Praise raised motivation; expected rewards for a task cut time on it afterwards
HelpedThe short version
Across 14 comparisons, praise led people to spend more time on the task afterwards.

- 14
- studies pooled
- 918
- children and college students in 96 experiments
- Recorded
- behaviour was recorded
- 2 tools
- tested
The review
Who took part
96 experiments compared people who got a reward with people who did not. They included children and college students. About 3,500 people were in the time-on-task analysis.
What they did
Rewarded people got praise, or a reward such as money, for doing an activity. The other people got no reward. Time on the activity was counted after the reward was taken away.
What happened
Praise led people to spend more time on a task, and to say they liked it more. Expected rewards for a task led people to spend a little less time on it later. They did not change how much people liked it. Across all rewards, time on the task did not change clearly. The reviewers did not rate the quality of the studies.
What the review found
- Helped
Across 14 comparisons, praise led people to spend more time on the task afterwards.
Pooled analysis of 14 studies with 918 people · vs people who got no reward · behaviour was recorded
- Helped
Across 12 comparisons, praise made people say they liked the task more.
Pooled analysis of 12 studies with 785 people · vs people who got no reward · they rated themselves
- Mixed
Across 40 comparisons, rewards for a task led people to spend less time on it later.
Pooled analysis of 40 studies with 2,015 people · vs people who got no reward · behaviour was recorded
- No difference
Across 32 comparisons, expected rewards for a task did not change how much people said they liked it.
Pooled analysis of 32 studies with 1,961 people · vs people who got no reward · they rated themselves
How solid is this?
- Behaviour was recorded
- The review didn't check the studies' quality or look for missing studies
Source
Reinforcement, Reward, and Intrinsic Motivation: A Meta-Analysis (1994)
DOI: 10.3102/00346543064003363 (opens in a new tab)

