Abstract:
Prosecutorial “tunnel vision” is an area of interest in the U.S. criminal justice system that is gaining increased attention as more and more wrongful convictions are brought to light. Legal scholars have raised concerns regarding the possibility that prosecutors are failing to recognize and/or disclose to the defense any knowledge of evidence that may exonerate the defendant, a requirement established by U.S. Supreme Court case Brady v. Maryland. This tunnel vision is considered to be due in part to the manifestation of confirmation bias, a well-documented phenomenon regarding the search for and interpretation of information. The current study investigates whether and to what extent prosecutors demonstrate confirmation bias in their review of evidence in a hypothetical homicide case. Active prosecutors were recruited via email and exposed to a fictional arrest report. After reporting their initial impressions of suspect guilt, participants were randomly assigned to three groups, manipulating exposure to new evidence by valence (inculpatory, exculpatory, and ambiguous). Participants evaluated the evidence in terms of credibility and degree of incrimination. Convergent with prior literature, it was hypothesized that initial ratings of guilt would predict case processing decisions and final impressions of guilt, but that this relationship would be mediated by evaluations of evidence credibility/incriminating power, evaluations which in turn would be moderated by evidence valence. Data from the study did not support the hypothesized moderated mediation model. Results indicated that prosecutors make case processing decisions based on appraisals of evidence that stand independent of initial impressions of suspect guilt. Implications regarding prosecutors’ objectivity, cognitive flexibility, and adherence to Brady are discussed.
