Retrieving effectively from source memory: Evidence for differentiation and local matching processes

The ability to distinguish between different explanations of human memory abilities continues to be the subject of many ongoing theoretical debates. These debates attempt to account for a growing corpus of empirical phenomena in item-memory judgments, which include the list strength effect, the strength-based mirror effect, and output interference. One of the main theoretical contenders is the Retrieving Effectively from Memory (REM) model. We show that REM, in its current form, has difficulties in accounting for source-memory judgments - a situation that calls for its revision. We propose an extended REM model that assumes a local-matching process for source judgments alongside source differentiation. We report a first evaluation of this model's predictions using three experiments in which we manipulated the relative source-memory strength of different lists of items. Analogous to item-memory judgments, we observed a null list strength effect and a strength-based mirror effect in the case of source memory. In a second evaluation, which relied on a novel experiment alongside two previously published datasets, we evaluated the model's predictions regarding the manifestation of output interference in item and lack of it in source memory judgments. Our results showed output interference severely affecting the accuracy of item-memory judgments but having a null or negligible impact when it comes to source-memory judgments. Altogether, these results support REM's core notion of differentiation (for both item and source information) as well as the concept of local matching proposed by the present extension.

Similarity leads to correlated processing: A dynamic model of encoding and recognition of episodic associations

We present a model of the encoding of episodic associations between items, extending the dynamic approach to retrieval and decision making of Cox and Shiffrin (2017) to the dynamics of encoding. This model is the first unified account of how similarity affects associative encoding and recognition, including why studied pairs consisting of similar items are easier to recognize, why it is easy to reject novel pairs that recombine items that were studied alongside similar items, and why there is an early bias to falsely recognize novel pairs consisting of similar items that is later suppressed (Dosher, 1984; Dosher & Rosedale, 1991). Items are encoded by sampling features into limited-capacity parallel channels in working memory. Associations are encoded by conjoining features across these channels. Because similar items have common features, their channels are correlated which increases the capacity available to encode associative information. The model additionally accounts for data from a new experiment illustrating the importance of similarity for associative encoding across a variety of stimulus types (objects, words, and abstract forms) and types of similarity (perceptual or conceptual), illustrating the generality of the model. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

Information and processes underlying semantic and episodic memory across tasks, items, and individuals

The development of memory theory has been constrained by a focus on isolated tasks rather than the processes and information that are common to situations in which memory is engaged. We present results from a study in which 453 participants took part in five different memory tasks: single-item recognition, associative recognition, cued recall, free recall, and lexical decision. Using hierarchical Bayesian techniques, we jointly analyzed the correlations between tasks within individuals-reflecting the degree to which tasks rely on shared cognitive processes-and within items-reflecting the degree to which tasks rely on the same information conveyed by the item. Among other things, we find that (a) the processes involved in lexical access and episodic memory are largely separate and rely on different kinds of information, (b) access to lexical memory is driven primarily by perceptual aspects of a word, (c) all episodic memory tasks rely to an extent on a set of shared processes which make use of semantic features to encode both single words and associations between words, and (d) recall involves additional processes likely related to contextual cuing and response production. These results provide a large-scale picture of memory across different tasks which can serve to drive the development of comprehensive theories of memory. (PsycINFO Database Record

Release from output interference in recognition memory: A test of the attention hypothesis

Retrieval results in both costs and benefits to episodic memory. Output interference (OI) refers to the finding that episodic memory accuracy decreases with increasing test trials. Release from OI is the restoration of original accuracy at some point during the test. For example, a release from OI in recognition memory testing occurs when the semantic similarity between stimuli decreases midway through testing, suggesting that item representations stored on early trials cause interference on tests occurring on later trials to the extent that the earlier items share features with the latter items. In two recognition memory experiments, we demonstrate release from OI for words and faces. We also test whether release from OI is the result of interference or is due to a boost in attention caused by reorienting to a novel stimulus type. A test for the foils presented during the initial test list supports the interference account of OI. Implications for models of memory are discussed.

Analysis of acute naproxen administration on memory in young adults: A randomized, double-blind, placebo-controlled study

Nonsteroidal anti-inflammatory drugs work by non-selectively inhibiting cyclooxygenase enzymes. Evidence indicates that metabolites of the cyclooxygenase pathway play a critical role in the process of learning and memory. We evaluated whether acute naproxen treatment impairs short-term working memory, episodic memory, or semantic memory in a young, healthy adult population. Participants received a single dose of placebo or naproxen (750 mg) in random order separated by 7-10 days. Two hours following administration, participants completed five memory tasks. The administration of acute high-dose naproxen had no effect on memory in healthy young adults.

Parallel interactive retrieval of item and associative information from event memory

Memory contains information about individual events (items) and combinations of events (associations). Despite the fundamental importance of this distinction, it remains unclear exactly how these two kinds of information are stored and whether different processes are used to retrieve them. We use both model-independent qualitative properties of response dynamics and quantitative modeling of individuals to address these issues. Item and associative information are not independent and they are retrieved concurrently via interacting processes. During retrieval, matching item and associative information mutually facilitate one another to yield an amplified holistic signal. Modeling of individuals suggests that this kind of facilitation between item and associative retrieval is a ubiquitous feature of human memory.

The response dynamics of recognition memory: Sensitivity and bias

Advances in theories of memory are hampered by insufficient metrics for measuring memory. The goal of this paper is to further the development of model-independent, sensitive empirical measures of the recognition decision process. We evaluate whether metrics from continuous mouse tracking, or response dynamics, uniquely identify response bias and mnemonic evidence, and demonstrate 1 application of these metrics to the strength-based mirror-effect paradigm. In 4 studies, we show that response dynamics can augment our current analytic repertoire in a way that speaks to the psychological mechanisms underlying recognition memory. We manipulated familiarity and response bias via encoding strength and the proportion of targets at test (Experiment 1) and found that the initial degree of deviation of the mouse movement toward a response is a robust indicator of response bias. In order to better isolate measures of memory strength, we next minimized response bias through the use of 2-alternative forced-choice tests (Experiments 2 and 3). Changes in the direction of movement along the x-axis provided an indication of encoding strength. We conclude by applying these metrics to the typical strength-based mirror effect design (Experiment 4) in an attempt to further discriminate between differentiation and criterion-shift accounts. (PsycINFO Database Record

Age and response bias: Evidence from the strength-based mirror effect

Performance in episodic memory is determined both by accurate retrieval from memory and by decision processes. A substantial body of literature suggests slightly poorer episodic memory accuracy for older than younger adults; however, age-related changes in the decision mechanisms in memory have received much less attention. Response bias, the willingness to endorse an item as remembered, is an important decision factor that contributes to episodic memory performance, and therefore understanding age-related changes in response bias is critical to theoretical development. We manipulate list strength in order to investigate two aspects of response bias. First, we evaluate whether criterion placement in episodic memory differs for older and younger adults. Second, we ask whether older adults have the same degree of flexibility to adjust the criterion in response to task demands as younger adults. Participants were tested on weakly and strongly encoded lists where word frequency (Experiment 1) or similarity between targets and foils (Experiment 2) was manipulated. Both older and younger adults had higher hit rates and lower false-alarm rates for strong lists than for weak lists (i.e., a strength-based mirror effect). Older adults were more conservative (less likely to endorse an item as studied) than younger adults, and we found no evidence that older and younger adults differ in their ability to flexibly adjust their criterion based on the demands of the task.

The shape of things to come: evaluating word frequency as a continuous variable in recognition memory

The role of experience in memory, specifically the word frequency (WF) mirror effect showing higher hit rates and lower false alarm rates for low-frequency words, is one of the hallmarks of memory. However, this "regularity of memory" is limited because normative WF has been treated as discrete (low vs. high). We evaluate the extent to which the prototypical WF mirror effect holds when WF is treated as a continuous variable. We find a clear nonmonotonic U-shaped relationship. Hit rates are higher for both low-frequency and high-frequency words. Linear and quadratic regression models were fit to the data at both the item and the participant level, and the quadratic model provided a better fit at both levels. This finding is inconsistent with the empirical and theoretical finding of a mirror effect and requires a novel approach to accounting for the role of experience in episodic memory.

Sources of interference in recognition testing

Recognition memory accuracy is harmed by prior testing (a.k.a., output interference [OI]; Tulving & Arbuckle, 1966). In several experiments, we interpolated various tasks between recognition test trials. The stimuli and the tasks were more similar (lexical decision [LD] of words and nonwords) or less similar (gender identification of male and female faces) to the stimuli and task used in recognition testing. Not only did the similarity between the interpolated and recognition tasks not affect recognition accuracy but performance of the interpolated task caused no interference in subsequent recognition testing. Only the addition of recognition trials caused OI. When we presented a block of LD trials or gender identification trials before the recognition test, a decrease in accuracy was observed in the subsequent recognition tests. These results suggest a distinction between temporal context and task context, such that recognition memory performance is determined by the salience of the context cues, and the use of temporal context cues is associated with OI.

A differentiation account of recognition memory: evidence from fMRI

Differentiation models of recognition memory predict a strength-based mirror effect in the distributions of subjective memory strength. Subjective memory strength should increase for targets and simultaneously decrease for foils following a strongly encoded list compared with a weakly encoded list. An alternative explanation for the strength-based mirror effect is that participants adopt a stricter criterion following a strong list than a weak list. Behavioral experiments support the differentiation account. The purpose of this study was to identify the neural bases for these differences. Encoding strength was manipulated (strong, weak) in a rapid event-related fMRI paradigm. To investigate the effect of retrieval context on foils, foils were presented in test blocks containing strong or weak targets. Imaging analyses identified regions in which activity increased faster for foils tested after a strong list than a weak list. The results are interpreted in support of a differentiation account of memory and are suggestive that the angular gyrus plays a role in evaluating evidence related to the memory decision, even for new items.

A causal contiguity effect that persists across time scales

The contiguity effect refers to the tendency to recall an item from nearby study positions of the just recalled item. Causal models of contiguity suggest that recalled items are used as probes, causing a change in the memory state for subsequent recall attempts. Noncausal models of the contiguity effect assume the memory state is unaffected by recall per se, relying instead on the correlation between the memory states at study and at test to drive contiguity. We examined the contiguity effect in a probed recall task in which the correlation between the study context and the test context was disrupted. After study of several lists of words, participants were given probe words in a random order and were instructed to recall a word from the same list as the probe. The results showed both short-term and long-term contiguity effects. Because study order and test order are uncorrelated, these contiguity effects require a causal contiguity mechanism that operates across time scales.

Overcoming the Negative Consequences of Interference From Recognition Memory Testing

Theories of why humans forget have been challenged by the newly discovered list-length/output-interference paradox, in which—under certain testing conditions—learning is not harmed by the amount of verbal material studied, whereas retrieval of that material becomes more difficult with increases in the number of items tested. The latter finding is known as output interference, and the results of the experiment reported here indicate that a release from output interference is obtained when the nature of the items is changed during testing. Specifically, when participants are asked to recognize items from two categories, output interference is minimized when items from each category are tested separately in large blocks. This finding supports models of forgetting that assume interference arises from information about the to-be-learned material that is stored in memory; in contrast, this finding is difficult to explain using models that assume forgetting is the result only of changing context.

List discrimination in associative recognition and implications for representation

Four experiments tested the predictions made by the model outlined in A. H. Criss and R. M. Shiffrin (2004b). Participants studied 2 successive lists of pairs followed by a recognition memory test for the most recent list. Some items and some pairs were repeated across the 2 lists. Critically, a given item could be repeated in the same or different type of pair. For associative recognition, performance was only affected by repetitions in the same pair type. However, in single-item recognition confusions occurred for both types of repetitions. The results are as predicted and confirm the assumption that different associative representations were stored even when the same token repeated in different pair types, whereas similar item representations were used regardless of pair type.

Interactions between study task, study time, and the low-frequency hit rate advantage in recognition memory

In studies of episodic recognition memory, low-frequency words (LF) have higher hit rates (HR) and lower false alarm rates (FAR) than do high-frequency words (HF), which is known as the mirror pattern. A few findings have suggested that requiring a task at study may reduce or eliminate the LF-HR advantage without altering the LF-FAR effect. Other studies have suggested that the size of the LF-HR advantage interacts with study time. To explore such findings more thoroughly and relate them to theory, the authors conducted 5 experiments, varying study time and study task. The full mirror pattern was found only in 2 cases: the standard condition requiring study for a later memory test and a condition requiring a judgment about unusual letters. The authors explain their findings in terms of the encoding of distinctive features and discuss the implications for current theories of recognition memory and the word frequency effect.

Pairs do not suffer interference from other types of pairs or single items in associative recognition

What is the source of interference on a memory test following study of a list containing different types of pairs? Many current models predict that pairs and singles of all types will jointly interfere and therefore harm memory. Such list length effects have often been observed for lists of a single-item type (e.g., a list of words). Here, we examine interference for lists containing multiple types of pairs (e.g., word-word, face-face, word-face). In three experiments, we manipulate the number of each type on the study list. In associative recognition, discrimination fell as the number of pairs of the same type rose, but the number of pairs of other types had little effect. That is, we found a list length effect within, but not between, classes of stimuli. We highlight the importance of representation and propose alternatives to current model representations that can predict such findings.

Context Noise and Item Noise Jointly Determine Recognition Memory: A Comment on Dennis and Humphreys (2001)

S. Dennis and M. S. Humphreys (see record 2001-17194-007) proposed a model with the strict assumption that recognition memory is not affected by interference from other items. Instead, confusions are due to noise generated by prior contexts in which the test item appeared. This model seems disparate from existing models of recognition memory but is similar in many ways that are not superficially obvious. One difference is the order in which item and context information are used as retrieval cues. A more critical difference is the assertion that only an item's history, and not other items, affects recognition memory. Conceptual arguments along with the results of 2 experiments make a persuasive case that both types of noise affect recognition. To illustrate the approach, the authors fit experimental data with a version of the retrieving effectively from memory model (R. M. Shiffrin & M. Steyvers, 1997) incorporating both sources of noise.