic buffer can also act as a “mental modelling space, allowing one to set up representations that might guide future actions” (Baddeley & Wilson, 2002, p. 1738). The current paper is concerned with the integration/binding of linguistic and semantic long-term knowledge with information from the “slave” verbal storage system, the phonological loop, during verbal episodic remembering and thinking tasks. It provides an exploratory investigation of whether the functioning of the episodic buffer in children with intellectual disabilities (IQ below 70, associated adaptive and daily living difficulties) is at a level commensurate with their mental age. Children with ID were compared to comparison groups of typically developing children matched for chronological and mental age. Such groups allow the evaluation of three competing cognitive accounts of ID: (1) according to the ‘developmental’ model, children with mixed aetiology familial ID should obtain mental age-appropriate levels of performance on cognitive tasks, as their cognitive development is delayed rather than different in comparison with typical controls (Zigler, 1969); (2) evidence for below mental age level performance on cognitive tasks supports the ‘difference’ model, that cognitive development in those with ID is different to those with typical development (Ellis, 1969 and Milgram, 1973); and (3) evidence for ‘optimal performance’ is found when levels of performance in those with ID approach chronological age level (Burack & Zigler, 1990).

The functioning of the proposed episodic buffer of working memory has received relatively little attention. One way of assessing its role in integrating activated long-term memory knowledge with the contents of other working memory systems to support online memory and processing, is to examine tasks such as prose recall, paired recall and category generation. The present findings were that children with ID showed an ability to bind together relevant long-term knowledge and information from short-term phonological memory to produce novel episodic memories and search semantic memory, at a level commensurate with their mental, but not their chronological age. These exploratory results supported the ‘developmental’ model of intellectual disability, that cognitive performance in the episodic buffer is delayed rather than different in this population (e.g. Zigler, 1969).