Relying on the external world: Working memory strategies in the context of memory capacity, external demands, and acquired brain injury

Abstract

Working memory is suggested to be an interactive and adaptive system whereby the amount of information we memorize or do not memorize is tuned to environmental demands. When information is readily available, people tend to look back and forth to this information (external sampling) and avoid memorization (internal storing). When information is less readily available—due to delays, physical distance, and alike–people start memorizing more. This thesis approached the trade-off between external sampling and internal storing from a neuropsychological angle: we tested whether and how brain injuries that reduce memory capacity influence the use of external sampling. Memory-impaired individuals were expected to rely more on the external world than those with normal memory function, even when it was relatively difficult to access external information. Patients with Korsakoff amnesia indeed relied more on external sampling than healthy controls. Although patients attempted to adjust their behaviour to changing information availability by decreasing the number of inspections and increasing memorization, they failed to do so as effectively as controls. Our results suggest a persistent reliance on external sampling, even when external information was less accessible. Individuals referred to a memory clinic inspected external information as often as controls but used longer inspections and showed poorer performance. Importantly, higher objective memory capacity correlated with fewer and shorter inspections, indicating that memory potential translated to memory usage. Memory self-efficacy (beliefs about one’s memory abilities) did not impact reliance on the external world. Patients recovering from a cerebrovascular accident (stroke) also showed increased reliance on the external world, with a subset of patients demonstrating an extreme dependence. Notably, this subset also showed abnormal memory capacity. Yet, the inverse was not always true: low memory scores (in either patients or controls) did not necessarily result in extreme reliance on the external world. Therefore, excessive reliance on external sampling in stroke patients cannot be attributed to reduced memory capacity alone. The extent to which one relies on the external world varied at the individual level. Low- and medium-loading strategies were common, but individuals seldom showed a high-loading strategy that approached working memory capacity limits. In an online study, healthy participants flexibly adjusted their strategy to changing task demands without negative effects on performance. In a lab setting, however, larger adjustments to an initial strategy resulted in poorer performance for both healthy adults and stroke patients. For these groups it appeared more beneficial to maintain a stable working memory strategy, even under changing conditions. Taken together, reduced memory capacity increased reliance on the external world, while negative beliefs about memory did not. Thus, memory capacity translates to memory usage in tasks that allow to choose how much to memorize. Importantly, full capacity often remains underutilized. We suggest that heterogeneity in cognitive and motoric profiles, personal characteristics, task-dependent demands, and effort minimization all differentially influence memory usage at any given instance. Neuropsychologists are encouraged to acknowledge these factors when assessing memory function.