Cue quality and criterion setting in recognition memory

Previous studies on how people set and modify decision criteria in old-new recognition tasks (in which they have to decide whether or not a stimulus was seen in a study phase) have almost exclusively focused on properties of the study items, such as presentation frequency or study list length. In contrast, in the three studies reported here, we manipulated the quality of the test cues in a scene-recognition task, either by degrading through Gaussian blurring (Experiment 1) or by limiting presentation duration (Experiment 2 and 3). In Experiments 1 and 2, degradation of the test cue led to worse old-new discrimination. Most importantly, however, participants were more liberal in their responses to degraded cues (i.e., more likely to call the cue “old”), demonstrating strong within-list, item-by-item, criterion shifts. This liberal response bias toward degraded stimuli came at the cost of increasing the false alarm rate while maintaining a constant hit rate. Experiment 3 replicated Experiment 2 with additional stimulus types (words and faces) but did not provide accuracy feedback to participants. The criterion shifts in Experiment 3 were smaller in magnitude than Experiments 1 and 2 and varied in consistency across stimulus type, suggesting, in line with previous studies, that feedback is important for participants to shift their criteria.     

Strength-based criterion shifts in recognition memory

In manipulations of stimulus strength between lists, a more lenient signal detection criterion is more frequently applied to a weak than to a strong stimulus class. However, with randomly intermixed weak and strong test probes, such acriterion shift often does not result. A procedure that has yieldeddelay-based within-list criterion shifts was applied to strength manipulations in recognition memory for categorized word lists. When participants made semantic ratings about each stimulus word, strength-based criterion shifts emerged regardless of whether words from pairs of categories were studied in separate blocks (Experiment 1) or in intermixed blocks (Experiment 2). In Experiment 3, the criterion shift persisted under the semantic-rating study task, but not under rote memorization. These findings suggest that continually adjusting the recognition decision criterion is cognitively feasible. They provide a technique for manipulating the criterion shift, and they identify competing theoretical accounts of these effects.     

Adaptive changes of response criterion in recognition memory

An experiment was conducted to investigate people’s ability to vary a response criterion strategically, in a recognition memory task, as a function of the length of time given to process the test stimuli (from 100 to 1,500 msec). The experiment used the response signal procedure, in which the participants responded after a signal that came at a variable time delay from stimulus onset. The proportion of new versus old test items was varied systematically with the time of the response signal, with the proportion of new test items rising, falling, or staying constant at later signals. It was found that the participants’ response biases changed adaptively, becoming more conservative at later signals in the rising condition, becoming less conservative in the falling condition, and not changing significantly in the constant condition. Theoretical and methodological implications for recognition memory research are discussed.     

Changing the criterion for memory conformity in free recall and recognition

People's responses during memory studies are affected by what other people say. This memory conformity effect has been shown in both free recall and recognition. Here we examine whether accurate, inaccurate, and suggested answers are affected similarly when the response criterion is varied. In the first study, participants saw four pictures of detailed scenes and then discussed the content of these scenes with another participant who saw the same scenes, but with a couple of details changed. Participants were either told to recall everything they could and not to worry about making mistakes (lenient), or only to recall items if they were sure that they were accurate (strict). The strict instructions reduced the amount of inaccurate information reported that the other person suggested, but also reduced the number of accurate details recalled. In the second study, participants were shown a large set of faces and then their memory recognition was tested with a confederate on these and fillers. Here also, the criterion manipulation shifted both accurate and inaccurate responses, and those suggested by the confederate. The results are largely consistent with a shift in response criterion affecting accurate, inaccurate, and suggested information. In addition we varied the level of secrecy in the participants’ responses. The effects of secrecy were complex and depended on the level of response criterion. Implications for interviewing eyewitnesses and line-ups are discussed.     

Changing the criterion for memory conformity in free recall and recognition

People's responses during memory studies are affected by what other people say. This memory conformity effect has been shown in both free recall and recognition. Here we examine whether accurate, inaccurate, and suggested answers are affected similarly when the response criterion is varied. In the first study, participants saw four pictures of detailed scenes and then discussed the content of these scenes with another participant who saw the same scenes, but with a couple of details changed. Participants were either told to recall everything they could and not to worry about making mistakes (lenient), or only to recall items if they were sure that they were accurate (strict). The strict instructions reduced the amount of inaccurate information reported that the other person suggested, but also reduced the number of accurate details recalled. In the second study, participants were shown a large set of faces and then their memory recognition was tested with a confederate on these and fillers. Here also, the criterion manipulation shifted both accurate and inaccurate responses, and those suggested by the confederate. The results are largely consistent with a shift in response criterion affecting accurate, inaccurate, and suggested information. In addition we varied the level of secrecy in the participants' responses. The effects of secrecy were complex and depended on the level of response criterion. Implications for interviewing eyewitnesses and line-ups are discussed.     

Detection and criterion change associated with different test contexts in recognition memory

Three groups of physics Ss and three groups of arts Ss viewed 20 physics words (PWs) and 20 common words (CWs) in a randomly ordered inspection sequence. Ss were then assigned to one of three recognition test conditions where the inspection items were embedded in different sequences of 120 additional words (60 PWs + 60 CWs; 100 PWs + 20 CWs, 20 pWs + 100 CWs). Signal-detection and Luce analyses examined the effects of the different test contexts on Ss′ true sensitivity and response bias. While all Ss found detection of PWs easier than CWs, these judgments varied according to the properties of the recognition sequences. This latter finding was interpreted as indicating that Ss adopted a conservative and probabilistic strategy for the recognition task as a whole rather than utilizing different strategies for PWs and the CWs where these would result in superior overall performance.     

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.     

Strength cues and blocking at test promote reliable within-list criterion shifts in recognition memory

In seven experiments, we explored the potential for strength-based, within-list criterion shifts in recognition memory. People studied a mix of target words, some presented four times (strong) and others studied once (weak). In Experiments 1, 2, 4A, and 4B, the test was organized into alternating blocks of 10, 20, or 40 trials. Each block contained lures intermixed with strong targets only or weak targets only. In strength-cued conditions, test probes appeared in a unique font color for strong and weak blocks. In the uncued conditions of Experiments 1 and 2, similar strength blocks were tested, but strength was not cued with font color. False alarms to lures were lower in blocks containing strong target words, as compared with lures in blocks containing weak targets, but only when strength was cued with font color. Providing test feedback in Experiment 2 did not alter these results. In Experiments 3A–3C, test items were presented in a random order (i.e., not blocked by strength). Of these three experiments, only one demonstrated a significant shift even though strength cues were provided. Overall, the criterion shift was larger and more reliable as block size increased, and the shift occurred only when strength was cued with font color. These results clarify the factors that affect participants’ willingness to change their response criterion within a test list.     

Age-related changes in recognition and response criterion

Recognition performance does not usually change along the lifespan, but the response criterion usually does, and in general, it changes from being conservative during youth to being liberal, in old age. The focus of the present study is to analyze the changes that take place, both in discrimination and response criterion, as a result of aging in two recognition tasks: one with neutral images, and the other with faces showing positive and negative emotional expressions. Two groups of participants performed both tasks: young (N = 21; age range, 17-33 years), older (N = 21; age range, 65-91 years). The analyses of several discrimination parameters (d' and probability of recognition) and the response criterion yielded significant age differences. Thus, results indicated that the ability to discriminate of older participants was better than that of younger participants when having to recognize neutral images, and faces with negative emotional expressions. The response criterion of younger participants was always conservative, whereas older participants only showed liberal criteria in front of faces with emotional expressions. In relation to the neutral images, the response criterion of older participants was optimum, because it led to more hits, without increasing the false alarms. The results are partially explained by the tasks differential difficulty, and are discussed within the frame of Simulation theory.     

Adaptive criterion setting in perceptual decision making

Pigeons responded in a perceptual categorization task with six different stimuli (shades of gray), three of which were to be classified as "light" or "dark", respectively. Reinforcement probability for correct responses was varied from 0.2 to 0.6 across blocks of sessions and was unequal for correct light and dark responses. Introduction of a new reinforcement contingency resulted in a biphasic process of adjustment: First, choices were strongly biased towards the favored alternative, which was followed by a shift of preference back towards unbiased choice allocation. The data are well described by a signal detection model in which adjustment to a change in reinforcement contingency is modeled as the change of a criterion along a decision axis with fixed stimulus distributions. Moreover, the model shows that pigeons, after an initial overadjustment, distribute their responses almost optimally, although the overall benefit from doing so is extremely small. The strong and swift effect of minute changes in overall reinforcement probability precludes a choice strategy directly maximizing expected value, contrary to the assumption of signal detection theory. Instead, the rapid adjustments observed can be explained by a model in which reinforcement probabilities for each action, contingent on perceived stimulus intensity, determine choice allocation.     

On the measurement of criterion noise in signal detection theory: the case of recognition memory

Traditional approaches within the framework of signal detection theory (SDT; Green & Swets, 1966), especially in the field of recognition memory, assume that the positioning of response criteria is not a noisy process. Recent work (Benjamin, Diaz, & Wee, 2009; Mueller & Weidemann, 2008) has challenged this assumption, arguing not only for the existence of criterion noise but also for its large magnitude and substantive contribution to individuals' performance. A review of these recent approaches for the measurement of criterion noise in SDT identifies several shortcomings and confoundings. A reanalysis of Benjamin et al.'s (2009) data sets as well as the results from a new experimental method indicate that the different forms of criterion noise proposed in the recognition memory literature are of very low magnitudes, and they do not provide a significant improvement over the account already given by traditional SDT without criterion noise.     

On the dynamic nature of response criterion in recognition memory: effects of base rate, awareness, and feedback

The authors examined whether participants can shift their criterion for recognition decisions in response to the probability that an item was previously studied. Participants in 3 experiments were given recognition tests in which the probability that an item was studied was correlated with its location during the test. Results from all 3 experiments indicated that participants' response criteria were sensitive to the probability that an item was previously studied and that shifts in criterion were robust. In addition, awareness of the bases for criterion shifts and feedback on performance were key factors contributing to the observed shifts in decision criteria. These data suggest that decision processes can operate in a dynamic fashion, shifting from item to item.     

Group collaboration in recognition memory

Group collaboration was examined in item and associative recognition. The present study distinguishes between group effects versus collaborative processes and defines the latter as interactive information exchange among group members. By that definition, many group effects do not involve collaboration. For example, group performance can exceed individual performance by pooling the increased resources of the group. Specifically, a group advantage can be obtained by deferring to a majority vote or to the group's best member. For both item and associative recognition, a group advantage was obtained that could not be accounted for by resource pooling. Collaborative facilitation was shown reliably in recognizing targets but not for rejecting distractors.     

Response reversals in recognition memory

Using a dynamic sequential sampling model and a recently proposed model for confidence judgments in recognition memory (T. Van Zandt, 2000b), the authors examine the tendency for rememberers to reverse their responses after a primary decision. In 4 experiments, speeded "old"-"new" decisions were made under bias followed by a 2nd response', either a confidence judgment or another simple choice. The data from these experiments showed that participants know when they have made a mistake and that they respond to this knowledge by reversing their responses. Response reversals are thus shown to be important for constructing models of the response-selection process in recognition memory.     

Age differences in recognition memory

It is generally believed that short-term memory (STM) performance decreases from maturity to old age, but the present paper questions the generality of this conclusion. The first experiment reported is a short-term recognition probe study—no significant age differences were found in either acquisition or the rate of forgetting. It is suggested that age decrements in STM are limited to situations where attention is divided at input or where the material must be manipulated during storage. Experiment II confirmed previous findings of minimal age losses in secondary memory word recognition. The results of this second study support the conclusion that there is poorer accessibility to stored verbal material with advancing age.     

Event-related potentials and recognition memory

According to dual-process models, recognition memory is supported by distinct retrieval processes known as familiarity and recollection. Important evidence supporting the dual-process framework has come from studies using event-related brain potentials (ERPs). These studies have identified two topographically distinct ERP correlates of recognition memory--the "parietal" and "mid-frontal" old/new effects--that are dissociated by variables that selectively modulate recollection and familiarity, respectively. We evaluate the extent to which ERP data support dual-process models in light of the proposal that recollection is a continuous rather than a discrete memory process. We also examine the claim that the putative ERP index of familiarity is a reflection of implicit rather than explicit memory. We conclude that ERP findings continue to offer strong support for the dual-process perspective.