Whatever happens in the laboratory stays in the laboratory: the prevalence and prevention of participant crosstalk.

Foreknowledge in research participants can undermine the validity of psychological research. Three studies examined a potentially major source of foreknowledge: participant crosstalk in an undergraduate subject pool. Participants in all three studies attempted to win extra experimental credit by guessing the number of beans in a jar-a nearly impossible task without foreknowledge of the answer. Participants guessing incorrectly were told the correct answer by the experimenter. In Study 1, 23 of 809 participants showed clear evidence of having received the correct answer from a prior participant. In Study 2, a classroom-based treatment asking students not to talk about experiments to other students significantly reduced crosstalk rates. In Study 3, a laboratory-based treatment supplemented the classroom-based treatment. After revealing the number of beans in the jar, the experimenter obtained a verbal commitment from participants that they would not tell anyone about the experiment. The combined treatment nearly eliminated crosstalk.