Where leading questions can lead: The power of conjecture in social interaction.

A theoretical analysis of the structure of social interaction led us to postulate two distinct processes through which leading questions may mislead people. When a questioner asks a leading question of a respondent, observers may use their knowledge of conversational rules to infer that the questioner had an evidentiary basis for the question. Hence, observers will treat the question as conjectural evidence for the view of the respondent implied by the question. In addition, observers who listen to respondents answer leading questions may be misled by the answers because, in an effort to cooperate with the questioner, respondents may supply beh(lvioral evidence that misrepresents their actual personalities. To test these hypotheses, we conducted two experiments in which observers listened to tape-recorded interviews. Questioners in these interviews asked respondents a series of leading questions that probed for evidence of either introversion or extroversion. The first experiment showed that it did not matter whether observers had access to the conjectural evidence in the interviewer's questions, the behavioral evidence in respondent's answers, or both. In each case observers inferred