Strategy shifts in classification skill acquisition: Does memory retrieval dominate rule use?

Bourne, Lyle E.; Healy, Alice F.; Kole, James A.; Graham, Susan M. · 2006 · OpenAlex

DOI: 10.3758/bf03193436

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the dynamics of strategy shifts during classification skill acquisition, specifically examining whether memory retrieval or rule-based processing dominates performance. The research challenges existing models (e.g., Logan, 1988; Rickard, 1997) that assume a unidirectional shift from rule-based to memory-based strategies as practice increases. Instead, the authors propose that strategy dominance depends on "shift affordance," determined by rule difficulty and cue salience. The study aims to determine if simplifying a rule prevents the shift to memory retrieval and how retention intervals affect strategy use. The researchers conducted two experiments using binary classification tasks. In Experiment 1, 36 participants were assigned to one of three conditions: "artificial hard" (complex alphabetical rule), "artificial easy" (simple alphabetical rule), or "natural" (pronunciation rule for the definite article "the"). Participants completed 30 blocks of practice trials, followed by a second session one week later. They were not instructed on the rules but received immediate feedback. After each response, participants self-reported their strategy: guessing, using a rule, retrieving memory, or other. Response times (RTs) and accuracy were recorded. Experiment 2 manipulated cue salience by scrambling letters in stimuli to test if focusing on a single cue (first letter) would alter strategy preferences compared to the natural task. Results from Experiment 1 demonstrated that strategy dominance varied by condition. In the artificial hard task, participants shifted from rule use to memory retrieval as practice progressed, consistent with previous models. However, in the natural task, rule use remained dominant throughout both sessions, contradicting the expectation that memory would eventually take over. The artificial easy task showed a compromise, with participants evenly split between rule and memory strategies. Crucially, after the one-week retention interval, participants in all conditions reverted to rule use initially, indicating that specific instances were forgotten more rapidly than the underlying rule. This "reverse shift" from memory to rule occurred because rule retention was superior to stimulus item retention. Furthermore, RTs were consistently faster for the dominant strategy in each condition. Responses to novel stimuli confirmed the validity of the strategy probes, as participants predominantly used rules for unseen items. The findings suggest that strategy shifts are not automatic or unidirectional but are governed by the relative costs and benefits of rule application versus memory retrieval. Rule difficulty significantly influences strategy preference; simpler rules sustain rule-based performance, while complex rules afford a shift to memory. Additionally, the durability of rule knowledge compared to instance memory explains the temporary reversion to rule use after delays. These results imply that existing models of skill acquisition must account for task-specific affordances and the differential forgetting rates of rules versus instances, rather than assuming a universal transition to memory-based processing.

Key finding

The dominant strategy in classification tasks is determined by rule difficulty, with simpler rules sustaining rule-based performance and more complex rules promoting a shift to memory retrieval.

Methodology

lab_experiment

Sample size: 36

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 (14 acquisition events logged).

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.

• automaticity skill acquisition