Coerced Confessions,Judicial Instruction,and Mock Juror Verdicts1

The present research assessed whether judicial instruction can curb jurors' inappropriate use of coerced-confession evidence. In Experiment 1, subjects read an auto theft trial in which the defendant had confessed on his own initiative (no constraint), after an offer of leniency (positive constraint), or after a threat of punishment (negative constraint). Subjects then received an instruction that simply directed them to ignore a coerced confession (short form), another that additionally defined both positive and negative inducement as coercive and hence unreliable (long form), or no instruction at all. As previously reported (Kassin & Wrightsman, 1980), subjects fully discounted the negatively constrained confession but not the positively induced one which, although judged involuntary, produced a high percentage of guilty verdicts. Neither form of instruction significantly reduced this latter tendency. In Experiment 2, subjects read an assault case involving a voluntary or positively coerced confession and one of four types of instruction. The positive coercion bias was replicated. An instruction that stressed both the unreliability and unfairness of an induced confession decreased voluntariness judgments but failed to lower the conviction rate. The theoretical basis for and practical implications of this phenomenon are discussed, and future research directions are proposed.