Affiliation: College of Arts and Sciences, Department of Psychology and Neuroscience
Prior research on encoding variability has often employed it as an auxiliary concept in an attempt to explain what gives rise to the spacing effect. However, tests of this hypothesis serendipitously revealed that spacing modulates the effects of encoding variability. It often results in superior performance (relative to encoding constancy) at short repetition lags, and this difference dissipates at longer lags. The chunking hypothesis is an extant but scarcely known theory that can account for this frequent pattern in the literature. This theory's core assumption is that encoding variability can enhance memory when something is recognized as a repetition, and then the information from both presentations is chunked into an enriched memory code (all of which is presumably easier at short lags). However, this core assumption about recognizing repetitions (a.k.a., study-phase retrieval) remains untested. The current study tested this assumption (as well as other ones implied by the chunking hypothesis) using methods akin to a continuous recognition memory paradigm. We also employed a metric of chunking to see if details from both presentations of a target stimulus exhibited statistical dependence between each other in ways predicted by the chunking hypothesis. The results largely supported the predictions of the chunking hypothesis, however, some amendments are needed to account for the effects of associative distance and retrieval difficulty. Regardless, the current study helped elucidate when and why encoding variability enhances memory.