Assessing the development of automatic processing: An application of dual-task and event-related brain potential methodologies

Previous research has found that properties of automatic processing do not always co-occur, suggesting that the acquisition rates may differ. The present study investigated the acquisition rate of several of these properties by employing additive factors logic, dual task methodology, and event-related brain potentials. Seven subjects participated in a ten session experiment in which they performed two tasks, a visual/memory search task and a pursuit step tracking task, both together and separately. RT and P300 latency measures indicated that parallel processing of the display was achieved early in training in the consistent mapping condition. This processing was unaffected by dual task demands. An analysis of RT/P300 ratios suggests that another form of perceptual efficiency was achieved later in practice in both the varied and consistently mapped search tasks. This effect was larger in the consistent mapping condition. Reductions in the slope of the memory set function occurred significantly earlier for P300 latency than for RT, suggesting that the stimulus evaluation processes became automated more rapidly than the response selection components of memory search. Consistent with an analysis of the processing demands of the two tasks, the introduction of the tracking task and an increase in tracking difficulty produced equivalent interference during consistent and varied mapping conditions. Results are discussed in terms of models of skill, skill acquisition and component task automaticity.