Dynamic shifts of limited working memory resources in human vision.
Level 3 - non-randomized controlled study
Laboratory-controlled human experimental behavioral study (graded by design analogy, not clinical CEBM).
PubMed 18687968 · doi:10.1126/science.1158023
What was done
The authors investigated human visual working memory architecture to test whether capacity is constrained by a fixed number of discrete item slots or functions as a continuous, dynamically shared resource allocated across visual scene items according to selective attention and upcoming eye movements.
What was found
The abstract reports no numerical values, sample sizes, effect sizes, or statistical metrics. It describes qualitative findings: memory capacity is not fixed by object count but is dynamically shared among items; resources shift flexibly with biases from attention and saccade targets; and recall precision relates to allocated resource proportion via a power law.
Why it matters
This study provides a foundational continuous-resource alternative to discrete-slot models of visual memory, offering a quantitative framework linking attentional selection to memory precision.
Limits
The abstract omits sample size, participant demographics, exact task design, and quantitative parameters of the power law. Findings from controlled laboratory psychophysics may not fully generalize to complex real-world visual behavior.
Cited by
- supports The quality and precision of information stored in working memory degrades as more items are added.