Forgetting from Short-Term Memory in Delayed Matching to Sample: A Reinforcement Context Model

Abstract

Short-term memory in nonhuman animals is typically studied in delayed matching to sample, with variation in the retention interval or delay between the to-be-remembered sample and subsequently presented choice or comparison stimuli. The forgetting function, which relates the systematic decrease in discriminability to increasing delay, is well described by an exponential in the square root of time, with an intercept and slope that vary systematically with different conditions, such as sample-stimulus disparity, retention-interval conditions, and reward parameters. We argue that the rewards for accurate matching are relative to the reinforcement context, which includes rewards R o for extraneous or other behaviors. Forgetting results from competition between R o and rewards for the delayed matching task. We suggest that R o acts to shift attention from the memory task to extraneous behavior, and that R o grows as a linear function of time in the retention interval. By incorporating these assumptions in the model proposed by White and Wixted (1999), we accurately predict the time course of forgetting under a variety of different conditions for delayed matching.

Direct remembering, mediated remembering, and atypical forgetting functions

“…Data from Grant () in which sample duration was varied in a delayed matching‐to‐sample task with predictions from the model of Nevin et al () p. 308, in the left panel, and predictions from the model of White and Brown (), p. 7, in the right panel.…”

“…The linear function is of course the most parsimonious, and in a very different model with “null” memory traces that block recall (Lansdale & Baguley, ), the null traces are assumed to increase as a linear function of time. White and Brown () therefore assumed a linear growth function with an intercept at time zero of R o (0) , and a slope of g . That is, the growth over time t is R o = R o (0) + g · t .…”

“…The source of interference in Nevin et al's () model is disruption of attending. The source of interference in White and Brown's () model is extraneous reinforcement. Interference as an account of forgetting must be favored over trace decay, because the atypical forgetting functions reviewed above suggest a dissociation of forgetting from the passage of time (Laming & Scheiwiller, ; Nairne, ).…”

“…An important difference between the two behavioral models is that the disruption of attending in Nevin et al's () model is specific to samples or comparisons, whereas extraneous reinforcement in White and Brown's () model competes with reinforcement for the delayed matching task as a whole. For White and Brown, delayed matching is a discriminated operant, that is, a behavioral unit.…”

“…Owing to the complexity of the task, interpretation of Wright et al's () result in terms of the behavioral theories considered here is not straightforward. The decrease in accuracy for fourth items with increasing retention interval (a diminishing recency effect) is simply understood in terms of the increasing effect of disruptors, q and v , in Nevin et al's () model or R o in White and Brown's () model. In terms of both theories, at short retention intervals, items later in the list might disrupt attending to the first item “as coded” (Nevin et al, ), or might generate task interference (White & Brown, ).…”