Binary ROCs in perception and recognition memory are curved

In recognition memory, a classic finding is that receiver operating characteristics (ROCs) are curvilinear. This has been taken to support the fundamental assumptions of signal detection theory (SDT) over discrete-state models such as the double high-threshold model (2HTM), which predicts linear ROCs. Recently, however, Br?der and Schütz (2009) challenged this argument by noting that most of the data on which support for SDT is based have involved confidence ratings. The authors argued that certain types of rating scale usage may result in curved ROCs even if the generating process is thresholded in nature. From this point of view, only ROCs constructed via experimental bias manipulations are useful for discriminating between the models. Br?der and Schütz conducted a meta-analysis and new experiments that compared SDT and the 2HTM using binary (yes-no) ROCs and found that many of these functions were linear, supporting 2HTM over SDT. We examine all the data reported by Br?der and Schütz, noting important limitations in their methodology, analyses, and conclusions. We report a new meta-analysis and 2 new experiments to examine the issue more closely while avoiding the limitations of Br?der and Schütz's study. These new data indicate that binary ROCs are curved in recognition, consistent with previous findings in perception and reasoning. Our results support classic arguments in favor of SDT and indicate that curvature in ratings ROCs is not task specific. We recommend the ratings procedure and suggest that analyses based on threshold models be treated with caution.     

ROCs in recognition with and without identification

Receiver operating characteristics (ROCs) for the recognition of unidentified word fragments were analysed and compared with ROCs for the recognition of identified word fragments. In all conditions reported here, the z-transformed ROCs (z-ROCs) for recognition without identification were linear, with slopes near 1.0. In contrast, the z-ROCs for recognition with identification each had both a linear and a quadratic component, and a slope of substantially less than 1.0. Because the recognition of unidentified word fragments likely results from a familiarity process, the present study examined how well the effect can be explained by global matching variants of signal detection theory. Global matching models predict that memory should be better for strong items from a mixed list than for strong items from a pure list, and should be worse for weak items from a mixed list than for weak items from a pure list; however, only the latter pattern was shown for the recognition of unidentified items.     

ROCs in recognition with and without identification

Receiver operating characteristics (ROCs) for the recognition of unidentified word fragments were analysed and compared with ROCs for the recognition of identified word fragments. In all conditions reported here, the z-transformed ROCs (z-ROCs) for recognition without identification were linear, with slopes near 1.0. In contrast, the z-ROCs for recognition with identification each had both a linear and a quadratic component, and a slope of substantially less than 1.0. Because the recognition of unidentified word fragments likely results from a familiarity process, the present study examined how well the effect can be explained by global matching variants of signal detection theory. Global matching models predict that memory should be better for strong items from a mixed list than for strong items from a pure list, and should be worse for weak items from a mixed list than for weak items from a pure list; however, only the latter pattern was shown for the recognition of unidentified items.     

Item recognition memory and the receiver operating characteristic

Four experiments were conducted to investigate the effects of study time, study repetition, semantic and orthographic similarity, and category length on item recognition memory receiver operating characteristics (ROCs). Analyses of ROC shape rejected A. P. Yonelinas's (1994) dual-process model. The normal unequal variance signal-detection model provided a better account of the data, except for a small but consistent excess of high-confidence errors. It was found that z-transformed ROC slope was increased by similarity, category length, and study item repetition, rejecting R. Ratcliff, G. McKoon, and M. Tindall's (1994) "constancy-of-slopes" generalization for these variables, but slope was relatively unaffected by massed study time.     

Familiarity supports associative recognition memory for face stimuli that can be unitised: Evidence from receiver operating characteristics

According to dual-process models, associative recognition memory mainly relies on recollection without benefiting from familiarity. This study investigates the circumstances under which familiarity may support associative recognition judgements by comparing recognition memory for arbitrarily paired items (i.e., pairs of face stimuli depicting two different persons; interitem associations) with recognition memory for pairs of items that are highly overlapping and can be unitised into a coherent whole (i.e., pairs of physically different but very similar face stimuli depicting the same person; intraitem associations). Estimates of familiarity and recollection were derived from receiver operating characteristics. Consistent with the hypothesis that familiarity is able to support associative recognition memory, but only when the to-be-associated stimuli can be unitised, results from two experiments revealed higher familiarity estimates for intra- compared to interitem associations. By contrast, recollection for recombined pairs was higher for inter- compared to intraitem associations. We propose a hypothetical model on how intraitem associations may benefit from familiarity.     

Word frequency and receiver operating characteristic curves in recognition memory: evidence for a dual-process interpretation

Dual-process models of the word-frequency mirror effect posit that low-frequency words are recollected more often than high-frequency words, producing the hit rate differences in the word-frequency effect, whereas high-frequency words are more familiar, producing the false-alarm-rate differences. In this pair of experiments, the authors demonstrate that the analysis of receiver operating characteristic (ROC) curves provides critical information in support of this interpretation. Specifically, when participants were required to discriminate between studied nouns and their plurality reversed complements, the ROC curve was accurately described by a threshold model that is consistent with recollection-based recognition. Further, the plurality discrimination ROC curves showed characteristics consistent with the interpretation that participants recollected low-frequency items more than high-frequency items.     

Recollection is a continuous process: Evidence from plurality memory receiver operating characteristics

Is recollection a continuous/graded process or a threshold/all-or-none process? Receiver operating characteristic (ROC) analysis can answer this question as the continuous model and the threshold model predict curved and linear recollection ROCs, respectively. As memory for plurality, an item's previous singular or plural form, is assumed to rely on recollection, the nature of recollection can be investigated by evaluating plurality memory ROCs. The present study consisted of four experiments. During encoding, words (singular or plural) or objects (single/singular or duplicate/plural) were presented. During retrieval, old items with the same plurality or different plurality were presented. For each item, participants made a confidence rating ranging from “very sure old”, which was correct for same plurality items, to “very sure new”, which was correct for different plurality items. Each plurality memory ROC was the proportion of same versus different plurality items classified as “old” (i.e., hits versus false alarms). Chi-squared analysis revealed that all of the plurality memory ROCs were adequately fit by the continuous unequal variance model, whereas none of the ROCs were adequately fit by the two-high threshold model. These plurality memory ROC results indicate recollection is a continuous process, which complements previous source memory and associative memory ROC findings.     

Source ROCs are (typically) curvilinear: comment on Yonelinas (1999)

On the basis of his assumption that recollection is a threshold process, A. P. Yonelinas (1999) predicted linear source-identification receiver operating characteristics (ROCs) and recently reported data that were consistent with this prediction. In this article, the authors present data showing curvilinear source-identification ROCs across various encoding and test conditions. On the basis of the source-monitoring framework (e.g., M. K. Johnson, S. Hashtroudi, & D. S. Lindsay, 1993), the authors argue that curvilinearity of source-identification ROCs is a result of differences in the qualitative characteristics of memories rather than simply the influence of undifferentiated familiarity as the dual-process model might suggest.     

Unraveling the contributions of the diencephalon to recognition memory: a review

Both clinical investigations and studies with animals reveal nuclei within the diencephalon that are vital for recognition memory (the judgment of prior occurrence). This review seeks to identify these nuclei and to consider why they might be important for recognition memory. Despite the lack of clinical cases with circumscribed pathology within the diencephalon and apparent species differences, convergent evidence from a variety of sources implicates a subgroup of medial diencephalic nuclei. It is supposed that the key functional interactions of this subgroup of diencephalic nuclei are with the medial temporal lobe, the prefrontal cortex, and with cingulate regions. In addition, some of the clinical evidence most readily supports dual-process models of recognition, which assume two independent cognitive processes (recollective-based and familiarity-based) that combine to direct recognition judgments. From this array of information a "multi-effect multi-nuclei" model is proposed, in which the mammillary bodies and the anterior thalamic nuclei are of preeminent importance for recollective-based recognition. The medial dorsal thalamic nucleus is thought to contribute to familiarity-based recognition, but this nucleus, along with various midline and intralaminar thalamic nuclei, is also assumed to have broader, indirect effects upon both recollective-based and familiarity-based recognition.     

Support for a continuous (single-process) model of recognition memory and source memory

Does memory retrieval occur in a continuous or an all-or-none manner? The shape of the receiver operating characteristic (ROC) has been used to answer this question, with curvilinear and linear memory ROCs indicating continuous and all-or-none retrieval processes, respectively. Signal detection models (e.g., the unequal variance model) correspond to a continuous retrieval process, whereas threshold models (including the multinomial model and the recollection component of the dual-process model) correspond to an all-or-none process. In studies of source memory, Slotnick et al. (2000) and others have observed curvilinear ROCs (supporting the unequal variance model), whereas Yonelinas (1999) observed linear ROCs (supporting the dual-process model). We resolve these seemingly inconsistent results, showing that source memory ROCs are naturally curvilinear but can appear linear when nondiagnostic source information is included in the analysis. Furthermore, the unequal variance model accounted for both recognition memory and source memory ROCs, supporting a continuous process of memory retrieval.     

Classification and recognition in artificial grammar learning: Analysis of receiver operating characteristics

In two experiments we investigated recognition and classification judgements using an artificial grammar learning paradigm. In Experiment 1, when only new test items had to be judged, analysis of z-transformed receiver operating characteristics (z-ROCs) revealed no differences between classification and recognition. In Experiment 2, where we included old test items, z-ROCs in the two tasks differed, suggesting that judgements relied on different types of information. The results are interpreted in terms of heuristics that people use when making classification and recognition judgements.     

Classification and recognition in artificial grammar learning: Analysis of receiver operating characteristics

In two experiments we investigated recognition and classification judgements using an artificial grammar learning paradigm. In Experiment 1, when only new test items had to be judged, analysis of z-transformed receiver operating characteristics (z-ROCs) revealed no differences between classification and recognition. In Experiment 2, where we included old test items, z-ROCs in the two tasks differed, suggesting that judgements relied on different types of information. The results are interpreted in terms of heuristics that people use when making classification and recognition judgements.     

Receiver operating characteristics from nonhuman animals: Some implications and directions for research with humans

Reviews of signal detection have largely overlooked the research involving nonhuman animal subjects. Some of this research is presented and reanalyzed here. Plots of receiver operating characteristics show that human and nonhuman signal-detection performance is very similar. The studies emphasize the series of discriminations that comprise signal-detection tasks and illustrate the systematic effects of different methods of arranging payoffs or feedback, of the consistency of that feedback, and of the physical disparity between response alternatives. The data provide some support for recent theoretical accounts that favor a criterion location measure of isobias over the likelihood ratio, but they also suggest that more systematic work is required in this area. Overall, this research supports many contemporary views concerning signal detection, and it provides an alternative way of looking at some recurrent issues. It also suggests that extensions of signal detection to analyze data from other research paradigms require some caution, and it offers directions for complementary research with human subjects.     

From humans to rats and back again: bridging the divide between human and animal studies of recognition memory with receiver operating characteristics

Receiver operating characteristics (ROCs) have been used extensively to study the processes underlying human recognition memory, and this method has recently been applied in studies of rats. However, the extent to which the results from human and animal studies converge is neither entirely clear, nor is it known how the different methods used to obtain ROCs in different species impact the results. A recent study used a response bias ROC manipulation with rats and demonstrated that speeding memory responses reduced the contribution of recollection, not familiarity. The current study confirms this finding in humans using a comparable response bias method. Moreover, a comparison of the response bias methods commonly used in animal studies and the confidence rating method typically employed in human studies produced similar ROC functions. The present results suggest that the analysis of recognition memory ROCs provides a fruitful method to bridge the human and animal memory literatures.     

Recognition memory ROCs for item and associative information: The contribution of recollection and familiarity

Receiver-operating characteristics (ROCs) were examined in three recognition memory experiments. ROCs for item information (i.e., was this word presented?) were found to be curvilinear. However, ROCs for associative information (i.e., were these two words presented together?) were found to be linear. The results are in agreement with the predictions of a dual-process model that assumes that recognition judgments are -based on familiarity and recollection. Familiarity reflects the assessment of a continuous strength dimension and is well described as a signal detection process, whereas recollection reflects the retrieval of qualitative information about the study episode and behaves like a discrete threshold process. The results showed that memory judgments about items relied on a combination of recollection and familiarity, but that judgments about associations relied primarily on recollection. Further examination of the associative ROCs suggested that subjects were able to recollect that old pairs of items were in the study list, and, under some conditions, that new pairs were not in the study list.     

Human recognition memory: a cognitive neuroscience perspective

For many years the cognitive processes underlying recognition memory have been the subject of considerable interest in experimental psychology. To account for a broad range of behavioral findings, psychologists have put forward a variety of 'dual-process' models, all of which propose that recognition memory is supported by two forms of memory - familiarity and recollection - that differ in their speed of operation and the specificity of the retrieved information. More recently, the dual-process framework has been extended to encompass findings from studies investigating the neural basis of recognition memory. Results from neuropsychological, ERP and functional neuroimaging studies can be accommodated within the framework, and suggest that familiarity and recollection are supported by distinct neural mechanisms.