Earwitnesses: The effect of type of voice lineup in identification accuracy and the realism in confidence judgments

This contribution is a partial report from a study of the identification accuracy and realism in the confidence judgments of the correctness in the identification reports in two kinds of target-present voice lineup. 24 men and 54 women were asked to identify a voice that they had heard previously in a dialogue context that simulated the planning of a burglary by two males 22 and 27 years old. The voice lineup either consisted of recordings of each of six male speakers reading a text from a book (text-lineup condition) or each of the same six speakers having a spontaneous dialogue with another male speaker (dialogue-lineup condition). Each recording lasted 30 seconds. The results showed a
tendency (p<.06) for better accuracy and better ability to separate correct from incorrect identification responses by means of ones’ confidence judgments for the text-lineup condition compared with the dialogue-lineup condition. The text-lineup condition also showed a tendency for lower overconfidence. These results deviate from expectations following from the encoding specificity principle in memory psychology (Tulving & Thomson, 1973), maybe because text reading provides a more varied representation of the features of the human voice compared to dialogues.