Aging and automaticity: Evaluation of instance-based and strength-based mechanisms

Experiments evaluated instance-based learning as a possible sole mechanism underlying development of automaticity. Age differences in extended practice lexical decision and factors that could effect age-related performance on lexical decision tasks were also examined. The first experiment was conducted to evaluate the role of stimulus-specific and strategic, task-specific learning underlying performance improvement in a trained lexical decision task. The first experiment consisted of a training and a transfer phase. The training phase assessed age-related differences on a lexical decision task where an attention response could, in principle, be developed (Search condition) and where an attention response could not develop (Nonsearch condition). The transfer phase was conducted to evaluate the role of strategic, task-specific factors on performance improvement observed in the training phase. Age-related differences in word-nonword response time differences were eliminated with practice in the nonsearch, but not the search, version of the task. Transfer tests also implicate strategic differences as a partial source of age differences in lexical decision performance. Experiment 2 was a 10-session (two-week) extended practice study which was conducted to provide a strong test of instance-based learning as a sole mechanism for automaticity. Contrary to predictions of instance theory, the Search and Nonsearch conditions converged for the young adults. Consistent with a strength-based theory of automaticity applied to cognitive aging, the conditions did not converge for older adults. The results provide further support for age differences in automaticity-as-attention training. Based on these and other results, the importance of considering varieties of automaticity in theories of cognitive aging is discussed.