The Effect of Exposure Duration on Perceived Similarity in Simultaneous Lineups

Abstract

Quantifying reliable and accurate eyewitness identification procedures which avoid wrongful convictions and give confidence to justice systems as to the accuracy of suspect guilt continues to be an area of intense research. Defining the parameters of a 'fair lineup' is central to this endeavour. Measures of similarity between lineup members have been key variables used to accurately describe what is and isn't a 'fair lineup'. To date little research has been done on how the perception of similarity may vary across groups and conditions, particularly as a result of memory encoding strength. This study aimed to understand how exposure time, a key variable for altering the encoding strength of a face, in the context of simultaneous lineups, may alter perceived similarity. Results showed that the observed data fit the Unequal Variance Signal Detection (UVSD) model well, however the predicted increases in discriminability with longer exposure duration and higher lineup similarity were not measured. Similarly no significant changes in perceived similarity were found between any of the conditions. Given observed differences in Hit (CID) and False Alarm (FA) rates between low and high similarity lineups this result suggests that judgements of perceived similarity between faces in a line up are unrelated to participants face familiarity judgements. This supports the independent observations assumption within the maximum likelihood method and indicates that overall a priori categorical classifications of lineups as wholly low or high in similarity are less important to discriminability than participants judgements about each faces familiarity to the memory of the target. This finding has implications for future research into 'fair lineup' design and measurement.