The authors varied the similarity between negative probes and study items in a short-term item-recognition task. Current models treat similarity as a function of the number of occurrences of the probe's features in the study set, a factor that is often confounded with the number of the probe's features occurring in the study set. Unconfounded comparisons showed that performance reflected only the latter factor, with response time a linear function of the number of probe features in the study set. The effect was obtained for both stimuli with manipulated features (colored shapes) and words. Number of presented features is a global property of the study list, but existing global models calculate familiarity by averaging across item matches and cannot readily accommodate the data. The authors proposed that the probe's features are compared with a global representation of the study set's features.

This article challenges the view that the generation effects occurs only for items represented in semantic memory. We obtained generation effects regardless of the lexical status of the cue used to generate the target and regardless of the lexical status of the target. Generation effects were comparable for wordlike and unwordlike nonwords. The effect for nonwords depended on displaying the target at the end of each generate trial. We argue that such feedback gives subjects appropriate visual experience with nonwords; otherwise, the disadvantage of never seeing the generated nonwords could overwhelm any memorial advantage conferred by generating. Generated items also suffer when study and test formats differ more for generated than for read items; we demonstrated that changing format reduces recognition for nonwords. We conclude that previous failures to demonstrate generation effects with nonwords reflect confoundings with such familiarity factors and that the generation effect does exist for nonwords.

We present an iterative-resonance model for recognition memory. On successive iterations, the probe is compared against a feature-by-feature profile of the study set. Yes decisions depend on the similarity of the probe to the profile; No decisions depend on a count of elements in the probe that are not in the profile. Successive iterations sharpen the evidence, and response latency is a function of the number of iterations needed to obtain a sufficiently clear result. The model successfully simulates classic data as well as recent data problematic for alternate models.

This study investigated short-term memory for odors using a four-alternative, forced-choiced recognition paradigm. Stimuli were the odors of 36 common food substances. Twelve subjects were tested in each offour conditions, which differed in the activity performed during the retention interval. Recognition performance was poorest when subjects free associated to an additional odorant presented during the retention interval. Thus, interference from interpolated events does occur in odor memory. Recognition performance was best when the subjects free associated to the name of the target odorant during the retention interval. Thus, the memory code for odors may incorporate semantic information. Remembering odors appears, therefore, to be governed by the same principles as remembering stimuli in other modalities.

No abstract

We propose that correct rejections are based on information that contradicts the study set, rather than on insufficient familiarity. Using two-dimensional stimuli, we varied the featural overlap between lures and the study set so that one feature of the lure had occurred during study and the second feature of the lure had not occurred. Familiarity varied with the number of times the studied feature had occurred, whereas detectability of the extralist feature varied with the number of studied alternatives on the same dimension. Lures increased in difficulty with familiarity only when familiarity of the studied feature was confounded with the number of alternatives to the extralist feature. Hence, the difficulty in rejecting the lure was controlled by the number of alternatives to the extralist feature, not by the familiarity of the lure. Current theory requires additional representation assumptions and a new comparison process to accommodate these data.

The effects of list organization on recognition decisions were studied in four experiments. Subjects memorized supraspan word lists which had been organized either into arbitrarily determined subsets or into subsets based on taxonomic categories. Subjects responded to a series of yes-no recognition probes for each list. The latency of response (RT) was reduced when a probe came from the same subset as the preceding probe; the relevant subset appears to be activated in the recognition process. RTs did not vary with the number of subsets comprising the study list; this result discredits models that posit scanning of the subsets as units. With categorized lists, RTs for both old and new probes increased with the size of the relevant studied subset. The slope of the RT function for old items was affected by the number of distractors from the studied categories. The overall pattern of results cannot be accommodated by existing search models, but the data may be compatible with a model in which associations to the probe item are examined for evidence of list membership.

In studies of iconic memory using the bar-probe task, subjects see a brief display of target letters and are probed by an arrow to report one of them. According to the classic early-selection account, subjects use the probe to select material for perceptual analysis from a precategorical (iconic) memory, but according to late-selection theories, subjects first identify the letters and then use the probe to select one letter for report from the set of categorized items. Pashler (1984) based his test for the locus of selection on a manipulation of display quality in previewed displays. He presented a target for 200 msec and then added a probe, together with the target, for an additional 150 msec. Reducing the target's stimulus quality increased response latency. If the subjects identified the characters before the probe appeared and then selected an item for report, the clarity of the original array should not have affected response latency. Hence, Pashler concluded that his subjects used the probe to select from a precategorical store (early selection). Pashler's experiment did not force subjects to rely on memory of the target; hence, although his experiment documented a situation in which subjects used early selection, it did not rule out late selection in studies of information persistence. We replicated Pashler's findings and, using his logic, showed that when subjects are forced to rely on memory of the target, they select from a categorized store.