Abstract
A total of 156 introductory psychology students heard a taped voice of a mock kidnapper for either 30 seconds or 8 minutes. The kidnapper had either a distinctive voice or a non‐distinctive voice, and spoke either in a whisper or in a normal tone of voice. Voice identification from six‐person, tape‐recorded lineups was tested 2 days later. Participants who initially heard the perpetrator speak in a normal tone were tested with normal tone lineups. Participants who initially heard the perpetrator speak in a whisper were tested either with whispered lineups or normal tone lineups. Results showed that identification performance was superior with longer voice‐sample durations. Voice disguise through whispering, distinctiveness of suspect's voice, and changes in tone of voice from initial hearing and lineup test significantly influenced identification performance on both suspect‐present and suspect‐absent lineups. Small but significant accuracy‐confidence correlations were found in both suspect‐present and suspect‐absent lineups. Duration estimations of the length of the speaker's voice‐sample were overestimated, particularly for the short 30‐second voice sample.