The nature of individual differences in working memory capacity: Active maintenance in primary memory and controlled search from secondary memory.
DOI: 10.1037/0033-295x.114.1.104
archive: archived pipeline: cataloged verified
Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)
Summary
This paper addresses the nature of individual differences in working memory capacity (WMC), proposing a framework that integrates experimental and differential psychology approaches. The authors argue that WMC limitations arise from two distinct components: a dynamic attention component termed primary memory (PM) and a probabilistic, cue-dependent search component termed secondary memory (SM). The research is motivated by the need to explain why individuals with low WMC perform poorly on complex cognitive tasks, such as reading comprehension and fluid reasoning, compared to those with high WMC. The central thesis is that individual differences in WMC are driven by variations in the ability to actively maintain information in PM and the ability to effectively search for and retrieve information from SM in the presence of interference. The authors review existing literature and present new evidence based on their theoretical framework. They define PM as a limited-capacity system (approximately four items) that maintains representations through the continued allocation of attention. When attention is diverted or capacity is exceeded, items are displaced to SM. Retrieval from SM requires a controlled, cue-dependent search process to discriminate relevant information from irrelevant competitors. The paper synthesizes data from various paradigms, including antisaccade tasks, Stroop tasks, and dichotic listening, to demonstrate how WMC differences manifest when active maintenance or controlled retrieval is required. Additionally, the authors introduce new experimental evidence regarding immediate free recall to test predictions about maintenance and retrieval mechanisms. The findings indicate that WMC differences are not universal but appear specifically when tasks require overriding automatic responses or retrieving information amidst interference. In antisaccade tasks, low WMC individuals showed higher error rates and slower latencies only when active maintenance of the task goal was necessary (e.g., intermixed pro- and antisaccade trials), not when the task was predictable. Similarly, in Stroop tasks, low WMC individuals made more errors in conditions with high congruency, where maintaining the goal to name ink color was critical to override the habitual reading response. In conditions where maintenance was not required, no WMC differences were observed. Regarding retrieval, the authors posit that low WMC individuals are less efficient at using contextual cues to delimit the search set in SM, leading to larger search sets containing more irrelevant information and thus poorer recall performance. The significance of this work lies in its unified explanation of WMC as a combination of active maintenance and controlled search capabilities. The authors conclude that high WMC individuals are superior because they can better sustain attention on relevant representations in PM and more effectively use cues to constrain search processes in SM. This framework clarifies why WMC predicts performance on higher-order cognitive tasks: these tasks demand the active maintenance of goals and the strategic retrieval of information in the face of distraction and interference. By linking these mechanisms, the paper provides a comprehensive model for understanding the cognitive underpinnings of individual differences in working memory.
Key finding
Individual differences in working memory capacity arise from variations in the ability to actively maintain information in primary memory and the ability to effectively use cues to control search processes in secondary memory.
Methodology
theoretical
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via openalex_abstract on 2026-05-08 (3 acquisition events logged).
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | — | — | — | 1 | 2026-05-07 |
| archive | success | canonical_url | — | — | 8 | 2026-06-06 |
| extract | success | cached | — | — | 3 | 2026-06-10 |
| clean | success | clean | — | — | 1 | 2026-06-04 |
| chunk | success | chunk | — | — | 1 | 2026-06-04 |
| embed | success | embed | Qwen/Qwen3-Embedding-8B | — | 1 | 2026-06-04 |
| enrich | success | normalization | — | — | 3 | 2026-05-28 |
| promote | success | — | — | — | 1 | 2026-05-07 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 2 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 15 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.