Posters: Memory II

Socio-cognitive theories of the own-race bias (ORB) propose that reduced recognition of other-race faces results from the failure to attend to individuating information in these faces during encoding. In support of this suggestion, individuating instructions that explicitly inform participants about the ORB and instruct them to pay close attention to other-race faces during learning can attenuate or even eliminate the ORB. In the present experiment, we investigated the effect of individuating instructions and encoding-related event-related potentials (ERPs) that contrast neural activity related to subsequently remembered and forgotten items (ERP Dm effects). In line with a socio-cognitive account, individuating instructions reduced the ORB in recognition memory, suggesting that increased attention to other-race faces can improve recognition. At the same time, individuating instructions increased ERP Dm effects for other-race faces, indicating that successful learning may require additional effort. Therefore, the present results suggest that although instructions to individuate can improve other-race face recognition, additional effort is needed to reduce difficulties resulting from a lack of perceptual expertise. This indicates that compensating for reduced experience with other-race faces is possible to some extent but requires additional resources.

Successful source memory retrieval is thought to rely on preretrieval processes including retrieval orientations (ROs), i.e. the specialized processing of retrieval cues depending on the type of searched-for information in memory. Contrasting event-related potentials (ERPs) on correctly rejected items (CRs) across retrieval tasks typically reveals slow-wave activity with task-specific timing, polarity and topography. However, the circumstances under which the adoption of ROs (as indicated by ERP slow waves) is beneficial for memory performance are not yet clear. One reason for heterogeneous results might be the induction of ROs by the use of memory exclusion tasks, in which participants are required to detect items of one target type (e.g. studied with a color) and reject items of another (nontarget) type (e.g. studied with a function) together with new items. This task enables the selective use of nontarget retrieval, probably reducing differences between retrieval tasks and thus RO. To overcome this limitation, the present research compared two types of source information (i.e. color or social group membership) and participants had to remember details within each source type, thus enforcing the retrieval of target information. Consistent with previous research, a positive frontal ERP component (600 - 800ms post-stimulus) differentiated between CRs in the group task and the color task. Moreover, this RO ERP effect was associated with better source memory performance. These results provide evidence for the beneficial effect of ROs on source memory performance, an effect that might be covered in memory exclusion task due to nontarget retrieval.

In the present study, we investigated behavioral aftereffects of ignoring task-irrelevant speech. Specifically, we were interested in whether participants’ subjective truth ratings would be higher for previously ignored statements than for new statements. To this end, participants performed a visual-verbal serial recall task either in silence or while task-irrelevant statements were presented over headphones. The truth status of these statements had been unknown to most participants in a pretest (e.g., “Singapore is the country with the highest population density.”). After the serial recall task, participants rated the truth of previously ignored and new statements. Previously ignored statements obtained significantly higher subjective truth ratings than new statements, suggesting that ignoring task-irrelevant speech can have profound effects on subsequent behavior.

The ubiquitous access to modern technological tools such as tablets allows the temporal externalization of working memory processes (i.e. cognitive offloading). Whereas using technological tools to offload cognitive processes improves immediate performance on different tasks, little is known about potential long-term consequences of offloading behavior. In the current set of experiments, we asked our participants to first solve a pattern copy task that allows for cognitive offloading and then to complete a subsequent memory test. In Experiment 1 (N=172), we observed a trade-off between immediate and subsequent consequences of offloading behavior. Decreasing costs for cognitive offloading resulted in more offloading and more efficient immediate task performance; however, this came along with less accurate performance in the (unexpected) memory test. In Experiment 2 (N=172), we observed less offloading accompanied by more accurate subsequent memory performance when the participants were aware of the upcoming memory test than when they were not aware of the memory test. In Experiment 3 (N=172), we replicated the detrimental effect of cognitive offloading on subsequent memory performance. Participants who were forced to maximally offload showed a reduced memory accuracy when they were not aware of the memory test. Interestingly, however, the participants were (at least partially) able to compensate the detrimental effects of offloading when they were aware of the memory test. Summarized, our experiments therefore suggest that offloading without the explicit intention to form memory representations diminishes memory performance. Nonetheless, such explicit memory intentions can help to evade the negative consequences of cognitive offloading.

Words that are rated in terms of their relevance for a hypothetical survival scenario tend to be remembered better than words processed in a scenario unrelated to survival. While evolutionary psychology considers this a clear example of adaptive memory, there is an ongoing debate about potential mechanisms that enable the positive effect survival processing has on memory. Recent research could not consistently identify whether the effect diminishes in conditions of high cognitive load. In further attempts to investigate the nature of resource dependency, we conducted three experiments, varying both the scenario in which words were processed (survival vs. non-survival) and the time each word was presented during encoding (5 seconds vs. 2 seconds). Assuming survival processing to be resource dependent, we predicted the reduction or even the disappearance of the survival processing benefit in the conditions with reduced presentation time. However, there was no consistent evidence for or against resource dependency across the experiments. We suggest and discuss further, more in-depth research concerning the influence of encoding time on the amount of the effect.

Immediate serial recall of visually presented verbal items is impaired by task-irrelevant background speech. ‎According to the “changing-state”-account of this “irrelevant speech effect” (ISE), pre-attentive, ‎obligatory processing of changing sound elements in a coherent speech stream results in cues to serial ‎order, which then interfere with the deliberate rehearsal of the serial order of the list items. In the current ‎experiment, we analyse whether the ISE is attenuated when the coherence of the irrelevant speech stream ‎is reduced by different talkers. In part 2 of the experiment, we assess the talker variability effect in serial ‎recall performance for spoken items, which has also been attributed to auditory streaming. We expect a ‎significant correlation between talker variability effects in both paradigms, resulting from individual ‎differences in pre-attentive auditory streaming. Results and implications for theoretical accounts of the ISE ‎and the talker variability effect in serial recall are discussed.‎

Our mind uses priors based on past experiences to predict outcomes and events in reality, thus enabling us to efficiently interact with our environment. Since the requirements of our environment are often changing, priors need to be constantly updated in order to meet new demands. One factor that can enhance these updating processes is prediction error (i.e., the difference between the predicted information and the actual evidence received). The strength of this errors reflects the amount of novel information that was not part of our mental representations of the world before the evidence was encountered. The present study measured the effects of three different levels of prediction error (high, medium and low) on memory performance. In four different phases we separately manipulated the establishment of the priors, the prediction verification as well as the immediate and delayed memory consequences. We were able to measure a significant effect of prediction error on memory performance: the effect of high prediction error exceeds the effect of low prediction error in the delayed memory test (i.e., after consolidation). Unexpectedly, we also measured an marked improvement in memory performance for medium prediction error. These results support the idea that predictive processing has a long-term influence on memory.

Retrieval practice has been shown to enhance the acquisition of new information in future learning opportunities by reducing proactive interference from prior lists. While prior work examined this forward effect of testing (FET) mainly with younger adults, there is almost no evidence of this mnemonic benefit in middle and old adulthood (but see Pastötter & Bäuml, 2019). To examine the FET as a function of age and to investigate semantic activation as a potential operating mechanism of the FET, we let 216 older subjects (30–69 years) study three lists of words in expectation to freely remember them in a final cumulative recall test; in between lists, they either recalled the items of each list (retrieval-practice group), studied them again (restudy group), or semantically judged them (judgement-practice group). After having studied all lists, subjects needed only to recall the last List 3. The results revealed a FET in both middle-aged and older subjects, with the retrieval-practice group recalling more list items and producing fewer intrusions than the restudy group. Critically, the judgement-practice group recalled less than the retrieval-practice group and did not reliably produce more intrusions. Together, these findings highlight retrieval practice as an education tool to reduce age-related memory decrements though they did not provide support for semantic activation as a key operating mechanism.