evidence EV-COMM-2026-0009
Subjective task ratings can correlate weakly with observed performance
Evidence Record
Evidence summary
In scenario-based testing with expert and novice nurses, subjective task-completion survey ratings had weak or marginal correlations with observed performance measures including task success and time.
Exact claim supported or contradicted
Supports treating satisfaction, perceived ease, and confidence as diagnostic secondary measures rather than proxies for correct completion.
Source provenance
Peer-reviewed empirical pilot study available through PubMed Central. The reported correlation analysis covered 108 expert-group and 63 novice-group task observations, not independent participants.
Relevant excerpt or data
The paper’s Table 3 reports correlations between task-completion survey ratings and mouse activity, time on task, and percent task success; the authors characterize the overall pattern as weak or marginal.
Interpretation
Subjective experience remains useful for explaining adoption and friction, but agreement with objective performance must be measured rather than assumed.
Limitations
Clinical software, repeated task observations, small pilot samples, and interface interaction may not generalize to standalone procedural documents. Correlation does not establish causal direction.
Counterevidence
Other studies find moderate associations between subjective and behavioral measures. Divergence is possible, not inevitable; both channels are needed.
Reproduction or verification notes
Inspect the Methods and Table 3 in the full-text article.