Why Errors in Alibis are Not Necessarily Evidence of Guilt

Abstract

Laypeople, police, and prosecutors tend to believe that a suspect’s alibi, if truthful, should remain consistent over time (see Burke, Turtle, & Olson, 2007; Culhane & Hosch 2012; Dysart & Strange, 2012). However, there is no empirical evidence to support this assumption. We investigated (a) whether some features of an alibi – such as what was happening, who with, where, and for how long – are more likely to produce errors than others; and (b) whether consistency in alibi stories is correlated with particular phenomenological characteristics of the alibi such as a person’s confidence and sense of reliving the event. We asked participants to imagine they were suspected of a crime and to provide their truthful alibi for an afternoon 3 weeks prior and to complete questions regarding the phenomenological characteristics of their memory. We also asked participants to locate evidence of their actual whereabouts for the critical period. Participants returned a week later, presented their evidence, re-told their alibi, and re-rated the phenomenological characteristics of the alibi. Our results revealed that participants were largely inconsistent across all aspects of their alibi, but there was variability across the different features. In addition, those who were inconsistent were less confident, recollected the time period in less detail and less vividly, and were less likely to claim to remember the time period. We conclude that inconsistencies are a normal byproduct of an imperfect memory system and thus should not necessarily arouse suspicion that a suspect is lying.