The Role of Decision Processes in Conscious Recollection

Dual-process models of recognition memory posit a rapid retrieval process that produces a general sense of familiarity and a slower retrieval process that produces conscious recollections of prior experience. The remember / know paradigm has been used to study the subjective correlates of these two processes, with remember judgments assumed to index conscious recollection and know judgments assumed to index familiarity. Recently, a two-criterion signal detection model has been proposed as an alternative account of this paradigm. This model assumes only a single memory process with a criterion remember from know responses. This report presents an empirical test of the model's critical prediction that manipulations that influence criterion placement should influence both remember and know judgments. An experiment confirmed this prediction, demonstrating that subjects who were told that 70% of the test items were study items produced more remember and know responses than subjects who were told that 30% of the test items were study items.     

The detection model of recognition using know and remember judgments

The signal detection model forknow andremember recognition judgments was tested in two experiments. In Experiment 1, two predictions of the model were tested: (1) that measures of memory sensitivity,A′, are equivalent in value when based on either the recognition (know or remember) criterion or on the remember criterion; and (2) that there is a positive correlation between recognition bias and the proportion of know judgments that are hits, but no correlation between recognition bias and proportion of remember hits (Donaldson, 1996). Both predictions were supported by the data. In Experiment 2, the context of test items was manipulated to make it more or less similar to learning context. The detection model requires that memory sensitivity be the same for both recognition and remember judgments, regardless of test context. Alternatively, if remember judgments reflect only the retrieval of episodic information from memory, the two measures of memory sensitivity should become more disparate in value as learning and test context are made more similar. Memory sensitivity was generally the same in value for recognition and remember criteria but different across context conditions, thus supporting the detection model. The nature of the memory continuum used in detection theory is also discussed.     

Modeling the conscious correlates of recognition memory: Reflections on the remember-know paradigm

Understanding how memory processes contribute to the conscious experience of memory is central to contemporary cognitive psychology. Recently, many investigators (e.g., Gardiner, 1988) have examined theremember-know paradigm to understand the conscious correlates of recognition memory. A variety of studies have demonstrated that variables have different effects on remember and know responses, and these findings have been interpreted in the context of dual-process models of recognition memory. This paper presents a single-process model of the remember-know paradigm, emphasizing the dependence of remember and know judgments on a set of common underlying processes (e.g., criterion setting). We use this model to demonstrate how a single-process model can give rise to the functional dissociations presented in the remember-know literature. We close by detailing procedures for testing our model and describing how those tests may facilitate the development of dual-process models.     

The role of decision processes in remembering and knowing

Participants in recognition memory studies are now often asked to partition recognized items into ones that are accompanied by some recollective experience (those they remember) and ones that are not so accompanied (but which they know were previously encountered). Rather than detecting separate memory systems, such attempts to distinguish between remembering and knowing are better understood as a division of positive recognition responses into those that lie above a second decision criterion (remember) and those that do not (know). As such, the amount of memory associated with knowing is strongly dependent on the placement of the decision criteria. A meta-analysis of published data and a simple experiment tested predictions from the decision process analysis of remember/know responses.     

A comparison of the multimemory and detection theories of know and remember recognition judgments

Know and remember recognition judgments were made in 2 experiments. In Experiment 1 either 3 (new, know, remember) or 4 (new, guess, know, remember) judgment alternatives were used. Also, the recognition test was either immediate, delayed 2 days, or delayed 1 week. In Experiment 2 the proportion of test items that were studied items was either .20, .50, or .80. Predictions were made from 2 multimemory theories regarding how proportions of know and remember judgments would change with these manipulations. Predictions were also made from 2 detection theories regarding the values of memory sensitivity and judgment criteria. The results supported the revised detection theory, which assumes unequal variances for studied and unstudied items, more than the multimemory theories. Furthermore, general processing tree versions of the 2 multimemory models were created including guessing processes. These extended models were shown to be inadequate in accounting for the data from the 2 experiments.     

Context effects on remembering and knowing: the expectancy heuristic

Three experiments are reported examining the effect of context on remember-know judgments. In Experiments 1 and 2, medium-frequency words were intermixed with high-frequency or low-frequency words at study or at test, respectively. Remember responses were greater for medium-frequency targets when they were studied or tested among high-frequency, as compared with low-frequency, words. The authors proposed a decision-based mechanism called "the expectancy heuristic" to explain why remember responses were more likely when items were studied or tested in the context of words that were relatively less distinct. According to the expectancy heuristic, when items on a recognition test exceed an expected level of memorability they will be given a remember judgment but when they do not, but are still more familiar than new words, they will be given a know judgment. Experiment 3, which varied expectancies about the strength of tested targets, demonstrated the use of the expectancy heuristic, indicating that it operates by selectively influencing the remember criterion rather than by influencing recollection of studied items.     

Cognitive effort and recollective experience in recognition memory

The difficulty of the cognitive operations required to process study items was manipulated in two experiments investigating recollective experience. In subsequent recognition tests, subjects indicated whether their recognition judgements for items processed in these tasks were based on recollection ("remember" responses) or on familiarity ("know" responses). In Experiment 1 target items were presented in the context of a category decision task. It was found that remember responses increased with the difficulty of the category decision. For positive instances, remember responses were greater for items of low instance frequency than for items of high instance frequency, while for negative instances remember responses were greater for items from similar categories than for items from dissimilar categories. These effects were not present in know responses. In Experiment 2, remember responses were more frequent when study items had been presented in the form of anagrams to be solved than when they had been presented in the form of words to be read aloud. The incidence of know responses was not affected by the format in which study items were presented. Source judgements were also more accurate when recognition was based on recollection. It is argued that the type of conscious awareness experienced during recognition is determined by the knowledge activated by items presented in the recognition test, which in turn is determined by the nature of the operations engaged at encoding.     

Familiarity and recollection in item and associative recognition

Recognition memory for item information (single words) and associative information (word pairs) was tested immediately and after retention intervals of 30 min and 1 day (Experiment 1) and 2 days and 7 days (Experiment 2) using Tulving's (1985) remember/know response procedure. Associative recognition decisions were accompanied by more "remember" responses and less "know" responses than item recognition decisions. Overall recognition performance and the proportion of remember responses declined at similar rates for item and associative information. The pattern of results for item recognition was consistent with Donaldson's (1996) single-factor signal detection model of remember/know responses, as comparisons based on A' between overall item recognition and remember item recognition showed no significant differences. For associative recognition, however, A' for remember responses was reliably greater than for overall recognition. The results show that recollection plays a significant role in associative recognition.     

Recognition memory in amnesia: effects of relaxing response criteria

In two experiments, using the remember/know paradigm, we examined whether recognition memory in amnesic patients can be improved by instructing patients to relax their response criterion. Experiment 1 was modeled after a study by Dorfman, Kihlstrom, Cork, and Misiaszek (1995), in which direct instructions to respond more leniently led to an increase in recognition accuracy in patients with ECT-induced amnesia. We failed to extend this finding to patients with global amnesia, but the manipulation was unsuccessful in control subjects as well. In Experiment 2, response criterion was manipulated indirectly by providing information about the alleged base rate of study items on the recognition test. This manipulation led to a criterion shift in control subjects and enhanced discriminability in amnesic patients. Analysis of “remember” and “know” responses suggests that improved accuracy in amnesia was associated with enhanced familiarity-based recognition.     

Measuring the speed of the conscious components of recognition memory: remembering is faster than knowing

Three experiments investigated response times (RTs) for remember and know responses in recognition memory. RTs to remember responses were faster than RTs to know responses, regardless of whether the remember-know decision was preceded by an old/new decision (two-step procedure) or was made without a preceding old/new decision (one-step procedure). The finding of faster RTs for R responses was also found when remember-know decisions were made retrospectively. These findings are inconsistent with dual-process models of recognition memory, which predict that recollection is slower and more effortful than familiarity. Word frequency did not influence RTs, but remember responses were faster for words than for nonwords. We argue that the difference in RTs to remember and know responses reflects the time taken to make old/new decisions on the basis of the type of information activated at test.     

Effects of exact and category repetition in true and false recognition memory

Two experiments used the distinction between remembering and knowing to investigate the effects of exact and category repetition in recognition memory. In Experiment 1, exact repetition enhanced remember responses but had no reliable effect on know responses. In Experiment 2, category repetition enhanced correct know responses but had no effect on correct remember responses. Category repetition also increased false positive remember and know responses. It is argued that exact repetition influences the recollection component of recognition memory via the creation of multiple episodic traces, each of which is potentially capable of supporting a remember response, whereas category repetition influences the familiarity component of recognition memory by enhancing the fluency with which test items are processed.     

Test-induced priming impairs source monitoring accuracy in the DRM procedure

Three experiments investigated the effects of test-induced priming (TIP) on false recognition in the Deese/Roediger-McDermott procedure (Deese, 1959; Roediger & McDermott, 1995). In Experiment 1, TIP significantly increased false recognition for participants who made old/new decisions at test but not for participants who made remember/know judgments or were given diagnostic information to help them avoid false recognition. In Experiment 2, a TIP effect was observed with old/new recognition but not when participants were required to remember whether study items were spoken by a male or a female speaker. In Experiment 3, false recognition increased when critical lures were preceded by 10 studied items but not when preceded by 5 studied and 5 unstudied items from the same list. These findings suggest that TIP increases false recognition by disrupting source monitoring processes.     

Illusions of familiarity caused by cohort activation

Subjects took part in an auditory lexical decision task followed by an auditory test of recognition memory for words presented in this task. Subjects categorized their recognition judgments as based on either recollection (“remember” responses) or familiarity (“know” responses). Distractor items in the recognition test included the base words from which the nonwords used in the lexical decision task were derived. Consistent with the findings of Wallace, Stewart, Sherman, and Mellor (1995), more false alarms were made to “late” base words (where the corresponding nonwords were created by changing a phoneme near the end of the word) than to “early” base words (corresponding nonwords were created by changing a phoneme at the beginning of the word). However, this effect was found in “know” responses but not in “remember” responses. The findings are attributed to enhanced fluency with which the base words are processed following their implicit activation at encoding.     

Sum-difference theory of remembering and knowing: a two-dimensional signal-detection model

In the remember-know paradigm for studying recognition memory, participants distinguish items whose presentations are episodically remembered from those that are merely familiar. A one-dimensional model postulates that remember responses are just high-confidence old judgments, but a meta-analysis of 373 experiments shows that the receiver operating characteristic (ROC) curves predicted by this model have the wrong slope. According to the new sum-difference Theory of remembering and knowing (STREAK), old items differ from new ones in both global and specific memory strength: The old-new judgment is based on a weighted sum of these dimensions, and the remember- know judgment is based on a weighted difference. STREAK accounts for the form of several novel kinds of ROC curves and for existing remember-know and item-recognition data.     

Investigating the attentional demands of recognition memory: Manipulating depth of encoding at study and level of attention at test

Two experiments investigated the effects of divided attention at test after manipulating levels of processing at study. In Experiment 1 items were studied either intact or as anagrams. In Experiment 2 items were studied with either full or divided attention (DA). In both experiments participants carried out a recognition test with either full or divided attention. Analysis of remember and know responses revealed that DA at test had no effect on remember responses to anagrams or to items studied with full attention. In contrast, Know responses decreased with DA at test, and this occurred for items studied in both deep and shallow encoding conditions. The present study confirms recent findings (Knott & Dewhurst, 2007a) that knowing can rely on more controlled retrieval processes, whereas remembering can rely on more automatic retrieval processes. Differences in the controlled processes associated with recollection and familiarity and remember and know responses are discussed.     

Mine to remember: The impact of ownership on recollective experience

Evaluating information with reference to self is associated with enhanced memory, the “self-reference effect”. The effect is found in recognition accompanied by recollective experience (remembering), but not in recognition based on a feeling of knowing. The current research employed an ownership procedure to investigate whether less evaluative forms of self-referential cognition produce similar enhancement of recollective experience. Participants were asked to sort items into baskets that belonged to themselves or a fictitious other. A subsequent remember–know recognition test showed that items encoded in the context of self-ownership were more likely to be correctly recognized than other-owned items. This ownership effect was found in remember, but not know, responses. This finding suggests that creating a self-referential encoding context leads to elaborative representations in episodic memory, even in the absence of explicit self-evaluation.     

Recollection can be weak and familiarity can be strong

The remember-know procedure is widely used to investigate recollection and familiarity in recognition memory, but almost all of the results obtained with that procedure can be readily accommodated by a unidimensional model based on signal-detection theory. The unidimensional model holds that remember judgments reflect strong memories (associated with high confidence, high accuracy, and fast reaction times), whereas know judgments reflect weaker memories (associated with lower confidence, lower accuracy, and slower reaction times). Although this is invariably true on average, a new 2-dimensional account (the continuous dual-process model) suggests that remember judgments made with low confidence should be associated with lower old-new accuracy but higher source accuracy than know judgments made with high confidence. We tested this prediction--and found evidence to support it--using a modified remember-know procedure in which participants were first asked to indicate a degree of recollection-based or familiarity-based confidence for each word presented on a recognition test and were then asked to recollect the color (red or blue) and screen location (top or bottom) associated with the word at study. For familiarity-based decisions, old-new accuracy increased with old-new confidence, but source accuracy did not (suggesting that stronger old-new memory was supported by higher degrees of familiarity). For recollection-based decisions, both old-new accuracy and source accuracy increased with old-new confidence (suggesting that stronger old-new memory was supported by higher degrees of recollection). These findings suggest that recollection and familiarity are continuous processes and that participants can indicate which process mainly contributed to their recognition decisions.     

Dissociating familiarity from recollection using rote rehearsal

Recollection-based recognition memory judgments benefit greatly from effortful elaborative encoding, whereas familiarity-based judgments are much less sensitive to such manipulations. In this study, we have examined whether rote rehearsal under divided attention might produce the opposite dissociation, benefiting familiarity more than recollection. Subjects rehearsed word pairs during the "distractor" phase of a working memory span task, and were then given a surprise memory test for the distractor items at the end of the experiment. Experiment 1 demonstrated that increasing rehearsal elevated the recognition rate for intact and rearranged pairs, but neither associative recognition accuracy nor implicit fragment completion benefited from rehearsal. The results suggest that rote rehearsal leads to a greater increase in familiarity than in recollection, and that the increase in observed familiarity cannot be attributed to effects of repetition priming. In Experiment 2, we tested item recognition with the remember/know procedure, and the results supported the conclusions of Experiment 1. Moreover, a signal detection model of remember/know performance systematically overpredicted rehearsal increases in remember rates, and this worsened when high-rehearsal items were assumed to be more variable in strength. The results suggest that rote rehearsal can dissociate familiarity from recollection at the time of encoding and that item recognition cannot be fully accommodated within a one-dimensional signal detection model.     

Binding of multidimensional context information as a distinctive characteristic of remember judgments

This research investigated the cognitive processes underlying remember-know judgments in terms of contextual binding in multidimensional source memory. Stochastic dependence between the retrieval of different context attributes, which formed the empirical criterion of binding, was observed for remembered items but not for known items. Experiment 1 showed that the qualitative difference in the stochastic relation holds even if quantitative source-memory performance is equated for items with remember and know judgments. Experiment 2 generalized the findings to context information from different modalities, and Experiment 3 ruled out a spurious stochastic dependence due to interindividual differences. Supporting recent dual-process models of remember-know judgments, the findings show that remember and know judgments differ with respect to binding processes that correspond to episodic recollection.     

Test context affects recollection and familiarity ratings: implications for measuring recognition experiences

The binary remember/know task requires participants to dichotomize their subjective recognition experiences into those with recollection and those only with familiarity. Many variables have produced dissociative effects on remember/know judgments. In contrast, having participants make independent recollection/familiarity ratings has consistently produced parallel effects, suggesting the dissociations may be artifacts of using binary judgments. Bodner and Lindsay (2003) reported a test-list context effect with binary judgments: Increased remembering but decreased knowing for a set of critical items tested with a set of less-memorable (vs. more-memorable) items. Here we report a parallel effect of test-list context on recollection and familiarity ratings, induced by a shift in response bias. We argue that independent ratings are preferable to binary judgments because they allow participants to directly report the co-occurrence of recollection and familiarity for each item. Implications for the measurement of self-reported recognition experiences, and for accounts of recognition memory, are discussed.