Key study: Vohs and Schooler (2008)

Aim: to see whether weakening people's belief in free will makes them more likely to cheat.
Participants: 30 university students.
Procedure: half read a passage from a book by scientist Francis Crick arguing that free will is an illusion (anti-free will). The other half read a passage from the same book about consciousness that did not mention free will (control).
They completed questionnaires on their belief in free will and their mood. Then they did 20 mental arithmetic problems on a computer. They were told a 'glitch' would show the answer unless they pressed the space bar. Not pressing it was a measure of cheating.

Experiment 2: 122 students were split into five groups. Some read statements supporting determinism, some read statements supporting free will, and one read neutral statements; two further groups were controls. They then answered 15 problems from a reasoning test and, in most groups, marked their own answers, shredded their sheets and paid themselves $1 for each correct answer, so they could cheat.
Results: in Experiment 1, the anti-free will group reported weaker belief in free will and cheated more (let the answer appear more often). In Experiment 2, the determinism group paid themselves more than the other groups, suggesting more cheating.
Conclusion: encouraging a belief in determinism increases cheating. Believing in free will helps people act morally.

Strengths: laboratory experiments with standardised procedures and controls, so cause and effect can be suggested. Two different experiments and two cheating measures gave similar results. Mood was measured, so mood was not the cause.
Weaknesses: small, narrow samples of university students. The tasks were artificial, so low ecological validity. Participants were deceived about the glitch. Some later attempts to replicate the findings did not get the same results.