A Meta-Analysis of Field Studies on Criteria-Based Content Analysis

Abstract: Criteria-based Content Analysis (CBCA) is employed in forensic contexts to assess the credibility of child sexual abuse (CSA) allegations. The presence of CBCA criteria in an account suggests the narrator personally experienced the described event. Because of differences between field and artificial laboratory settings we conducted a meta-analysis focusing specifically on field studies and quasi-experiments carried out in the field. We formally assessed strict ground-truth criteria (and other methodological aspects) to select only methodologically sound studies. Effect sizes for individual criteria were much larger than effect sizes for most deception cues in previous meta-analyses. We propose an unweighted mean summary score as an alternative to traditional summary scores. Effect sizes were similar for CSA allegations and other kinds of autobiographical events but varied depending on ground-truth base rates.