Do item-dependent context representations underlie serial order in cognition? Commentary on Logan (2021)

Logan (2021) presented an impressive unification of serial order tasks including whole report, typing, and serial recall in the form of the context retrieval and updating (CRU) model. Despite the wide breadth of the model’s coverage, its reliance on encoding and retrieving context representations that consist of the previous items may prevent it from being able to address a number of critical benchmark findings in the serial order literature that have shaped and constrained existing theories. In this commentary, we highlight three major challenges that motivated the development of a rival class of models of serial order, namely positional models. These challenges include the mixed-list phonological similarity effect, the protrusion effect, and interposition errors in temporal grouping. Simulations indicated that CRU can address the mixed list phonological similarity effect if phonological confusions can occur during its output stage, suggesting that the serial position curves from this paradigm do not rule out models that rely on inter-item associations, as has been previously been suggested. The other two challenges are more consequential for the model’s representations, and simulations indicated the model was not able to provide a complete account of them. We highlight and discuss how revisions to CRU’s representations or retrieval mechanisms can address these phenomena and emphasize that a fruitful direction forward would be to either incorporate positional representations or approximate them with its existing representations.

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.

The forward effect of testing: Behavioral evidence for the reset-of-encoding hypothesis using serial position analysis

The forward effect of testing refers to the finding that retrieval practice of previously studied information increases retention of subsequently studied other information. It has recently been hypothesized that the forward effect (partly) reflects the result of a reset-of-encoding (ROE) process. The proposal is that encoding efficacy decreases with an increase in study material, but testing of previously studied information resets the encoding process and makes the encoding of the subsequently studied information as effective as the encoding of the previously studied information. The goal of the present study was to verify the ROE hypothesis on an item level basis. An experiment is reported that examined the effects of testing in comparison to restudy on items’ serial position curves. Participants studied three lists of items in each condition. In the testing condition, participants were tested immediately on non-target lists 1 and 2, whereas in the restudy condition, they restudied list 1 and 2. In both conditions, participants were tested immediately on target list 3. Influences of condition and items’ serial learning position on list 3 recall were analyzed. The results showed the forward effect of testing and furthermore that this effect varies with items’ serial list position. Early target list items at list primacy positions showed a larger enhancement effect than middle and late target list items at non-primacy positions. The results are consistent with the ROE hypothesis on an item level basis. The generalizability of the ROE hypothesis across different experimental tasks, like the list-method directed-forgetting task, is discussed.

Articulatory Rehearsal Is More Than Refreshing Memory Traces

This study examined whether additional articulatory rehearsal induced temporary durability of phonological representations, using a 10-s delayed nonword free recall task. Three experiments demonstrated that cumulative rehearsal between the offset of the last study item and the start of the filled delay (Experiments 1 and 3) and a fixed rehearsal of the immediate item during the subsequent interstimulus interval (Experiments 2 and 3) improved free recall performance. These results suggest that an additional rehearsal helps to stabilize phonological representations for a short period. Furthermore, the analyses of serial position curves suggested that the frequency of the articulation affected the durability of the phonological representation. The significance of these findings as clues of the mechanism maintaining verbal information (i.e., verbal working memory) is discussed.