Enisman · Journal of experimental psychology. Learning, memory, and cognition 2026 · Meta-analytic reanalysis of 6 behavioral studies plus 2 preregistered experimental replications · n=1140

It was the best of times, it was the worst of times: Evidence for switch cost beyond stimulus-response tasks.

Cited 0 times in the scientific literature.

Level 2 - randomized trial

Experimental cognitive psychology reanalysis and preregistered laboratory experiments (graded by design analogy, not clinical CEBM)

PubMed 42406489 · doi:10.1037/xlm0001643 · record verified 2026-08-26

What was done

Researchers investigated whether cognitive switch costs extend beyond simple stimulus-response tasks to higher-level decision making. They conducted frequentist and Bayesian meta-analyses reanalyzing data from six existing studies (N = 694) comparing choices between positive options versus negative options. They further evaluated the effect in two new preregistered experimental studies (N = 241 and N = 205) testing whether switching between positive and negative choice sets (versus repeating them) induces a switch cost.

What was found

The reanalysis of the six studies (N = 694) revealed a consistent pattern of switch costs when alternating between positive and negative choices, supported by both frequentist and Bayesian meta-analyses. This switch cost effect was successfully replicated in both preregistered experimental studies (N = 241 and N = 205). The abstract does not report specific numerical effect sizes, reaction time differences, or statistical test values.

Why it matters

This work shows that task-switching paradigms and cognitive switch costs apply to complex, value-based decisions rather than just basic stimulus-response pairings. It suggests that choosing between positive versus negative options engages distinct cognitive task sets and proposes task-switching as an implicit method to distinguish cognitive concepts.

Limits

The abstract reports no numerical effect sizes, response times, error rates, or statistical parameters. The sample characteristics and demographics are not described. The findings reflect laboratory cognitive tasks, and generalizability to real-world, naturalistic decision-making was not evaluated.

Cited by