Familiarity and categorization processes in memory search

A fundamental distinction in tasks of memory search is whether items receive varied mappings (targets and distractors switch roles across trials) or consistent mappings (targets and distractors never switch roles). The type of mapping often produces markedly different performance patterns, but formal memory-based models that account quantitatively for detailed aspects of the results have not yet been developed and evaluated. Experiments were conducted to test a modern exemplar-retrieval model on its ability to account for memory-search performance involving a wide range of memory-set sizes in both varied-mapping (VM) and consistent-mapping (CM) probe-recognition tasks. The model formalized the idea that both familiarity-based and categorization-based processes operate. The model was required to fit detailed response-time (RT) distributions of individual, highly practiced subjects. A key manipulation involved the repetition of negative probes across trials. This manipulation produced a dramatic dissociation: False-alarm rates increased and correct-rejection RTs got longer in VM, but not in CM. The qualitative pattern of results and modeling analyses provided evidence for a strong form of categorization-based processing in CM, in which observers made use of the membership of negative probes in the “new” category to make old–new recognition decisions.     

Effect of delay on recognition decisions: Evidence for a criterion shift

Recent evidence indicates that in intermixed recognition testing of different stimulus classes, people can apply different decision criteria (a criterion shift) to stimulus classes distinguished by the study-test delay (Singer, Gagnon, & Richards, 2002), but not by a conspicuous strength manipulation (Stretch & Wixted, 1998b). In an attempt to reconcile these differences, we applied Singer et al.'s text retrieval method to word recognition. People first studied blocked items from each of five categories. After a delay, five new category lists were presented. After each one, the participants recognized intermixed targets and distractors from the current category and one of the earlier ones. At delays of up to 40 min, the answering criteria for immediate and delayed categories were indistinguishable. At delays of 2 days, in contrast, however, both yes-no and confidence-rating data indicated that more lenient criteria were applied to delayed than to immediate test items. This suggests that people can use the delay between study and test to flexibly adjust the decision criteria of word recognition.     

Low target prevalence is a stubborn source of errors in visual search tasks

In visual search tasks, observers look for targets in displays containing distractors. Likelihood that targets will be missed varies with target prevalence, the frequency with which targets are presented across trials. Miss error rates are much higher at low target prevalence (1%-2%) than at high prevalence (50%). Unfortunately, low prevalence is characteristic of important search tasks such as airport security and medical screening where miss errors are dangerous. A series of experiments show this prevalence effect is very robust. In signal detection terms, the prevalence effect can be explained as a criterion shift and not a change in sensitivity. Several efforts to induce observers to adopt a better criterion fail. However, a regime of brief retraining periods with high prevalence and full feedback allows observers to hold a good criterion during periods of low prevalence with no feedback.     

Application of a Multidimensional Nested Logit Model to Multiple-Choice Test Items

Nested logit models have been presented as an alternative to multinomial logistic models for multiple-choice test items (Suh and Bolt in Psychometrika 75:454–473, 2010) and possess a mathematical structure that naturally lends itself to evaluating the incremental information provided by attending to distractor selection in scoring. One potential concern in attending to distractors is the possibility that distractor selection reflects a different trait/ability than that underlying the correct response. This paper illustrates a multidimensional extension of a nested logit item response model that can be used to evaluate such distinctions and also defines a new framework for incorporating collateral information from distractor selection when differences exist. The approach is demonstrated in application to questions faced by a university testing center over whether to incorporate distractor selection into the scoring of its multiple-choice tests. Several empirical examples are presented.     

On the nature of the decision axis in signal-detection-based models of recognition memory

Most models of recognition memory involve a signal-detection component in which a criterion is placed along a decision axis. Older models generally assume a familiarity-decision axis, but newer models often assume a likelihood ratio axis instead because it allows for a more natural account of the ubiquitous mirror effect. In 3 experiments reported here, item strength was differentially manipulated to see whether a mirror effect would occur. Within a list, the items from 1 category were strengthened by repetition, but the items from another category were not. On the subsequent recognition test, the hit rate was higher for the strong category, but the false-alarm rates for the weak and strong categories were the same (i.e., no mirror effect was observed). This result suggests that the decision axis represents a familiarity scale and that participants adopt a single decision criterion that they maintain throughout the recognition test.     

An object-based cost of visual filtering

Although evidence for object-based attention has been reported in a variety of paradigms, few studies have examined directly the relationship between efficiency in the processing of targets and the number of intervening distractors. In five experiments, observers judged whether the vertices of two relevant shapes were of the same height. Experiments 1 and 2 manipulated observers' perceptual set so that identical stimulus displays were perceived as containing either intervening or flanking distractors. The observers were faster when the distractors were flanking rather than intervening between the targets. Experiments 3-5 varied the number of intervening distractors directly. The observers' response latencies correlated positively with the distractor set-size. Because the distractors were highly discriminable from the targets and the spatial separation between the targets and their interactions with the adjacent distractors were held constant, it was unlikely that the differential reaction times across the conditions were caused by lateral inhibition or response competitions from the distractors. The results suggest the existence of an object-based filtering cost. The implications of the present data for attentional selection over noncontiguous regions are also discussed.     

Individual differences in shifting decision criterion: A recognition memory study

An ability to flexibly shift a decision criterion can be advantageous. For example, a known change in the base rate of targets and distractors on a recognition memory test will lead optimal decision makers to shift their criterion accordingly. In the present study, 95 individuals participated in two recognition memory tests that included periodic changes in the base rate probability that the test stimulus had been presented during the study session. The results reveal a wide variability in the tendency to shift decision criterion in response to this probability information, with some appropriately shifting and others not shifting at all. However, participants were highly reliable in their tendency to shift criterion across tests. The goal of the present study was to explain what factors account for these individual differences. To accomplish this, over 50 variables were assessed for each individual (e.g., personality, cognitive style, state of mind). Using a regression model that incorporated different sets of factors, over 50?% of the variance was accounted for. The results of the analysis describe the total, direct, and mediating effects on criterion shifting from factors that include memory strength, strategy, and inherent characteristics such as a fun-seeking personality, a negative affect, and military rank. The results are discussed with respect to understanding why participants rarely chose an optimal decision-making strategy and provide greater insight into the underlying mechanisms of recognition memory.     

Spreading suppression and the guidance of search by movement: Evidence from negative color carry-over effects

A growing number of studies have shown that significant impairments to search and selection can occur if the target item carries a feature of the irrelevant distractors currently being ignored Braithwaite, Humphreys, and Hodsoll (Journal of Experimental Psychology: Human Perception and Performance, 29, 758–778, 2003). However, these effects have been documented only when search has been extended over time (i.e., in preview search), and not in standard search displays with simultaneously presented items. Here, we present the first evidence that similar costs to selection can occur in simultaneous displays under appropriate circumstances. In the present experiment, participants searched a display for a moving target letter among static and moving distractors. Search efficiency was significantly enhanced for a moving target when half of the letters moved (and half remained static), allowing the static items to be excluded from search. However, if the moving target then shared its color with the irrelevant static items, significant costs emerged, relative to baselines. These results are consistent with the involvement of a general feature-based suppression mechanism in selection, operating over space as well as time.     

Evaluating the performance of different procedures for constructing confidence intervals for coefficient alpha: A simulation study

Reliability is one of the most important aspects of testing in educational and psychological measurement. The construction of confidence intervals for reliability coefficients has important implications for evaluating the accuracy of the sample estimate of reliability and for comparing different tests, scoring rubrics, or training procedures for raters or observers. The present simulation study evaluated and compared various parametric and non-parametric methods for constructing confidence intervals of coefficient alpha. Six factors were manipulated: number of items, number of subjects, population coefficient alpha, deviation from essentially parallel condition, item response distribution and type. The coverage and width of different confidence intervals were compared across simulation conditions.     

Repetition and laterality effects on recognition memory for words and pictures*

Recognition memory for a list of words was tested by presenting a series of items with Ss instructed to make positive responses to targets (list items) and negative responses to distractors (nonlist items). The test items were either words or pictures, and they were presented tachistoscopically either to the left or right visual field. The results showed mean response latencies to be generally faster for stimuli presented to the right visual field. Response times were faster for target and distractor stimuli on their second test presentations than on initial tests, but this effect was much larger for targets. Repetitions were shown to decrease the amount of time necessary to execute the stimulus encoding and initial retrieval stages of recognition. This was also true, although to a lesser extent, if different stimulus forms (words or pictures) were used on the two tests. Subsequent recognition stages, including memory search and decision processes. were apparently independent of test stimulus form.     

Measuring efficiency of selection from briefly exposed visual displays: a model for partial report

In the proposed model for partial report, performance reflects the number of targets in a short-term memory buffer. The total number of items (targets, distractors, or extraneous noise) entering the buffer is independent of the number of targets and distractors in the stimulus. Entrance is determined by selective sampling according to a Luce (1959) ratio rule. The model was tested in a variety of conditions with partial reports based on brightness, color, shape, or alphanumeric class. With three parameters, the model accounted for 99% of the variance with number of targets and number of distractors in data obtained by averaging across conditions. Parameter K (number of items entering the buffer) showed little variation with the selection criterion, and estimates for parameter epsilon (total impact of extraneous noise with impact per target as the unit) were rather small. Estimates for parameter alpha (impact per distractor with impact per target as the unit) varied widely across conditions. Parameter alpha is a measure for the efficiency of selecting targets rather than distractors.     

Spreading the sparing: against a limited-capacity account of the attentional blink

The identification of the second of two targets presented in close succession is often impaired—a phenomenon referred to as the attentional blink. Extending earlier work (Di Lollo, Kawahara, Ghorashi, and Enns, in Psychological Research 69:191–200, 2005), the present study shows that increasing the number of targets in the stream can lead to remarkable improvements as long as there are no intervening distractors. In addition, items may even recover from an already induced blink whenever they are preceded by another target. It is shown that limited memory resources contribute to overall performance, but independent of the attentional blink. The findings argue against a limited-capacity account of the blink and suggest a strong role for attentional control processes that may be overzealously applied.     

Rejecting salient distractors: Generalization from experience

Distraction impairs performance of many important, everyday tasks. Attentional control limits distraction by preferentially selecting important items for limited-capacity cognitive operations. Research in attentional control has typically investigated the degree to which selection of items is stimulus-driven versus goal-driven. Recent work finds that when observers initially learn a task, the selection is based on stimulus-driven factors, but through experience, goal-driven factors have an increasing influence. The modulation of selection by goals has been studied within the paradigm of learned distractor rejection, in which experience over a sequence of trials enables individuals eventually to ignore a perceptually salient distractor. The experiments presented examine whether observers can generalize learned distractor rejection to novel distractors. Observers searched for a target and ignored a salient color-singleton distractor that appeared in half of the trials. In Experiment 1, observers who learned distractor rejection in a variable environment rejected a novel distractor more effectively than observers who learned distractor rejection in a less variable, homogeneous environment, demonstrating that variable, heterogeneous stimulus environments encourage generalizable learned distractor rejection. Experiments 2 and 3 investigated the time course of learned distractor rejection across the experiment and found that after experiencing four color-singleton distractors in different blocks, observers could effectively reject subsequent novel color-singleton distractors. These results suggest that the optimization of attentional control to the task environment can be interpreted as a form of learning, demonstrating experience’s critical role in attentional control.     

Visual and semantic factors in recognition from long-term memory

In recognition tests, physical and semantic relationships between targets and distractors have been shown, in separate manipulations, to affect the latency of subject's decision. Recognition was tested for distractors which were visually similar or dissimilar to targets and which belonged to the target categories or to nontarget categories in order to examine the interaction of these dimensions. Rejection latency was longer for target category than for nontarget category distractors. Latency was also longer for visually similar than visually dissimilar distractors, but only when combined with target category probes. This interaction can be explained by the hypothesis that word recognition depends on the analysis of several dimensions of the probe stimulus, and rejection can occur before all such analyses have been completed.     

Memory strength and the decision process in recognition memory

We investigated the role that memory strength plays in the decision process by examining the extent to which strength is used as a cue to dynamically modify recognition criteria. The study list consisted of strong and weak items, with strength a function of study duration or repetition. The recognition test list was divided into two consecutive blocks; strong items appeared in one block, weak items in the other. If the change in item strength across blocks leads to a shift in criterion, the false alarm rate should change accordingly. In four experiments, the false alarm rates did not change across blocks, even when the difference between the strong and the weak items was magnified and marked with semantic cues. However, the strength of the items in the first test block affected the false alarm rate. Thus, strength cues influence initial criterion placement but fail to induce criterion shifts following permanent and even dramatic changes in item strength. These null findings are contrasted with those in a fifth experiment, in which accuracy feedback produced dynamic criterion shifts.     

A dissociative word-frequency X levels-of-processing interaction in episodic recognition and lexical decision tasks

The effects of levels-of-processing and word frequency were directly compared in three different memory tests. In the episodic recognition test, the subjects decided whether or not a word or a pronounceable nonword had been previously studied. In the two lexical decision tests with either pronounceable or unpronounceable nonwords as distractors, the subjects decided whether a test item was a word or a nonword. There were four main results: (1) in all three tests, reaction times (RTs) in response to studied words were faster if they had received semantic rather than rhyme processing during study; (2) in the episodic recognition test, RTs were faster for low- than for high-frequency words; in both lexical decision tests, RTs were faster for high- than for low-frequency words, though less so when the nonword distractors were unpronounceable; (3) prior study facilitated lexical decisions more in response to low- than to high-frequency words, thereby attenuating the word-frequency effect, but more so when the nonword distractors were pronounceable; (4) in the lexical decision test with pronounceable nonword distractors, relative to prior rhyme processing, prior semantic processing facilitated performance more for high- than for low-frequency words, whereas the opposite was the case in the episodic recognition test. Discussion focused on the relationship of these results to current views of the mechanisms by which (1) word frequency and depth of processing affect performance in implicit and explicit memory tests, and (2) repetition priming attenuates word-frequency effects for lexical decisions.     

The depth of distractor processing in search with clutter

Some search tasks involve looking for a category target in clutter. This is the task faced, for example, by a baggage screener looking for weapons in a suitcase. Such tasks presumably involve the segmentation and recognition of the target object, but it is unknown whether they also involve the segmentation and recognition of the distractor objects. To examine the depth of distractor processing in this task, we had observers search through cluttered displays composed of normal and chimerical distractors. The normal distractors were photographs of recognizable objects, while the chimerical distractors were created by interchanging parts between the normal objects. The obsever's task was to identify the display quadrant that contained an animal or a vehicle target. We varied the difficulty of the search task by varying target and distractor discriminability, target uncertainty, and target occlusion. Only when the target was partially occluded did we find an effect of distractor type. In this case, observers may have found the target through a process of mentally eliminating whole distractor objects. When the target was unoccluded, we found no evidence that observers selected and rejected whole distractors during search. This second result supports our previous claim that often the items for search in clutter are not whole objects.     

Phonological encoding in the attentional blink paradigm

The attentional blink refers to a deficit in reporting a second target that follows a first target within a few hundred milliseconds, when both targets occur in a sequence of distractors shown serially at rates of about 10 items per second. In four experiments, phonological similarity of post-Target 1 distractors impaired dual target report within the interval in which the attentional blink occurs. Similarity of letter targets had a smaller, less reliable effect on performance. Phonological similarity of letter distractors did not affect single target identification (Experiment 3), but it continued to impair dual target report (Experiment 5), even when the targets belonged to a different category—namely, digits. The results demonstrated that, not only targets, but also distractors are encoded phonologically, despite the fact that distractors are irrelevant and never have to be reported.     

Variations in constrained retrieval

Mentally reinstating encoding operations at retrieval might improve access to memories; however, such constrained retrieval is an effortful process that may not always be used. The memory-for-foils procedure (Jacoby, Shimizu, Daniels, & Rhodes, Psychonomic Bulletin & Review 12, 852-857, 2005) infers participant-initiated mental reinstatement of encoding operations during attempts at recognition from the differential memory that accrues to foils during a test of deeply processed items versus during a test of shallowly processed items, as indicated by performance on a final recognition memory test for the foils. Experiment 1 tested whether differential memory for foils is due to the evocation of task context during recollection of neighboring old items. Experiment 2 tested whether inducing a set to respond without much effort on a prior recognition test affects the likelihood of constrained retrieval on later tests. Experiment 3 tested whether constrained retrieval is less likely to occur when the deep versus shallow source of test items is intermixed, rather than blocked in separate tests. These experiments provide evidence that people query memory by mentally reinstating encoding operations and identify conditions that affect the probability of constrained retrieval.     

Paradoxical effects of presentation modality on false memory.

Two experiments are reported in which the effects of presentation modality on false memory in recall and recognition are studied. False recall and recognition of critical targets are lower for non-presented items related to a study list when that study list is presented visually than when presented auditorily. This pattern of low levels of false memory for critical targets holds even when participants read the visually presented study items aloud. These results suggest that recollection of visual detail plays a role in the prevention of false memory. However, both the hit rates (true memory) and the false-alarm rates to weakly related distractors (non-critical targets) were higher for visual presentation than for auditory presentation, suggesting that more than one mechanism may underlie false recognition.