Encoding-specificity instead of online integration of real-world spatial regularities for objects in working memory

In daily life scenarios, most objects are not independent of each other; rather, they show a high degree of spatial regularity (e.g., beach umbrellas appear above beach chairs, not under them). Previous studies have shown a benefit of spatial regularities in visual working memory (VWM) performance of real-world objects, termed the spatial regularity effect. However, the mechanisms underlying this effect remain unclear. The spatial regularity effect can be explained by an “encoding-specificity” hypothesis or a “perception-alike” hypothesis. The former suggests that spatial regularity will enhance the visual encoding process but will not operate in information integration during VWM maintenance, while the latter suggests that spatial regularity will play a role in both the visual encoding and VWM maintenance processes. We tested these two hypotheses by investigating whether VWM integrates sequentially presented real-world objects by focusing on the existence of the spatial regularity effect. In Experiment 1, we manipulated the presentation (simultaneous vs. sequential) and regularity (with vs. without regularity) of memory arrays among pairs of real-world objects. The spatial regularity of memory objects improved the VWM performance in simultaneous presentation trials, but not in sequential presentation trials. In Experiment 2, we examined whether overburdened memory load hindered the spatial regularity effect in sequential presentation trials. We again found an absence of the spatial regularity effect, regardless of the memory load. These results suggest that participants were unable to integrate real-world objects into pairs based on spatial regularity during the VWM maintenance process. Therefore, the present results support the “encoding-specificity” hypothesis, implying that although the spatial regularity of real-world objects can enhance the efficiency of the encoding process in VWM, VWM cannot exploit spatial regularity to help organize sampled sequential information into meaningful groups.

Perpetrator pose reinstatement during a lineup test increases discrimination accuracy

We examined how encoding view influences the information that is stored in and retrieved from memory during an eyewitness identification task. Participants watched a mock crime and we varied the angle from which they viewed the perpetrator. In Experiment 1, participants (N = 2904) were tested with a static photo lineup; the viewing angle of the lineup members was the same or different from the perpetrator at encoding. In Experiment 2, participants (N = 1430) were tested with a novel interactive lineup in which they could rotate the lineup faces into any angle. In both experiments, discrimination accuracy was greater when the viewing angle at encoding and test matched. Participants reinstated the angle of the interactive faces to match their encoding angle. Our results highlight the importance of encoding specificity for eyewitness identification, and show that people actively seek out information in the testing environment that matches the study environment to aid memory retrieval.

Testing encoding specificity and the diagnostic feature-detection theory of eyewitness identification, with implications for showups, lineups, and partially disguised perpetrators

The diagnostic feature-detection theory (DFT) of eyewitness identification is based on facial information that is diagnostic versus non-diagnostic of suspect guilt. It primarily has been tested by discounting non-diagnostic information at retrieval, typically by surrounding a single suspect showup with good fillers to create a lineup. We tested additional DFT predictions by manipulating the presence of facial information (i.e., the exterior region of the face) at both encoding and retrieval with a large between-subjects factorial design (N = 19,414). In support of DFT and in replication of the literature, lineups yielded higher discriminability than showups. In support of encoding specificity, conditions that matched information between encoding and retrieval were generally superior to mismatch conditions. More importantly, we supported several DFT and encoding specificity predictions not previously tested, including that (a) adding non-diagnostic information will reduce discriminability for showups more so than lineups, and (b) removing diagnostic information will lower discriminability for both showups and lineups. These results have implications for police deciding whether to conduct a showup or a lineup, and when dealing with partially disguised perpetrators (e.g., wearing a hoodie).

Exploring the Encoding Specificity Principle and Context-Dependent Recognition in Virtual Reality

A virtual reality (VR) training system’s effectiveness is determined by how well the knowledge-and skills-gained in the virtual environment transfers to real-world performance. The purpose of this study was to examine the efficacy of virtual reality training by comparing semantic memorization in congruent (e.g., memorization task in VR and recognition task in VR) versus incongruent environments (e.g., memorization task in VR and recognition task in the real word). In the present study, we semi replicated Godden and Baddeley’s 1980 study on context-dependent recognition memory by using a photorealistic virtual reality environment in place of the underwater, scuba environment. Results revealed participants that learned semantic information in the virtual environment performed highly on the memory recognition task in the material, real-world environment (and vice versa). These findings replicate and extend Godden and Baddeley’s original results and provide evidence for the use of VR training to support semantic-based knowledge transfer.

Enabling witnesses to reinstate perpetrator pose during a lineup test increases accuracy

Recently, the field of eyewitness identification has undergone a radical transformation, using signal-detection theory, models, and associated analyses to answer the important applied question of how police should test a witness’s memory of a criminal perpetrator. Here, we used these analytical techniques and the basic science of face memory to examine how encoding view influences the information that is stored in and retrieved from memory during an eyewitness identification task. Participants watched a simulated crime and we varied the angle from which they viewed the perpetrator. In Experiment 1, participants (N=2,904) were tested with a static photo lineup; the viewing angle of the lineup members was the same or different from the perpetrator at encoding. In Experiment 2, participants (N=1,430) were tested with a novel interactive lineup in which they could rotate the lineup faces into any angle. In both experiments, discrimination accuracy was greater when the viewing angle at encoding and test matched. Interestingly, participants also naturally reinstated the angle of the interactive faces to match their encoding angle. Our results highlight the importance of encoding specificity for eyewitness identification, and reveal that people actively seek out information in the testing environment that matches the study environment to aid memory retrieval. Our study demonstrates how memory findings and analytical techniques from the basic science literature can enhance our understanding of applied eyewitness memory processes, and ultimately reduce memory errors that can have devastating consequences in the real world.

Memory-related encoding-specificity paradigm: Experimental application to the exercise domain

The Encoding-Specificity Paradigm indicates that memory recall will be superior when contextual factors are congruent between memory encoding and memory retrieval. However, unlike other contextual conditions (e.g., verbal context, mental operations, global feature context, mood dependency, and physical operations), this paradigm has nearly been ignored in the exercise domain. Thus, the purpose of this study was to examine the Encoding-Specificity Paradigm in the context of exercise and rest conditions. 24 young adults (age: M = 21 years) completed a within-subject, counterbalanced experiment involving four laboratory visits, including 1) R-R (rest-rest) condition, 2) R-E (rest-exercise) condition, 3) E-R (exercise-rest) condition, or 4) E-E (exercise-exercise) condition. The exercise bout included a 15-minute moderate-intensity walk on a treadmill. Memory recall was assessed via a 15 word-list task. Memory recall was greater for R-R (8.71 ± 3.1) versus R-E (7.46 ± 2.8), and similarly, for E-E (8.63 ± 2.7) versus E-R (8.21 ± 2.7). The mean word recall for the congruent and incongruent conditions, respectively, was 8.67 (2.4) and 7.83 (2.4). There was a statistically significant condition effect (F = 5.02; P = .03; partial η² = .18). This experiment provides direct support for the Encoding-Specificity Paradigm in the exercise domain.

Long-Term Memory of Past Events in Great Apes

It has been claimed that the ability to recall personal past events is uniquely human. We review recent evidence that great apes can remember specific events for long periods of time, spanning months and even years, and that such memories can be enhanced by distinctiveness (irrespective of reinforcement) and follow a forgetting curve similar to that in humans. Moreover, recall is enhanced when apes are presented with features that are diagnostic of the event, consistent with notions of encoding specificity and cue overload in human memory. These findings are also consistent with the involuntary retrieval of past events in humans, a mode of remembering that is thought to be less cognitively demanding than voluntary retrieval. Taken together, these findings reveal further similarities between the way humans and animals remember past events and open new avenues of research on long-term memory in nonhuman animals.