Investigating and explaining the effects of successive relearning on long-term retention.

Abstract

Research on techniques for enhancing long-term retention has focused almost exclusively on single-session learning conditions. However, even the most potent initial learning manipulations typically do not yield retention levels sufficient for successful performance in many real-world contexts. In contrast, <i>successive relearning</i> (i.e., practicing to some level of mastery in multiple learning sessions) is a promising technique for enhancing long-term retention, but surprisingly few studies have been conducted to date. In two experiments, the advantage of successive relearning over single-session learning (i.e., <i>relearning potency</i>) was substantial, ds = 1.52 to 4.19. Additionally, the effects of initial lag on retention were sizable prior to relearning (ds = 1.07 to 1.87) but attenuated after relearning (ds = -.20 to .38; the extent to which relearning attenuates the effects of initial learning conditions is referred to as <i>relearning override</i>). To evaluate a possible explanation of these effects, we formally tested the <i>relearning attenuates</i><i>decay</i> (RAD) model. The RAD model accurately accounted for relearning potency and relearning override effects (and several other secondary effects). We discuss the practical and theoretical implications of these findings and advocate that successive relearning is an important next frontier for educationally relevant memory research. (PsycINFO Database Record