nucleation

LIT-557: A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation

3 min read

Properties6
StatusActive — read in full 2026-10-03 (NOTE-438), from an OCR of the scanned published article the authors post on selfdeterminationtheory.org; worth reading as the largest meta-analysis of the reward-undermining effect and the evidential base SDT cites for it. Across 128 experiments on interesting tasks, expected tangible rewards undermined free-choice intrinsic motivation (d = −0.36), engagement-, completion- and performance-contingent rewards all did (−0.40, −0.36, −0.28), unexpected and task-noncontingent rewards did not, and positive feedback enhanced it (0.33), though not for children's free choice. It reanalyses Cameron and Pierce (LIT-556) study by study. Its categories are CET's, and the rival authors reply, by the abstract of Cameron, Banko and Pierce (LIT-562, unread), that the negative effects are circumscribed, not pervasive.
Version2
Tagsmotivation, social-science, agency, philosophy-of-science, learning-and-conditioning
Compared againstLIT-556 — Reinforcement, Reward, and Intrinsic Motivation: A Meta-Analysis, LIT-562 — Pervasive negative effects of rewards on intrinsic motivation: The myth continues, LIT-754 — Effects of reward on intrinsic motivation—Negative, neutral, and positive: Comment on Deci, Koestner, and Ryan (1999)
Cited as “paper” byNOTE-438 — A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation
Cited as “source” byTHEORY-046 — Expected tangible rewards for merely engaging in an interesting task lower later free-choice engagement with it, THEORY-049 — Whether an external event lowers or raises intrinsic motivation depends on whether it is experienced as controlling or as information about competence

Graph View

Edward L. Deci, Richard Koestner and Richard M. Ryan (1999), Psychological Bulletin 125(6):627–668 — DOI-10.1037/0033-2909.125.6.627

Key takeaways

  • Tangible rewards undermine intrinsic motivation for interesting tasks when they are expected and contingent on doing, finishing or doing well at the task. Free choice: engagement-contingent d = −0.40, completion-contingent −0.36 (−0.44 with one outlier removed), performance-contingent −0.28.
  • They do not when unexpected (d = 0.01) or task-noncontingent (−0.14, not significant).
  • Positive feedback enhances free choice (0.33) and self-reported interest (0.31), but for children’s free choice the effect was nil (0.11, not significant).
  • The worst case is the common one. Performance-contingent rewards that give some people less than the maximum, with a no-feedback control, were the most detrimental of all (d = −0.88, p. 644).
  • Undermining is not transitory in children: −0.40 immediately, −0.49 within a week, −0.53 after a week (p. 650). Only children were studied with delays.

Standing in the record

Filed on 2026-10-03 at the owner’s request, as the evidence SDT’s summary papers (LIT-559, LIT-558, LIT-560) cite for the reward-undermining effect, and as one side of the record’s first empirical dispute in the motivation topic.

Relations to Cameron and Pierce (1994), LIT-556.

  • compared_against: this paper ran the comparison. Its Appendix A lists every study it shares with Cameron and Pierce and says, study by study, whether the coding, control group and effect size match theirs, and Appendix B lists the studies of theirs it excluded and why.
  • Not declared corrects. This paper says Cameron and Pierce were wrong on specific things. The clearest is Boal and Cummings (1981), which they counted as a rewarded-versus-unrewarded comparison although every group in it was paid (pp. 633, 668). Others are codings in Appendix A and the collapsing of positive and negative feedback in Rosenfield et al. (1980) (p. 655). But the record has read neither Cameron and Pierce nor their reply, so it cannot say which of these charges stand, and declaring corrects would take this paper’s side. The reply (LIT-562) has since been read (NOTE-587). It answers these charges by re-coding, not by conceding them, and the two analyses agree on the core effect. Neither shows the other wrong on something it claimed, so no corrects is declared in either direction.

The reply, LIT-562, records the comparison from the other side (compared_against this paper). Re-coding by procedure, it finds negative effects only for tangible, expected rewards “loosely tied to level of performance” or earned at less than the maximum. Rewards for surpassing a score come out neutral and rewards for exceeding others positive. Neither relation says which overall conclusion is right.

No instruction for machine-learning practice; nothing here belongs in the anthology.

Backlinks