Investigating models of human performance

Abstract

In Decision and Stress, Broadbent (1971) highlighted the deficiencies of steady‐state information‐processing systems and recommended a methodology for studying change and individual variability in human performance. Some implications of his approach are considered here by reviewing studies of the ways in which performance in choice reaction time, visual search and simple memory tasks is affected by individual differences in age and intelligence, by alcohol and by practice. The results are discussed in terms of the successes and limitations of five types of model: (1) ‘box‐and‐arrow’ information flow models, (2) ‘control‐process’ models, (3) algebraic models of decision latencies, (4) ‘single‐factor’ models derived from psychometrics, and (5) connectionist‐network models.