Enhancing and Undermining Intrinsic Motivation: Task-Involving and Ego-Involving Evaluation
Butler, R. · 1988
grade Crctindependentreplicated
Sample
132 Israeli 5th-6th graders, 12 classes
Population
5th-6th graders (high and low classroom achievers).
Design
The canonical grades-vs-comments experiment: numerical grades vs individualized task comments vs both, across sessions. Researcher-scored outcomes; exact cell values paywalled.
Key findings
The classic: task-involving comments raised interest and performance; ego-involving grades depressed both; and grades+comments behaved like grades alone — adding a grade neutralizes the formative comment. A strong existence proof, not an effect-size anchor.
Genetic confound
Randomized; the achievement-level moderation (high achievers tolerate grades) tracks prior ability, which is heritable — grading regimes plausibly differentially harm low-ability pupils.
Replication notes
Direction replicated across the Butler programme and pooled by Koenka 2021. No modern large preregistered replication; magnitudes unknown.
DOI / URL
10.1111/j.2044-8279.1988.tb00874.x
Effects
| Outcome | Metric | Value | Measure | Timing | Vs | Horizon | Class |
|---|---|---|---|---|---|---|---|
| Performance & interest | direction | comments > grades AND > grades+comments; the grade cancels the comment | researcher-designed | immediate | active-alternative | end-of-treatment | non-cognitive |
| High achievers expecting more grades | direction | exception: maintained interest/performance — normative grading rewards those it ranks well | researcher-designed | immediate | active-alternative | end-of-treatment | non-cognitive |
Cited by
- Feedback and formative assessmentmixedconf: highgc: low