What the research says

Do Rewards Kill Motivation? 40 Years of Research

Do prizes and points crowd out the joy of learning? A meta-analysis of 183 studies and over 200,000 people found that inner motivation and rewards both matter, for different things.

By the Giga Quizzes team4 min read
Three wrapped gift boxes with bows on a red background
Photo: Ubaid E. Alyafizi, Unsplash, Unsplash License, cropped.
In this post
  1. 1Two kinds of motivation
  2. 2The meta-analysis
  3. 3What they found
  4. 4What this means for classroom rewards
  5. 5The limits
  6. 6Using rewards and interest together

For more than forty years, psychologists have argued about a worrying idea called the undermining effect: that paying people, or rewarding them, for something they already enjoy can make them enjoy it less. If it is true, every sticker chart and every point in a classroom game carries a hidden cost. In 2014 Christopher Cerasoli, Jessica Nicklin and Michael Ford stepped back from that argument and asked a more practical question. When it comes to how well people actually perform, how do inner motivation and outside rewards compare, and do they really work against each other?

Two kinds of motivation

Intrinsic motivation is doing something because it is interesting or satisfying in itself: solving a puzzle because puzzles are fun, reading because you want to know what happens. Extrinsic incentives are rewards from outside: money, grades, prizes, points, praise tied to a result.

The worry has always been that the second can push out the first. A child who loved drawing, once paid for each picture, may come to see drawing as work. Much of the long debate has been about whether people keep choosing an activity once the rewards stop. This review asked something different: how well people perform, and what predicts it.

The meta-analysis

Cerasoli and his colleagues gathered studies from school, work and sport that measured both intrinsic motivation and performance. In all there were 183 independent samples, covering 212,468 people. They also sorted the studies by two features that earlier reviews had not examined systematically.

  • Type of performance: quality, meaning how good the work was, or quantity, meaning how much of it was done.
  • How directly the reward was tied to performance: directly, as in being paid per item completed, or indirectly, as in a salary or a general reward for taking part.

What they found

Intrinsic motivation was a medium to strong predictor of performance, with correlations ranging from about 0.21 to 0.45 across the different groupings. People who found a task interesting in itself tended to do it better. That held whether or not incentives were also on offer.

Incentives mattered too, but for a different thing. Considered together, intrinsic motivation predicted more of the quality of performance, while incentives were better predictors of the quantity. Rewards get people to do more; interest gets them to do it well.

The crowding-out effect did appear, but in a specific form. When incentives were tied directly to performance, intrinsic motivation mattered less for how people performed. When incentives were only indirectly tied to performance, intrinsic motivation mattered more. In other words, a tightly coupled reward can take over as the reason for doing the task, and then the person’s own interest has less influence on the result.

The distinction between quality and quantity is the most useful idea in the paper for anyone designing a classroom activity. It means the question “do rewards work?” has no single answer. Rewards work for getting more done. Whether the work is done well depends much more on whether students find it worth doing for its own sake.

The authors’ conclusion is the opposite of the undermining story’s gloomiest version. Incentives and intrinsic motivation are not necessarily antagonistic, they write, and are best considered together.

What this means for classroom rewards

The findings fit what teachers see. A points race can make a class answer more questions faster. It does not by itself make them think harder about each one. For depth, interest in the material is what predicts quality.

The finding about direct incentives also gives a practical rule. Rewards tightly tied to each piece of performance, such as a prize per correct answer, risk becoming the main reason for doing the work. Rewards that are looser, such as recognition for the class as a whole or a celebration at the end of a game, leave more room for interest to drive the quality of the work.

The limits

A meta-analysis of correlations shows which things go together, not exactly what causes what. The studies spanned schools, workplaces and sport, and the balance may differ for children in a classroom. The reviewed research also measured performance, not long-term interest after the rewards stop, which is the question earlier undermining studies focused on. Both questions matter; this review answers the first.

Using rewards and interest together

Points, scores and celebrations in a quiz game are extrinsic incentives. The research does not say to remove them. It says to know what they are good for.

A balance that fits the evidence

  • Use points and speed when the goal is volume: practising lots of facts quickly.
  • Slow down and drop the race for questions that need careful thought.
  • Make the material itself interesting: surprising facts, real puzzles, questions students care about.
  • Prefer loose rewards, such as class celebrations, over prizes for each correct answer.
  • Ask students what they found interesting, not only what they scored.

Self-determination theory, the background to much of this research, is covered in what motivates students, and the specific effects of points and leaderboards in do leaderboards and points help?.

Sources

  1. Cerasoli, C. P., Nicklin, J. M. & Ford, M. T. (2014). Intrinsic motivation and extrinsic incentives jointly predict performance: A 40-year meta-analysis. Psychological Bulletin, 140(4), 980–1008.

More from the blog

All posts →