Robert Eisenberger, W. David Pierce and Judy Cameron (1999), Psychological Bulletin 125(6):677–691 — DOI-10.1037/0033-2909.125.6.677
Key takeaways
From Cameron, Banko and Pierce (2001, LIT-562), not a reading. Their account of their own earlier comment.
- It distinguishes two control groups for the studies Deci et al. call performance-contingent. A complete control is told the performance objective and given feedback; a partial control is told neither (LIT-562, p. 13).
- Comparing reward against each, it found one small difference on free choice. For rewards offered for exceeding others, reward against a partial control was positive and not significant, and reward against a complete control was “significantly positive” (LIT-562, p. 13).
- It examined the procedures Deci et al. pooled as performance-contingent and found “very different effects on measures of intrinsic motivation” (LIT-562, p. 10).
- An analysis in it “indicates that reward contingencies enhance perceptions of autonomy or self-determination”, against cognitive evaluation theory’s prediction (LIT-562, p. 26).
Standing in the record
Filed Deferred because it was not read. It is compared_against Deci,
Koestner and Ryan (LIT-557), whose meta-analysis it comments on in the same
journal issue. Their reply follows it there and is not in the record.
It matters for THEORY-049. That account’s refutation clause asks for rewards tied to a performance standard to raise free choice “against a control group given the same performance feedback”. By LIT-562’s account this paper reports that comparison for rewards for exceeding others, and finds it positive. A reading would say whether it does so with enough studies, and with what effect size, to meet the clause. Until then nothing should lean on it.
No instruction for machine-learning practice; nothing here belongs in the anthology.