Life-span changes in task switching

Theories of cognitive development and aging suggest that age-related changes in one or more general abilities are sufficient to account for changes in cognitive processing across the life-span. Empirical studies demonstrate that specific abilities, in addition to general processing changes, are needed to account for age-related changes during performance of some tasks. Two experiments presented here examine whether a specific task switching ability develops and declines at a different rate than general changes in processing speed and working memory capacity. Results suggest that processing speed alone, working memory alone, or a combination of these general abilities are unable to explain life-span changes in speed of changing tasks. Both experiments examine theoretical task switching abilities which could explain age-related changes in task switching, independent from age-related changes in processing speed or working memory. Experiment 1 examines interactions between two theoretical components of task switching, task set inertia and task set preparation. Task set inertia results suggest that the rate of task set decay slows throughout the life-span, from childhood to old age. This slowing has a relatively small effect on reaction time increases when changing tasks (switch cost). In contrast, the ability to prepare in advance for changes in task set is present throughout the life-span, and is optimal in young adulthood. Providing time to prepare in advance of stimulus presentation has a relatively large effect on decreases in switch cost. Experiment 2 examines interactions between response compatibility and task set inertia using non-overlapping versus overlapping stimulus-response (S-R) representations (univalent versus bivalent tasks, respectively). Only older adults (age 71–85) were affected by increased task set inertia (i.e., buildup of S-R representation strength). All age groups, especially children and older adults, were slowed by the requirement to select between competing S-R mapping rules (i.e., response incompatibility). Taken together, these experiments suggest that age-related changes in a general ability fail to explain fully age-related changes in task switching. In addition to general age-related changes in processing speed, changes in active allocation and passive decay of task sets modulate switch cost across the life-span.