The Effect of Distributed Learning on the Identification of Normal-Tone and Whispered Voices

The Effect of Distributed Learning on the Identification of Normal-Tone and Whispered Voices

One hundred and twenty participants (listeners) heard 18-second or 6-minute voice samples, presented in one massed presentation, two distributed presentations, or three distributed presentations. Listeners heard either normal-tone voices or whispered voices. Accuracy of voice identification, confidence in decisions, and remember/know judgements were assessed after a 24-hour delay. Earwitness performance was significantly better for normal-tone voices than whispered voices. There was a significant interaction between frequency of exposure and tone of voice for accuracy of identification. A significant positive relationship was found between accuracy and confidence for normal-tone voices. The relationship between identification accuracy and remember and know responses depended on the tone of voice, the length of the voice sample, and the number of voice samples presented. The forensic significance of the results is discussed.