Encoding serial items depletes cognitive resource: An adversarial comparison of reconfiguration and resource-depletion accounts of working memory
Year: 2025
Type: preprint
Abstract: When people encode items into working memory, the time taken to complete a secondary task slows as the memory load increases but becomes faster during retention intervals. The findings are best explained by either a reconfiguration process needed only during encoding, or by a resource-depletion model in which encoding of each item drains cognitive resource that recovers during retention. We tested diverging predictions of these accounts with a running span complex span task. List-length was unpredictable, and participants had to recall only the most recently presented N memory items. According to the reconfiguration account, after N items are presented, response times for the secondary task should stabilize. This should occur because the amount of updating required at each memory item beyond N should be equivalent. According to the resource-depletion account, each new encoding event beyond N should continue to deplete a limited resource. This depletion results in a continued slowing of response times for the secondary task. Across four experiments, secondary task response times increased with each new memory item, even after the required memory load was reached. The only exception was with memory loads of 2 where secondary task response times plateaued after 2 items. These findings place limits on reconfiguration as a comprehensive explanation for short-term memory but are mostly in line with a modified resource-depletion account.
Language: English
Subfield: Experimental and Cognitive Psychology
Field: Psychology
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W4415617108