Stephen Madigan

Reminiscence and item recovery in free recall

The item recovery or reminiscence component of recall in RTT procedures was investigated in two free recall experiments. In the first, Erdelyi and Becker’s (1974) “hypermnesia” effect was found with pictures as the to-be-remembered material: total amount recalled increased over two successive test trials, and included a large reminiscence effect, with some 27% of previously unrecalled items appearing in the second test. The second experiment, with word lists, showed that the frequency of occurrence of new items was greater following a 12-min separation of two test trials than in two relatively massed tests. This kind of item recovery is relevant to models of output interference and retrieval limitations in free recall, and may be also related to spontaneous recovery effects.

Perfect recall and total forgetting: A problem for models of short-term memory

In two paired-associate probe experiments, the items in the last input position in five-pair lists were recalled perfectly in immediate recall, but practically never in a subsequent delayed recall test. Other items in the list were recalled at intermediate levels in both tests. This “negative recency” effect in delayed recall occurred regardless of whether a pair had or had not been tested in immediate recall. These results, at variance with predictions derived from some current models of memory, suggest that pairs in terminal input positions are deliberately not processed and stored by S s in a manner allowing their later recall from long-term memory.

The serial position curve in immediate serial recall

Two different kinds of serial position curves are produced by immediate serial recall of supraspan lists. One is the regular bow-shaped curve, generally regarded as the standard shape. The other is a curve in which recall decreases from the first to about the third serial position, levels off at the fourth position, and then decreases more sharply at the fifth serial position. In many studies, this marked pattern suggests a discontinuity in the serial position curve. This article reviews studies producing each kind of curve, presents some analyses suggesting that a chunking process is involved in the production of the second kind of curve, and discusses the significance of the problem for studies of serial recall and order information.

Retention of item attributes in free recall

A free recall experiment using mixed-modality lists was conducted to assess retention of information about presentation format of to-be-remembered words. Lists consisted of items presented once only, auditorily or visually, or twice, either in the same modality or once in each modality. Ss were required to remember items only (standard free recall) or to remember items as well as their modality and frequency attributes. With all recall from secondary memory, there were no effects of the modality variable (auditory, visual, or mixed) on recall of repeated items, although Ss were able to retain a considerable amount of information about the modality and frequency characteristics of recalled items. Processing this additional information had the effect of lowering word recall scores relative to the standard free recall test. It is argued that recall of modality attributes is based on some direct representation of this information in memory, and is not solely the result of a “tagging” process.

Negative effects of recency in recall and recognition

Following the immediate probe recall of each of 40 five-pair lists, Ss were retested for retention of previously tested and previously untested pairs by cued recall, noncued recall, or recognition tests of either stimulus or response items. Negative effects of recency were obtained in all conditions: items in the final input position of the originally presented lists were recalled and recognized at significantly lower levels than were items from earlier input positions. Neither the individual items in a pair or any long-term association or unitized form of the items seemed to have been well registered in memory at the time they were presented.

False recognition and word length: A reanalysis of Roediger, Watson, McDermott, and Gallo (2001) and some new data

Roediger, Watson, McDermott, and Gallo (2001) reported a multiple regression analysis of the variables that predicted rates of false recall and recognition across lists in the Deese/Roediger— McDermott paradigm. They concluded that false recollection was predictable from the backward associative strength of critical words and list words and the veridical recall level of list words, with no independent contribution of frequency, concreteness, or length of critical words. A reanalysis of their data shows that critical word length does contribute to false recognition when it is measured relative to the length of other words in the list. Relative word length in the form of an unsignedz-score has a larger correlation with false recognition than any of the variables used by Roediger et al. and is also independently predictive of false recognition. This relationship was confirmed by the results of two recognition experiments in which false positives were significantly less frequent for words of lengths that never occurred in a studied list than for words of lengths that did occur in the study list.

Facilitating word-fragment completion with hidden primes

Two experiments tested the effectiveness of priming on word-fragment completion when the priming operation consisted of the presentation of target words in the instructions for the fragment-completion task. Reliable priming effects were obtained in both experiments. The second experiment also examined the influence of case change of printed materials on the extent of priming effects. No simple outcome occurred, as the effects of case change or constancy were not symmetrical with respect to upper- and lowercase.

Representational storage in picture memory

Just as subjects can be highly accurate in identifying the original sense modality of presentation of words, they are also easily able to remember whether items were originally presented in verbal or pictorial form. This finding suggests some representational storage in terms of symbolic as well as in terms of sensory modality.

Ineffectiveness of visual distinctiveness in enhancing immediate recall

Glenberg (1984) and others have theorized that greater recency effects are obtained with auditory as opposed to visual presentation because of greater temporal distinctiveness of items in auditory sequences. We tested a number of ways of enhancing visual distinctiveness, including the use of color, spatial location, and minimized visual interference. None of the seven experiments provided any evidence of improved recall from enhanced visual distinctiveness. In particular, no increase in recency effects was obtained with increased distinctiveness. Additional analyses of pairwise dependency in recall across serial positions also failed to show any evidence of the near-independence of recall of the terminal item that characterizes recall of auditory sequences. Visual-perceptual distinctiveness does not get mapped in any simple way onto memorial distinctiveness in an immediate-serial-recall task.

Retrieval latency and "at-risk" memories.

Does retrieval latency reflect variations in the strength of associations in episodic memory? In three experiments, subjects were given a single study and test trial on each of five lists of 10 paired associates. Spoken recall latencies were measured. When the subjects were later given a second test, initial recall latency was systematically related tointertest retention—that is, the faster the initial correct recall of a pair, the more likely a pair was to be recalled at the second test. This effect occurred at retention intervals of 5 min, 30 min, and 24 h and was present in the data for individual subjects. The results are consistent with the classical view of latency as a measure of trace strength and stand in sharp contrast with results reported by Benjamin, Bjork, and Schwartz (1998) that showed that fast retrievals from semantic memory were more poorly retained than slower ones.