Testing improves performance as well as assesses learning: A review of the testing effect with implications for models of learning.

Taking a test of previously studied material has been shown to improve long-term subsequent test performance in a large variety of well controlled experiments with both human and nonhuman subjects. This phenomenon is called the testing effect. The promise that this benefit has for the field of education has biased research efforts to focus on applied instances of the testing effect relative to efforts to provide detailed accounts of the effect. Moreover, the phenomenon and its theoretical implications have gone largely unacknowledged in the basic associative learning literature, which historically and currently focuses primarily on the role of information processing at the time of acquisition while ignoring the role of processing at the time of testing. Learning is still widely considered to be something that happens during initial training, prior to testing, and tests are viewed as merely assessments of learning. However, the additional processing that occurs during testing has been shown to be relevant for future performance. The present review offers an introduction to the historical development, application, and modern issues regarding the role of testing as a learning opportunity (i.e., the testing effect). We conclude that the testing effect is seen to be sufficiently robust across tasks and parameters to serve as a compelling challenge for theories of learning to address. Our hope is that this review will inspire new research, particularly with nonhuman subjects, aimed at identifying the basic underlying mechanisms which are engaged during retrieval processes and will fuel new thinking about the learning-performance distinction. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

Retroactive interference: Counterconditioning and extinction with and without biologically significant outcomes.

Following cue-outcome (X-O) pairings, 2 procedures that reduce conditioned responses to X are extinction, in which X is presented by itself, and counterconditioning, in which X is paired with a different outcome typically of valence opposite that of training. Although studies with animals have generally found counterconditioning more efficient than extinction in reducing responding, data from humans are less clear. They suggest counterconditioning is more efficient than extinction at interfering with emotional processing, but there is little difference between the two procedures regarding their impact on the verbal assessment of the probability of the outcome given the cue. However, issues of statistical power leave conclusions ambiguous. We compared counterconditioning and extinction in highly powered experiments that exploited a novel procedure. A rapid streamed-trial procedure was used in which participants were asked to rate how likely a target outcome was to accompany a target cue after being exposed to acquisition trials followed by extinction, counterconditioning, or neither. In Experiments 1 and 2, evaluative conditioning was assessed by asking participants to rate the pleasantness of the cues after treatment. These studies found counterconditioning more efficient than extinction at reducing evaluative conditioning but less efficient at decreasing the assessment of the conditional probability of the outcome given the cue. The latter effect was replicated with neutral outcomes in Experiments 3 and 4, but the effect was inverted in Experiment 4 in conditions designed to preclude reinstatement of initial training by the question probing the conditional probability of the outcome given the cue. Effect sizes were small (Cohen's d of 0.2 for effect on evaluative conditioning, Cohen's d of 0.3 for effect on the outcome expectancy). If representative, this poses a serious constraint in terms of statistical power for further investigations of differential efficiency of extinction and counterconditioning in humans. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

Effects on Memory of Early Testing and Accuracy Assessment for Central and Contextual Content.

Memory for an event is influenced by many factors including retention interval, frequency of assessment, and type of information assessed concerning the event. We examined the usefulness of observer memory for contextual information in assessing accuracy of memory for central information. Participants viewed a video of a purse being stolen and were asked questions concerning the perpetrator and surrounding context of the event, including where and when the event occurred and who else was present. Participants tested immediately after seeing the video exhibited better memory than those tested for the first time 48-hour after the event. Additionally, testing immediately after viewing the video reduced forgetting over the 48-hour delay (i.e., early testing attenuated subsequent forgetting). Moreover, memory for the context of the event correlated positively with memory of the central information (i.e., perpetrator), and memory concerning other people at the event tended to have the highest correlation with perpetrator memory.

Visual Gender Cues Guide Crossmodal Selective Attending to a Gender-Congruent Voice During Dichotic Listening.

Visual input of a face appears to influence the ability to selectively attend to one voice over another simultaneous voice. We examined this crossmodal effect, specifically the role face gender may have on selective attention to male and female gendered simultaneous voices. Using a within-subjects design, participants were presented with a dynamic male face, female face, or fixation cross, with each condition being paired with a dichotomous audio stream of male and female voices reciting different lists of concrete nouns. In Experiment 1a, the female voice was played in the right ear and the male voice in the left ear. In Experiment 1b, both voices were played in both ears with differences in volume mimicking the interaural intensity difference between disparately localized voices in naturalistic situations. Free recall of words spoken by the two voices immediately following stimulus presentation served as a proxy measure of attention. In both sections of the experiment, crossmodal congruity of face gender enhanced same-gender word recall. This effect indicates that crossmodal interaction between voices and faces guides auditory attention. The results contribute to our understanding of how humans navigate the crossmodal relationship between voices and faces to direct attention in social interactions such as those in the cocktail party scenario.

Extinction training can make the extinction context a stimulus-specific inhibitor: A potential mechanism of experimental renewal.

Renewal is the recovery of extinguished responding to a conditioned stimulus when testing occurs outside the extinction context. Renewal has been explained as the extinction context becoming a negative occasion setter during extinction. However, other mechanisms may contribute. Two recent studies showed (a) after extinction of a discrete cue, the extinction context can serve as a conditioned inhibitor, and (b) in some circumstances operational extinction of a conditioned inhibitor can reduce inhibition with respect to a transfer excitor while retaining inhibition with respect to the excitor used in inhibitory training. Here we examine the potential contribution of these phenomena to renewal. In the present experiment, all rats received fear-conditioning with a target cue in one context and extinction of that cue in a second context. Then half of the subjects received massive extinction of the extinction context (i.e., 24 h) while the other half received only handling. Finally, some subjects in each condition were tested for responding to the target cue in the extinction context, others in a second familiar context, and yet others in a third transfer context in which another fear cue had been extinguished. The results showed ABC renewal independent of whether subjects had or had not received context extinction. However, transfer of the inhibitory potential of the extinction context was observed only in subjects that did not receive context extinction. These results suggest an extinction context can serve as a stimulus-specific conditioned inhibitor, thereby contributing to renewal by decreasing responding to the target cue in an ABB control condition.

Adaptive Memory: Generality of the Parent Processing Effect and Effects of Biological Relatedness on Recall.

The adaptive memory framework posits that human memory is an evolved cognitive feature, in which stimuli relevant to fitness are better remembered than neutral stimuli. There is now substantial evidence that processing a neutral stimulus in terms of its relevancy to an imagined ancestral survival scenario enhances recall, although there is still disagreement concerning the proximate mechanisms responsible for this effect. Several other mnemonic biases have recently been discovered that similarly appear to reflect evolutionary pressures, including a bias to remember items relevant to an imagined parenting scenario. We tested the generality of this parent processing effect by varying the biological relatedness of the imagined child. We also varied the biological relatedness of a child during an imagined third-person survival processing scenario. Across four experiments, we found evidence that simply altering the described biological relatedness of a child in the parenting scenario and third-person survival processing scenario can affect recall, such that items are better remembered when made relevant to a biological child compared to an adopted child. How these findings inform the general adaptive memory framework is discussed.

Extinction of a Pavlovian-conditioned inhibitor leads to stimulus-specific inhibition.

Conditioned inhibitors have been shown to be largely unaffected by non-reinforced exposure (i.e., extinction treatment). Although excitatory associations are readily diminished by extinction treatment, so-called inhibitory associations appear to be largely immune to them. In two fear-conditioning experiments with rats, it was found that a decrease in inhibitory control can result from a massive number of extinction exposures to the inhibitor. Experiment 1 provided evidence that extinction treatment attenuated negative summation between the potential inhibitor and a transfer excitor. However, the extinction treatment had no influence on responding to the original training compound, indicating that some stimulus-specific inhibitory potential remained even after massive extinction. Experiment 2 indicated that retarded excitatory acquisition to the inhibitory stimulus observed after extinction treatment of the inhibitor is no greater than that following a similar amount of stimulus pre-exposure without prior inhibition training (i.e., latent inhibition). The findings indicate that inhibitory associations can be extinguished with large numbers of extinction trials, but they appear to be much more resistant to extinction than excitatory associations.

Associative structure of conditioned inhibition produced by inhibitory perceptual learning treatment.

Exposure to a set of complex stimuli yields an enhanced ability to discriminate between these stimuli. In previous experimental studies, two distinguishable stimuli, X and A, were each repeatedly paired with a common Stimulus B to create compound Stimuli XB and AB. Prior evidence suggests that unique Features X and A form mutually inhibitory associations. This was evidenced by pairing Feature A with a biologically relevant stimulus (i.e., an unconditioned stimulus [US]) and observing that Stimulus X alone later serves to inhibit anticipatory behaviors for that US. These observations may reflect the mutually inhibitory nature of the two Features X and A. However, by assessing the influence of X on behavior that anticipates the US rather than Feature A, these experiments tested inhibition only indirectly. In the present experiments, a more direct measure of inhibition is proposed and tested with rats. We found evidence of retardation and negative summation of associations between unique Features X and A in their capacity to serve as competing cues during overshadowing treatments. Stimulus X was less susceptible to overshadowing by A (which is indicative of retardation of the establishment of an X-A within-compound association) and was able to suppress overshadowing by A of another stimulus (Y) when X was presented with Y at test (which is indicative of negative summation of the representation of A by X). Thus, XB/AB trials were seen to establish an inhibitory relationship between X and A.

Inhibition and mediated activation between conditioned stimuli: Parallels between perceptual learning and associative conditioning.

This series examines the associative basis of inhibitory perceptual learning. Four experiments demonstrate that inhibitory perceptual learning, like Pavlovian conditioned inhibition, is affected by manipulating the number of training trials. Specifically, many interspersed XB/AB training trials (in which letters represent initially neutral stimuli such as tones, clicks, and flashing lights) followed by A-US pairings caused X to act like a conditioned inhibitor (Experiment 1), which is presumed to suggest that an inhibitory association between conditioned stimuli X and A had been formed (i.e., inhibitory perceptual learning). Conversely, few XB/AB training trials followed by A-US pairings produced conditioned responding to X (Experiment 2), which suggests that an excitatory association between X and A had been formed. Additionally, associations with the common element, B, appear to play an inconsistent role across inhibitory and excitatory perceptual learning situations, as extinction of B attenuated excitatory (Experiment 3) but failed to have an influence after inhibitory (Experiment 4) pretraining. The viability of several different accounts of perceptual learning is discussed in light of these observations. (PsycINFO Database Record

Proactive interference by cues presented without outcomes: Differences in context specificity of latent inhibition and conditioned inhibition.

This report is part of a larger project examining associative interference as a function of the nature of the interfering and target associations. Lick suppression experiments with rats assessed the effects of context shifts on proactive outcome interference by latent inhibition (LI) and Pavlovian conditioned inhibition (CI) treatments on subsequently trained Pavlovian conditioned excitation treatment. LI and CI were trained in Context A during Phase 1, and then excitation treatment was administered in Context B during Phase 2, followed by tests for conditioned excitation in Contexts A, B, or C. Experiment 1 preliminarily established our LI and CI treatments and resulted in equally retarded acquisition of behavioral control when the target cue was subsequently trained as a conditioned excitor and tested in Context A. However, only CI treatment caused the target to pass a summation test for inhibition. Centrally, Experiment 2 consisted of LI and CI treatments in Context A followed by excitatory training in Context B. Testing found low excitatory control by both LI and CI cues in Context A relative to strong excitatory control in Context B, but CI treatment transferred to Context C more strongly than LI treatment. Experiment 3 determined that LI treatment failed to transfer to Context C even when the number of LI trials was greatly increased. Thus, first-learned LI appears to be relatively context specific, whereas first-learned CI generalizes to a neutral context. These observations add to existing evidence that LI and CI treatments result in different types of learning that diverge sharply in transfer to a novel test context.

Sources of maladaptive behavior in 'normal' organisms.

A basic assumption of most researchers is that behavior is generally functional, and indeed, in most instances the function is obvious. But in a number of cases, some behaviors of neurophysiologically 'normal' organisms appear to be maladaptive. Considerable research has been conducted to understand the basis of such behavior as well as how the frequency of such behavior can be reduced. Here we provide a brief panoramic review of the major sources of maladaptive behavior in neurophysiologically 'normal' organisms: a) altered environmental contingencies relative to those faced by ancestral generations in their environment of evolutionary adaptation, b) altered environmental contingencies within the lifespan of the animal, c) linked behaviors in which the dysfunctional behavior is a linked companion of a more valuable beneficial trait, and d) the labeling of some behaviors as 'maladaptive' when more careful examination finds that they provide net benefit. Most of our attention is on the consequences of altered contingencies across and within a generation, with altered contingencies within a generation constituting a form of associative interference. The central issue in these two cases can be framed in terms of insufficient or excessive transfer of training resulting in maladaptive behavior. We discuss the functional basis of successful and unsuccessful near transfer (i.e., stimulus and response generalization) and far transfer (including rule learning and abstraction).

Associative structure of second-order conditioning in humans.

Second-order conditioning (SOC; i.e., conditioned responding to S2 as a result of S1-US pairings followed by S2-S1 pairings) is generally explained by either a direct S2→US association or by an associative chain (i.e., S2→S1→US). Previous research found that differences in responses to S2 after S1 was extinguished often depended on the nature of the S2-S1 pairings (i.e., sequential or simultaneous). In two experiments with human participants, we examined the possibility that such differences result from S1 evoking S2 during extinction of S1 following simultaneous but not sequential S2-S1 pairings. This evocation of S2 by S1 following simultaneous pairings may have paired the evoked representation of S2 with absence of the outcome, thereby facilitating mediated extinction of S2. Using sequential S2-S1 pairings, both Experiments 1 and 2 failed to support this account of how extinction of S1 reduced responding to S2. Experiment 1 found that extinguishing S1 reduced responding to S2, while extinguishing S2 had little effect on responses to S1, although forward evocation of S1 during extinction of S2 paired the evoked representation of S1 with absence of the outcome. In Experiment 2, evocation of S2 during S1 nonreinforced trials was prevented because S2-S1 pairings followed (rather than proceeded) S1-alone exposures. Nevertheless, responding to S2 at test mimicked S1 responding. Responding to S2 was high in the context in which S1 had been reinforced and low in the context in which S1 had been nonreinforced. Collectively, these experiments provide additional support for the associative-chain account of SOC.

Adaptive memory: Is there a reproduction-processing effect?

Like all biological systems, human memory is likely to have been influenced by evolutionary processes, and its abilities have been subjected to selective mechanisms. Consequently, human memory should be primed to better remember information relevant to one's evolutionary fitness. Supporting this view, participants asked to rate words based on their relevance to an imaginary survival situation better recall those words (even the words rated low in relevancy) than the same words rated with respect to non-survival situations. This mnemonic advantage is called the "survival-processing effect," and presumably it was selected for because it contributed to evolutionary fitness. The same reasoning suggests that there should be an advantage for recall of information that has been rated for relevancy to reproduction and/or mate seeking, although little evidence has existed to assess this proposition. We used an experimental design similar to that in the original survival-processing effect study (Nairne, Thompson, & Pandeirada, 2007) and across 3 experiments tested several newly designed scenarios to determine whether a reproduction-processing effect could be found in an ancestral environment, a modern mating environment, and an ancestral environment in which the emphasis was on raising offspring as opposed to finding a mate. Our results replicated the survival-processing effect but provided no evidence of a reproduction-processing effect when the scenario emphasized finding a mate. However, when rating items on their relevancy to raising one's offspring in an ancestral environment, a mnemonic advantage comparable to that of the survival-processing effect was found. (PsycINFO Database Record

Stepping back from 'persistence and relapse' to see the forest: Associative interference.

Historically, there has been considerable interest in a large variety of forms of associative interference. However, various factors including interest in clinical application and perhaps recent funding priorities have resulted in a narrowed focus on one particular instance of interference, extinction, with relative neglect of other types of interference. We have been using the existing literature and conducting new experiments to determine whether there is a consistent set of rules governing the occurrence and persistence of two-phase associative interference across (a) proactive and retroactive interference, (b) cue and outcome interference, (c) the type of training in phase 1 (excitatory, inhibitory, or simple nonreinforcement), and (d) the type of training in phase 2 (excitatory, inhibitory, or simple nonreinforcement). Our hope is that a return to more general questions concerning associative interference might reveal broad truths concerning the nature of forgetting. Identifying global principles of associative interference may also help us better appreciate the nature of extinction, including how it can be enhanced and made more enduring, as well as how it can be minimized and made more fleeting.

Retrieval-induced versus context-induced forgetting: Does retrieval-induced forgetting depend on context shifts?

Retrieval-induced forgetting (RIF) is the observation that retrieval of target information causes forgetting of related nontarget information. A number of accounts of this phenomenon have been proposed, including a context-shift-based account (Jonker, Seli, & Macleod, 2013). This account proposes that RIF occurs as a result of the context shift from study to retrieval practice, provided there is little context shift between retrieval practice and test phases. We tested both claims put forth by this context account. In Experiment 1, we degraded the context shift between study and retrieval practice by implementing a generative study condition that was highly similar to retrieval practice. We observed no degradation of RIF for these generated exemplars relative to a conventional study control. In Experiment 2, we conceptually replicated the finding of RIF following generative study, and tested whether context differences between each of the three phases affected the size of RIF. Our findings were again contrary to the predictions of the context account. Conjointly, the 2 experiments refute arguments about the potential inadequacy of our context shifts that could be used to explain either result alone. Overall, our results are most consistent with an inhibitory account of RIF (e.g., Anderson, 2003).

Comparing the context specificity of extinction and latent inhibition.

Exposure to a cue alone either before (i.e., latent inhibition treatment) or after (i.e., extinction) the cue is paired with an unconditioned stimulus results in attenuated conditioned responding to the cue. Here we report two experiments in which potential parallels between the context specificity of the effects of extinction and latent inhibition treatments were directly compared in a lick suppression preparation with rats. The reversed ordering of conditioning and nonreinforcement in extinction and latent inhibition designs allowed us to examine the effect of training order on the context specificity of what is learned given phasic reinforcement and nonreinforcement of a target cue. Experiment 1 revealed that when conditioned-stimulus (CS) conditioning and CS nonreinforcement were administered in the same context, both extinction and latent inhibition treatments had reduced impacts on test performance, relative to excitatory conditioning when testing occurred outside the treatment context. Similarly, Experiment 2 showed that when conditioning was administered in one context and nonreinforcement was administered in a second context, the effects of both extinction and latent inhibition treatments were attenuated when testing occurred in a neutral context, relative to the context in which the CS was nonreinforced. The observed context specificity of extinction and latent inhibition treatments has been previously reported in both cases, but not in a single experiment under otherwise identical conditions. The results of the two experiments convergently suggest that memory of nonreinforcement becomes context dependent after a cue is both reinforced and nonreinforced, independent of the order of training.

Behavioral techniques for attenuating the expression of fear associations in an animal model of anxiety.

BACKGROUND AND OBJECTIVES: Recent data indicate that extinguished fear often returns when the testing conditions differ from those of treatment. Several manipulations including extensive extinction training, extinction in multiple contexts, and spacing the extinction trials and sessions reduce the return of fear. Moreover, extensive extinction and extinction in multiple contexts summate in reducing return of fear, and the spacing of the extinction trials and the spacing of extinction sessions summate in reducing return of fear. Here we evaluated whether these techniques also attenuate the context specificity of latent inhibition, and whether they summate to further decrease fear responding at test. METHODS: In two experiments, with rats as subjects in a lick suppression preparation, we assessed the effects of massive CS preexposure, CS preexposure in multiple contexts, and of spacing the CS-preexposure trials and sessions, in reducing the context specificity of latent inhibition. RESULTS: Fear responding was attenuated by all four manipulations. Moreover, extensive CS preexposure in multiple contexts, and conjoint spacing of the CS-preexposure trials and sessions, were more effective in reducing the context specificity of latent inhibition than each manipulation alone. LIMITATIONS: Our experimental designs evaluated degrees of context specificity of latent inhibition but omitted groups in which latent inhibition was assessed without a context shift away from the context of latent inhibition treatment. This precluded us from drawing conclusions concerning absolute (as opposed to relative) levels of recovery from latent inhibition. CONCLUSIONS: Techniques effective in decreasing the return of conditioned fear following extinction are also effective in decreasing the context specificity of latent inhibition in an animal model of anxiety. Fear and anxiety disorders might be prevented in anxious human participants with the same techniques used here, but that is still an empirical question.

Trial frequency effects in human temporal bisection: implications for theories of timing.

To contrast the classic version of the Scalar Expectancy Theory (SET) with the Behavioral Economic Model (BEM), we examined the effects of trial frequency on human temporal judgments. Mathematical analysis showed that, in a temporal bisection task, SET predicts that participants should show almost exclusive preference for the response associated with the most frequent duration, whereas BEM predicts that, even though participants will be biased, they will still display temporal control. Participants learned to emit one response (R[S]) after a 1.0-s stimulus and another (R[L]) after a 1.5-s stimulus. Then the effects of varying the frequencies of the 1.0-s and 1.5-s stimuli were assessed. Results were more consistent with BEM than with SET. Overall, this research illustrates how the impact of non-temporal factors on temporal discrimination may help us to contrast associative models such as BEM with cognitive models such as SET. Deciding between these two classes of models has important implications regarding the relations between associative learning and timing. This article is part of a Special Issue entitled: Associative and Temporal Learning.

Failure to observe renewal following retrieval-induced forgetting.

Recent studies have pursued the nature of inhibition observed in retrieval-induced forgetting (RIF) tasks. In a RIF paradigm, participants are trained on category-exemplar pairs in Phase 1. Then, some exemplars from select categories (Rp+ items) receive further practice in Phase 2. At test, impaired recall for non-practiced exemplars of the practiced categories (Rp- items) is observed relative to exemplars from non-practiced categories (Nrp items). This difference constitutes RIF. Prior reports of spontaneous recovery from RIF indicate that RIF represents a lapse rather than a loss of memory. Empirical analogs and theoretical considerations suggest that RIF should also be reversible through a change of context between Phase 2 and testing (i.e., renewal). We conducted two experiments using human participants to evaluate the context dependency of RIF. In both experiments, Phases 1 and 2 occurred in distinctly different contexts with subsequent testing occurring in either the Phase 1 context or the Phase 2 context. RIF was observed in both experiments. Experiment 1 additionally found that the magnitude of RIF was not reduced by testing in the Phase 1 context relative to testing in the Phase 2 context. Experiment 2 further tested context dependency of RIF by (1) increasing the dissimilarity between the two contexts and (2) inserting a retention interval between Phase 2 and test for half of the participants in each test context condition. The data again indicated no effect of the context manipulation. Thus, no renewal from RIF was observed in either experiment; moreover, these null findings were supported by Bayesian analyses. These results are compared with analogous inhibitory processes in the animal memory literature that typically show both physical and temporal context dependency.

Associative structure of integrated temporal relationships.

According to the temporal-coding hypothesis (TCH; Savastano & Miller, Behavioural Processes 44:147-162, 1998), acquired associations include temporal information concerning the interval between the associated elements. Moreover, the TCH posits that subjects can integrate two independently acquired associations that share a common element (e.g., S2-S1 and S1-US), which results in the creation of a third association with its own temporal relationship (S2-US). Some evidence has suggested that such temporal integration occurs at the time of testing (Molet, Miguez, Cham, & Miller, Journal of Experimental Psychology: Animal Behavior Processes 38:369-380, 2012). Here we report two fear-conditioning experiments with rats conducted to identify the associative structure of the integrated temporal relationship. The goal was to distinguish between two possible associative structures that could exist following an initial test on which temporal integration occurs: (1) Conditioned responding to S2 on subsequent tests could be the result of recurring successive activation of two independently learned temporal maps that remain independently stored in memory (i.e., S2-S1 plus S1-US). (2) Temporal integration at the moment of initial testing could result in the formation of a direct S2-US (or S2-response) temporal map. Integration was found to occur at test and to produce a new association that was independent of associations with the common element (S1). However, the associative status of S1 appeared to modulate whether or not the new association with S2 was US-specific (S2-US) or directly activated a fear response (S2-response).