Using response time distributions and race models to characterize primacy and recency effects in free recall initiation

Memory models have typically characterized retrieval in free recall as multi-alternative decision making. However, the majority of these models have only been applied to mean response times (RTs), and have not accounted for the complete RT distributions. We show that RT distributions carry diagnostic information about how items enter into competition for recall, and how that competition impacts on the dynamics of recall. We jointly fit RT distributions and serial position functions of free recall initiation with both a racing diffusion model and the linear ballistic accumulator (LBA: Brown & Heathcote, 2008) model in a hierarchical Bayesian framework while factorially varying different assumptions of how primacy and recency are generated. Recency was either a power law or an exponential function. Primacy was treated either as a strength boost to the early list items so that both primacy and recency items jointly compete to be retrieved, a rehearsal process whereby the first item is sometimes rehearsed to the end of the list to make it functionally recent, or due to reinstatement of the start of the list. While serial position curves do not distinguish between these accounts, they make distinct predictions about how RT distributions vary across serial positions. Results from a number of datasets strongly favor the reinstatement account of primacy with an exponential recency function. These results suggest that models of free recall can be more constrained by considering complete RT distributions.

Three Semantic Serial Position Functions at the Same Time

Serial position functions are so ubiquitous that researchers frequently use buffer items to control for primacy and recency effects regardless of the memory task. However, most theories offer different explanations for different types of tests. In contrast, the relative distinctiveness principle offers one explanation for all tasks: items with fewer close neighbors will generally be more distinct and therefore better remembered than items with more close neighbors. An experiment assessed two predictions of this account. (1) When undergraduates place seven US states in three different orders (by area, year of statehood, and population), serial position functions and error gradients will be observed that resemble those observed in episodic tasks. (2) States that are accurately placed in order because they are an early or late item on one dimension will be placed in order far less accurately when they become mid-list items on a different dimension. The results confirm both predictions.

Serial Position Functions for Recognition of Olfactory Stimuli

Two experiments examined item recognition memory for sequentially presented odours. Following a sequence of six odours participants were immediately presented with a series of two-alternative forced-choice (2AFC) test odours. The test pairs were presented in either the same order as learning or the reverse order of learning. Method of testing was either blocked (Experiment 1) or mixed (Experiment 2). Both experiments demonstrated extended recency, with an absence of primacy, for the reverse testing procedure. In contrast, the forward testing procedure revealed a null effect of serial position. The finding of extended recency is inconsistent with the single-item recency predicted by the two-component duplex theory (Phillips & Christie, 1977). We offer an alternative account of the data in which recognition accuracy is better accommodated by the cumulative number of items presented between item learning and item test.