David RandContributions
- Demonstrates empirically that monetary incentives do not reduce the illusory truth effect, contrary to some prior expectations about motivation's power to override heuristics.
- Shows the illusion persists even when participants receive explicit, repeated reminders about reward opportunities for accuracy.
- Finds that more analytically reflective participants actually showed larger illusory truth effects (in Experiment 2), suggesting that reflective thinking alone does not protect against fluency heuristics.
- Provides evidence that motivation may be a weaker tool for combating misinformation than commonly assumed.
Method
Two preregistered experiments on Amazon Mechanical Turk (Exp 1: N=602; Exp 2: N=586).
Procedure: Participants first rated 16 statements for interest, then in a truth-rating phase rated the truthfulness of 32 statements (16 repeated, 16 new). Half the participants were told they could earn a \$1 bonus for a randomly selected correct answer (incentive condition); the other half received no monetary incentive (standard condition).
Materials: 32 general-knowledge facts (16 true, 16 false) selected from prior work to be of similar difficulty. Half were created by converting true facts into plausible falsehoods.
Key manipulation: Experiment 1 presented the incentive information once at the start of the truth-rating phase. Experiment 2 presented item-by-item reminders before each statement, making the reward opportunity maximally salient.
Results
Main finding: Repeated statements were judged true significantly more often than new ones in both conditions — replicating the classic illusory truth effect.
- Exp 1: Repeated M=0.77, New M=0.49, F(1,600)=530.08, p<.001, η²=.47
- Exp 2: Repeated M=0.77, New M=0.62, F(1,584)=178.49, p<.001, η²=.23
Incentive effect: Even with monetary incentives, participants continued to show the illusion:
- Exp 1 incentive condition: Repeated M=0.76, New M=0.52, t(303)=14.61, p<.001, d=0.83
- Exp 2 incentive condition: Repeated M=0.76, New M=0.62, t(294)=9.28, p<.001, d=0.54
The condition-by-repetition interaction was marginal in Exp 1 (F(1,600)=6.76, p=.010, η²=.01) but not significant in Exp 2 (F<1).
Analytic thinking: An exploratory analysis found no correlation between illusory truth and performance on a cognitive reflection test in Exp 1 (r=0.08, p=.185). Surprisingly, in Exp 2, more reflective participants showed a larger illusory truth effect (r=0.29, p<.001) — suggesting that analytical ability does not protect against fluency heuristics.
Notes
Strength: Large, preregistered experiments with clear experimental design. Extends prior work by testing whether motivation (absent in most prior studies) can reduce the effect. Item-by-item reminders in Exp 2 represent a maximally salient incentive condition, making the null result particularly striking.
Limitations: Tested only on general-knowledge facts of moderate difficulty; unclear whether results generalize to more emotionally charged misinformation or political claims. The effect sizes in Exp 2 are smaller than Exp 1, suggesting some contextual variation. The puzzling finding that analytic thinking correlates with larger (not smaller) illusory truth in Exp 2 deserves follow-up.
Implications for the field: Challenges the assumption that raising the stakes or making people more motivated to be accurate will substantially reduce reliance on fluency and repetition. Aligns with growing evidence that simple, motivation-based interventions are often insufficient to counteract misinformation belief. Suggests that preventing initial exposure to false information (rather than trying to correct belief after exposure) may be more effective.