Previous cognitive modelling work has suggested that the decline of short-term memory (STM) capacity is the dominant factor of age-related decline on recall ability. We report the results of a set of experiments investigating in further detail the effect of varying short-term memory capacity on the recall of chess positions using the CHREST cognitive architecture, and demonstrate a good qualitative match for human data. We then use these results to suggest a potential functional reason for the size of STM capacity in humans and go on to explore the relationships between STM capacity, ageing, knowledge, presentation time, and recall ability.
Richard Ll. Smith, Peter C. R. Lane, Fernand Gobet