Working memory: maintenance, manipulation and competing tasks
Compare dual-task costs to test resource-sharing hypotheses without treating working memory as a fixed universal number.
Question and prediction
Why might one secondary task disrupt a primary task more than another?
Predict whether a verbal and a spatial secondary task necessarily produce equal costs on verbal recall.
Mechanism and evidence
Working memory concerns temporary maintenance and use of information. Baddeley and Hitch’s model challenged the idea that one simple short-term store explained all reasoning and recall. Later multicomponent formulations distinguish verbal and visuospatial maintenance, attentional control and integration. These are functional models supported by patterns of evidence, not visible boxes or exact labels for single brain regions.
A dual-task study asks participants to perform two tasks together. If tasks compete for overlapping processes, performance may suffer more than when demands differ. But unequal secondary-task difficulty, strategy changes and trade-offs can also produce a cost. Measure both tasks, preserve single-task baselines and avoid inferring a resource merely because a result fits its name.
In the fictional table, verbal recall is 8 of 10 alone, 5 with a verbal task and 7 with a spatial task. Costs are 3 and 1 items relative to the same baseline. The pattern is compatible with greater overlap in the verbal combination. If participants almost ignore the spatial task, however, its small cost would not establish separate capacity. The secondary performance records are essential.
Chunking reorganizes information using learned structure; it does not create an unlimited storage capacity. Capacity estimates depend on material, task, prior knowledge and scoring. The original 1974 report supplies a historical model-development context, while the new numbers are invented. The activity evaluates task architecture rather than ranking learners by a memory-span score.
Inspect the relationships
| Condition | Primary recall | Cost from alone | Secondary accuracy |
|---|
| Alone | 8/10 | 0 | — |
| Verbal secondary | 5/10 | 3 items | 90% |
| Spatial secondary | 7/10 | 1 item | 45% |
Worked example
Primary costs are 8−5=3 and 8−7=1 items. But secondary accuracy differs sharply. A smaller primary cost may reflect doing less of the spatial task, so the pattern alone does not establish separate resources.
Your investigation
Draw a primary-versus-secondary performance plot and identify the trade-off. Compare the original model’s question with the fictional design. Write a 200-word revision specifying baseline matching, both outcome measures and a result that would challenge a shared-resource account.
Unfamiliar transfer
A familiar sequence is remembered as three meaningful groups instead of nine unrelated symbols. What mechanism fits? Explain your reasoning before checking feedback.
Reading
Baddeley and Hitch (1974): Working Memory. The records below are original fictional teaching cases, not research results from this source.
Fictional dual-task outcomes Primary verbal recall: alone 8/10; with verbal secondary task 5/10; with spatial secondary task 7/10. Secondary accuracy: verbal 90%, spatial 45%. Equal trial counts.
Fictional design revision Researchers propose adjusting secondary tasks to comparable single-task accuracy and monitoring both outcomes under dual-task conditions. No new results are available.
Authentic evidence and its limits
Original model-development question — Alan D. Baddeley and Graham Hitch (1974). The authors examined the role of temporary memory in reasoning, comprehension and recall, including performance under concurrent memory load, and proposed a working-memory account.
Paraphrase of the original chapter’s research program. Later additions to the model are not all part of its 1974 formulation. The invented dual-task table is not an original result.
Read the original source. Original LearnAI paraphrase and attributed factual results; no reproduced figure or substantial verbatim passage. Original publication retains its rights.
