Search Abstracts | Symposia | Slide Sessions | Poster Sessions
The role of attention in verbal short-term memory
Poster Session A, Friday, September 12, 11:00 am - 12:30 pm, Field House
Priyanka Shah-Basak1, Sara Pillay2, Nismeta Kabilovic3, Leonardo Fernandino4, Jeffrey Binder5; 1Medical College of Wisconsin
Attention is intimately linked with verbal short-term memory (STM) in temporarily maintaining information either by selecting and prioritizing linguistic representations in the long-term memory itself, or within specialized buffers. However, the precise role of attention in supporting verbal STM, particularly auditory-verbal STM, remains unclear. Auditory-verbal STM—the ability to maintain spoken information over a brief period—is frequently impaired in stroke survivors with aphasia (SWA). Cognitive-linguistic aphasia treatments often target auditory-verbal STM to strengthen both retrieval and maintenance of linguistic representations. In this study, we examined the role of attention in maintaining speech sounds over a short delay in 18 chronic SWA and 13 age-matched healthy individuals using a dual-task paradigm. The overarching goal of this work is to inform treatment development targeting maintenance of speech sounds at a single word level, and within phrases, sentences and dialogues. We investigated attention as a limited-capacity resource supporting both phonological maintenance and concurrent cognitive operations. Our framework builds on prior work suggesting that: (1) attention is a domain-general resource, insensitive to its content (e.g., visual-spatial, linguistic); (2) STM traces decay in the absence of attention; (3) the proportion of time that attention is occupied by a concurrent distractor task is the proportion of time during which memory traces decay without being refreshed, as described by the time-based resource sharing model. Our dual-task paradigm consisted of a maintenance task with consonant-vowel (CV) syllable strings and a visual distractor task involving shape judgments. Maintenance load was manipulated by varying CV set sizes (1-5), and cognitive load by varying the distractor shape judgment difficulty (hard, easy). We expected responses to hard shape trials to take longer than those to easy trials, thereby modulating cognitive load. Maintenance performance was assessed using a probe span recognition task after an 8-second maintenance period. We hypothesized that increased cognitive load would impair phonological maintenance. A hard distractor condition would create a bottleneck by limiting the time devoted to maintenance of phonological traces, resulting in their decay and ultimately poorer recognition performance. Data were analyzed using signal detection theory and linear mixed effects modeling. In both SWA and healthy individuals, cognitive load affected probe recognition sensitivity (d’) for set sizes 4 and 5, with poorer recognition after hard vs. easy distractor trials (p=0.032). No effects were found for set sizes 1 or 2. Unexpectedly, for set size 3, recognition sensitivity was better after hard than easy trials (p=0.017). While recognition performance declined from set sizes 1 to 3 during easy trials, no such decline occurred during hard trials, resulting in relatively better performance at set size 3. We interpret this paradoxical finding as a switch to a more efficient maintenance mechanism under dual-task timing constraints. Overall, results suggest a bottleneck in attentional resource sharing, which affects higher maintenance loads, underscoring the role of attention in phonological STM. Intermediate maintenance loads may benefit from attentional mechanisms in SWA, providing compelling preliminary support for testing the use of dual-task paradigms as a treatment strategy to improve auditory-verbal STM.
Topic Areas: Control, Selection, and Executive Processes, Disorders: Acquired