Aspects of remembering: Current topics in experimental memory research

People can purposefully forget information that is irrelevant and no longer needed. In the lab, such forgetting of outdated information is often examined by means of list-method directed forgetting (LMDF). Several accounts have been proposed to explain LMDF, but despite decades of research, there is still no agreement concerning the mechanism(s) mediating it. Here, we used prolonged retention intervals to examine two specific accounts of LMDF, namely mental context change and selective rehearsal. Experiment 1 probed the mental context change account by contrasting LMDF with context-dependent forgetting across two delay intervals. The results showed intact LMDF, but eliminated context-dependent forgetting with longer delay, which is inconsistent with the context-change account. Experiment 2 probed the selective-rehearsal account by contrasting the longevity of LMDF across intentional and incidental encoding. The results showed persistent LMDF for both types of encoding, which indicates that selective rehearsal is not critical for the persistence of LMDF. Together, the findings indicate that context change and selective rehearsal cannot account for persistent LMDF, thus more generally challenging contemporary noninhibitory accounts of LMDF.

People expect to remember words printed in larger fonts better than words printed in smaller fonts, although actual memory is similar for smaller and larger words. Incidental observations raise the possibility that this font size illusion in metamemory may not occur with nonwords. If so, this might provide insights into when and why font size dissociates metamemory and memory and, in addition, point to limitations of current theorizing about metamemory. Three experiments obtained judgments of learning (JOLs) for words (e.g., metal), pseudowords (e.g., unsle), and nonwords (e.g., abrtz) presented in four different font sizes between 9 point and 294 point. Results revealed illusory effects of font size on JOLs for words and pseudowords in all experiments and conditions. Provided that participants attempted to master nonwords, font size also affected JOLs for nonwords. These results demonstrate that font size dissociates metamemory and memory for various types of items and inform our knowledge about metamemory.

The present study was a preregistered replication of the so-called mnemonic time-travel effect. In a recent paper it was claimed that memory improves when people experience backward motion before a memory test in comparison to when they experience forward motion or no motion because the backward motion brings them back to the moment of encoding. In the original study this mnemonic time-travel effect was reported to be robust for different manipulations of backward motion—such as real, simulated, and imagined motion. To test the robustness of the effect, we performed a close replication of the experiment with the largest mnemonic time-travel effect. We found no significant differences among the different motion conditions despite sufficient statistical power to detect an even considerably smaller effect than the one reported before. The present results thus contradict the idea that experiencing backward motion improves memory for past events. Our findings also demonstrate the importance of close replications—especially of seemingly spectacular novel findings such as the one scrutinized here.

Retrieval practice is more effective than restudy practice to potentiate subsequent learning of new information. Here, we examined this forward effect of testing (FET) across adulthood and to test predictions of the episodic context account. Due to the latter, retrieval practice, relative to restudy practice, invites subjects to reinstate the prior context of the initial study, which should help to reduce proactive interference. 108 younger adults (18–39 years) and 108 older adults (55–85 years) studied three separate, new lists of words in expectation to freely recall them in a final cumulative recall test. During the initial study phase of Lists 1–3, subjects were presented with two separate sublists (of Lists 1–3). Then, they were instructed to either recall or restudy each list. Importantly, we let a third group indicate the sublist, which the items occurred in originally. Thus, subjects needed to think back to and reinstate the original temporal context. After the study phase of List 3, all groups were instructed to selectively recall List 3. For both age groups, retrieval practice led to greater List-3-recall performance and fewer prior-list intrusions than restudy practice and temporal discrimination. Critically, practicing temporal discrimination let older adults, but not younger adults, recall more items than restudy practice and also produce less intrusions from prior lists independent of age. Together, these findings suggest that retrieval practice is a robust learning strategy of new information and that the FET (in older adults) can be partly explained via the episodic context account.

Source memory refers to memory for the origin of information, such as who told you something. In a typical source-monitoring task, participants study items (e.g., words) presented by one of two sources (e.g., speakers). At test, the items are presented in a source-neutral manner (e.g., printed, not spoken) and participants have to judge whether this item was presented by Source A, Source B, or is new. Here, a source memory failure may simply be due to difficulties with retrieving stored source details (e.g., perceptual features of the speaking voice) from memory rather than a lack of source storage. In two experiments with student participants (Experiment 1: N = 146, Experiment 2: N = 108), we demonstrate that the reinstatement of the original source at test (i.e., spoken presentation of the test items) facilitates retrieval of stored source details from memory. In Experiment 1, retrieval facilitation from source reinstatement was most effective for difficult to discriminate (i.e., highly similar) sources. In Experiment 2, a partial reinstatement of the speaker’s face and name (but not the voice) was already effective for facilitating source retrieval. In Experiment 3, we tested 72 older adults (≥ 60 years old) with the difficult to discriminate sources and full reinstatement. In contrast to the consistent reinstatement benefit observed in younger adults, older adults only benefitted from source reinstatement for weakly encoded sources. We will discuss implications of our results for research on source memory and the age-related deficit in source memory.

A number of studies have documented older adults' tendency to falsely remember past events that are in part or entirely false. Older adults' increased propensity for false memory has been documented in paradigms including semantic manipulations, across tasks that vary perceptual similarity of test items as well in paradigms relying on source memory. However, little is known about the degree to which the effects of semantic, perceptual or source similarity on false memory are comparable, or whether older adults are particularly vulnerable to some of those factors. We tested this question in older adults from the Berlin Aging Study who completed a verbal learning task, including a recognition test. During the recognition test older adults encountered studied words along with lure words that were either semantically related to the studied words, phonetically similar to those words, or were familiar from a previous task. Initial results suggest that the effects of lure similarity are indeed variable, with larger age-related increases in false alarms for familiar and phonetically similar lures than for semantically related lures. Further analyses demonstrated that individual differences in learning rates play an important role in predicting individual differences in false recognition. Taken together, while older adults show a general tendency for increased false memory, these effects may be rather nuanced and may critically depend on the type of information being tested.