Paying students to read
An after-school program serving 120 fourth- and fifth-graders introduces a 'reading points' system: students earn one point per 20 minutes of independent reading, redeemable for small prizes. Over the first 10 weeks, logged reading time rises from a mean of 22 to 61 minutes per week. When the program ends in May and prizes stop, teachers report that by September many former participants read less than a comparison group that never had the program - a drop concentrated among students who had been the most frequent readers before the incentive began. Staff propose relaunching with a bigger prize menu.
1. Account for the September pattern using motivational theory, and say which students the theory predicts would be most affected.
Read the worked answer
The pattern fits overjustification: when an activity that was already intrinsically interesting is paired with a salient, contingent external reward, people may reattribute their behavior to the reward ('I read because I get points'). When the reward is withdrawn, the inferred intrinsic motive is weaker than before, so behavior falls below baseline.
The theory predicts the drop is largest precisely where intrinsic motivation was highest to begin with - the heavy readers, whose behavior was most plausibly self-determined before the program. That matches the reported concentration. Students who read rarely at baseline had little intrinsic motive to erode, so their post-program behavior should regress toward their own baseline rather than below it.
Note the alternative explanation to rule out first: a simple contrast/withdrawal effect (behavior was reward-supported, reward removed, behavior drops to baseline). The discriminating evidence is below-baseline reading in the heaviest readers - that is what pushes the explanation toward motivational undermining rather than mere reinforcement removal.
2. The relaunch proposes a larger prize menu. Predict the effect and justify your prediction.
Read the worked answer
Scaling up the same contingency should increase reading while the program runs (larger, more salient reinforcers sustain higher response rates) but should worsen the post-program erosion, because a bigger, more attention-grabbing reward makes the external attribution even more plausible and the withdrawal even more noticeable.
This is a case where the short-run outcome measure (logged minutes during the program) and the outcome of interest (reading after the program) point in opposite directions - so a trial that only measured in-program minutes would 'confirm' the wrong policy.
A prediction worth stating: the correct test is post-program reading at follow-up, with the incentive arm compared to a no-incentive arm, and with baseline intrinsic interest measured so the moderating effect can be detected.
3. Describe a redesign that would plausibly preserve in-program reading gains without the September drop.
Read the worked answer
Shift from controlling to informational/autonomy-supportive design:
- Reward effort, choice, and completion of self-selected goals (e.g., 'you finished the book you picked') rather than minutes logged, which targets competence feedback rather than behavioral control.
- Make rewards unexpected and non-contingent (a surprise book fair) rather than a predictable payoff for each unit of reading.
- Preserve choice of texts and give students a role in setting targets - autonomy support is the component most associated with maintaining intrinsic motivation.
- Keep some structural supports that don't function as payment: access to books, peer book clubs, adult modeling.
Rationale: the undermining effect depends on the reward being perceived as controlling and expected. Change those two features and the same reinforcer no longer supplies an external causal attribution for reading.