Factors predicting conviction in stranger rape cases

Abstract

Background: Despite there being no legal distinction between different types of rapes (e.g., those committed by strangers to the victim versus those committed by perpetrators known to the victim), stereotypical beliefs about rape have meant that these can be treated differently by the justice system. The aim is to explore the factors that predict juries' decisions to convict or acquit in stranger rape cases. Methods: We measured the importance of a range of 20 perpetrator-, victim-, and offense-related factors in predicting outcomes for 394 stranger rape cases tried by a jury. A four-stage analytic process was employed: (a) Kendall's tau-b measured intercorrelations among the factors (predictors); (b) Chi-square and Welch t-tests measured associations between factors and verdicts; (c) binary logistic regression measured the power of factors in predicting verdicts; and (d) Stein's formula was used to cross-validate the model. Results: Jury verdicts were predicted by five offense-related factors and one victim-related factor. None of the perpetrator-related factors were significant predictors of convictions for stranger rape. Conclusion: The findings have potential implications for victims of stranger rape, as well as prosecution and courtroom policy. We show that if a perpetrator is identified and charged, the likelihood of securing a conviction by a jury is high for victims of stranger rape. We suggest that prosecutors could gather as much information as possible from victims about the factors found to be of importance to juries, and judges could instruct juries on assumptions about the characteristics of the offense in order to challenge incorrect beliefs and stereotypes. Ultimately, this could be used to encourage victims of stranger rape to report and testify in court.