Positional and temporal clustering in serial order memory

Abstract

The well-known finding that responses in serial recall tend to be clustered around the position of the target item has bolstered positional-coding theories of serial order memory. In the present study, we show that this effect is confounded with another well-known finding—that responses in serial recall tend to also be clustered around the position of the prior recall (temporal clustering). The confound can be alleviated by conditioning each analysis on the positional accuracy of the previously recalled item. The revised analyses show that temporal clustering is much more prevalent in serial recall than is positional clustering. A simple associative chaining model with asymmetric neighboring, remote associations, and a primacy gradient can account for these effects. Using the same parameter values, the model produces reasonable serial position curves and captures the changes in item and order information across study-test trials. In contrast, a prominent positional coding model cannot account for the pattern of clustering uncovered by the new analyses.

Sequential dependencies in recall of sequences: Filling in the blanks

“…These results contrast strikingly with those observed by Solway et al (2012) in the four data sets they examined, where displacements following the first order error tended to peak at negative values. Before attempting to explain the discrepancy in these results, we first report analyses of temporal clustering and fill-in/infill errors.…”

“…However, a recent article by Solway, Murdock, and Kahana (2012) reported striking evidence of sequential dependencies more in line with chaining models. Solway et al applied two analyses that they used as indices of positional clustering and temporal clustering, respectively.…”

“…Positive values represent postponements (i.e., recalling items too late), and negative values represent anticipations (recalling items too early). Following Solway et al (2012), these probabilities were corrected for the number of opportunities to make positional confusions over various distances (e.g., there are many ways to confuse adjacent items, whereas a confusion over the longest possible distance can involve only the first and last items on the list); repetition and extralist intrusion errors were ignored. The pattern in Fig.…”

“…When one attempts to adjudicate between extant models of serial recall, the results of Solway et al (2012) Table 1 for the source of the data illustrated in each panel Henson (1996) and Surprenant et al (2005) speak against chaining as a process supporting serially ordered behavior, the analyses of Solway et al point to chaining as a key mechanism. To address these apparent inconsistencies, we analyzed a number of existing serial recall data sets gathered for other purposes, using the measures employed by Solway et al and Henson (1996).…”

“…This is similar to an infill error in the analyses of Henson (1996) and Surprenant et al (2005) and is consistent with participants navigating a chain of associations. Solway et al (2012) also measured temporal clustering (the tendency for recalled items to cluster around the position of the last item recalled), using the lag conditional response probability (lag-CRP) function, which has been commonly applied to examine sequential dependencies in free recall (e.g., Farrell & Lewandowsky, 2008;Howard & Kahana, 1999;Kahana, 1996;Ward, Tan, & GrenfellEssam, 2010). The lag-CRP function records the probability of recalling an item k as a function of the lag k − j between that item and the item just recalled, j.…”