Blind to the Truth? Testing the Impact of Sequencing Case Information on Contextual Bias in Credibility Assessments

The credibility assessment of a victim statement can provide crucial information in sexual abuse cases. A content analysis of the statement, often using Criteria-Based Content Analysis (CBCA), is a core element in this assessment procedure but might be prone to bias by contextual case information. We first wanted to confirm whether such contextual bias indeed affects credibility assessments. To then mitigate this, we tested a sequencing approach regulating the release of case information based on its relevance. In a preregistered experimental study, 467 participants assessed a victim statement on four CBCA criteria and judged its overall credibility. Contextual information was provided either before or after the CBCA scoring or not at all, resulting in three conditions. Results showed that the CBCA scores remained unaffected by contextual information, but the final credibility judgement showed strong contextual bias. Contrary to our expectations, sequencing did not have the expected protective effect. These results suggest that CBCA might be more robust to bias than previously assumed, but boundary conditions of such robustness need to be explored further. More generally, we recommend to increase standardisation of the entire procedure to reduce contextual influences in credibility assessments.