Reaction time techniques were used to examine the role of attention in the construction and maintenance of expectancies. In Experiment I, a primed letter matching task, expectancies were observed through a delay (cost) in responding to misprimed letter arrays. A secondary probe task was interpolated between prime and array letters on some of the trials, with attentional demands inferred from delayed responding to probes. By varying the amount of time between onset of the prime and either a probe or letter array, it was found that there is attentional involvement (as reflected in probe inhibition) prior to the observation of expectancies (as reflected in letter matching cost). It was also found that the interpolation of an attention-demanding probe task did not entirely disrupt primed expectancies. Experiment II found that an expectancy persists even when an interpolated distractor task signals that the expectancy is no longer valid. These expectancies were found to decay as a function of time. The implications of these results for attention allocation and memory activation views of expectancy were discussed.
