Abstract
The retention interval has become a critical factor in explaining inconsistent results in relative differences between anticipation and study-test methods. The retention interval in paired-associate (and verbal discrimination) learning consists of two qualitatively different components: Types I (rest, intercycle) and II (work, intervening events) intervals. To determine relative effects of the two components, six sets of comparisons were made of the two methods; massed practice (omitting Type I intervals) was employed with various paired-associate lists and presentation rates, using 140 college students. In the absence of the Type I interval, both significant and nonsignificant performance differences between the two methods occurred. Thus, performance inconsistency seems not easily explained by Type I interval effects, as e. g., done in the spaced practice view. Specifically, the results from five of the six experiments contradicted the spaced practice view. By contrast, the retention interval hypothesis, based on Type II interval effects, is supported more strongly-and therefore suggests developing a better theory.