Earwitness Testimony: Never Mind the Variety, Hear the Length

Abstract

Three aspects of voice recognition were investigated in the study reported here: memory for familiar voices, memory for the words spoken, and the relative effects of length and variation in a voice extract on long‐ and short‐term memory. In Experiment 1, recognition memory for the briefly heard voice of a stranger was superior with longer extracts (p<0.01), but increasing vowel variety did not improve performance. This pattern was repeated for short‐term memory (p<0.01) in Experiment 2. Scores for the above task correlated significantly (p<0.05) with scores for recognizing well‐known voices. In a further test of well‐known voice memory in Experiment 3, a weak and non‐significant positive correlation (r=0.29) was found between memory for well‐known voices and memory for a once‐heard voice. Memory for the words spoken did not correlate significantly with memory for the unknown voice itself. The possibilities of a memory‐for‐voices general ability, and forensic applications of the findings are discussed. © 1997 by John Wiley & Sons, Ltd.

Eyewitness Accuracy When Making Multiple Identifications Using the Elimination Line-Up

“…Research on identifications independent of facial identifications has generally focused on using voice or body characteristics as identifiers, with varying degrees of success (Cook & Wilding, 1997;Philippon, Cherryman, Bull, & Vrij, 2006;Pryke et al, 2004). Accuracy in identifying a voice sample is dependent on a number of factors, including familiarity of the speaker's voice, length of exposure to the voice and number of voice foils in the line-up (Cook & Wilding, 1997). Most research has therefore found it to be difficult for witnesses to accurately identify a suspect through their voice (Cook & Wilding, 1997;Olsson, Juslin, & Winman, 1998).…”

“…Accuracy in identifying a voice sample is dependent on a number of factors, including familiarity of the speaker's voice, length of exposure to the voice and number of voice foils in the line-up (Cook & Wilding, 1997). Most research has therefore found it to be difficult for witnesses to accurately identify a suspect through their voice (Cook & Wilding, 1997;Olsson, Juslin, & Winman, 1998).…”

“…Descriptions of perpetrators, however, typically include details regarding additional descriptors (e.g., Pozzulo & Warren, 2003). Research on identifications independent of facial identifications has generally focused on using voice or body characteristics as identifiers, with varying degrees of success (Cook & Wilding, 1997;Philippon, Cherryman, Bull, & Vrij, 2006;Pryke et al, 2004). Accuracy in identifying a voice sample is dependent on a number of factors, including familiarity of the speaker's voice, length of exposure to the voice and number of voice foils in the line-up (Cook & Wilding, 1997).…”

Earwitness identification performance: the effect of language, target, deliberate strategies and indirect measures

“…These favourable conditions included short delay, long exposure to the target voice during encoding, with the speech samples for the events lasting between 45 and 50 seconds. This was well above the duration necessary for optimum voice identification (Cook & Wilding, 1997;Pollack, Pickett, & Sumby, 1954).…”

Overhearing the Planning of A Crime: Do Adults Outperform Children As Earwitnesses?

“…We have only found a handful of empirical studies concerned with children's voice-memory. Some of them have limited relevance in the present context since they are concerned with memory for familiar voices (Bartholomeus 1973;Murry and Cort 1971;Spence et al 2002), and it is not possible to generalise findings about familiar voices to identification of unfamiliar voices (Cook and Wilding 1997;Van Lancker and Kreiman 1985).…”