Behavioral Sciences & the Law · 2010 · 109 citations · 30 references
Forensic PsychologyExpert Witness ResearchEducationPsychometricsResearch EthicsCommunicationMisinformationPsychologyJournalismBiasFactor AnalysisExpert WitnessPsychological EvaluationPost-truthReliabilityCognitive ScienceTest DevelopmentTrustSubjective DeterminationFact CheckingTrust MetricOutcome MeasureConfirmatory ResearchWitness Credibility ScaleEpistemologyArtsDeception DetectionEvidence-based PracticePersuasion
The judge or jury must subjectively assess expert credibility, yet no standardized measure has been published. This study developed and cross‑validated the Witness Credibility Scale (WCS) to fill that gap. The WCS was built from a 41‑item pool derived from the Osgood Semantic Differential, refined to 20 adjectives across four subscales—likeability, believability, trustworthiness, and intelligence—using ratings from 264 undergraduates and subsequent studies. Factor analysis revealed four factors (knowledge, likeability, trustworthiness, confidence), and the scale successfully distinguished expert groups in five additional experiments, establishing a basis for future outcome research.
The judge or jury makes a subjective determination of when an expert is credible. However, no published measure exists for assessment of the credibility of expert witnesses. The current study addressed this gap by developing and cross-validating the Witness Credibility Scale (WCS). Drawing on the narrative literature, we hypothesized that credibility was a product of four factors: "likeability," "believability," "trustworthiness," and "intelligence." A 41-item measure was initially constructed based on successive iterations of ratings by a panel of judges using items from the Osgood Semantic Differential measure and was subsequently administered to 264 undergraduates. A factor analysis of the data yielded a factor structure that consisted of four factors labeled, "knowledge," "likeability," "trustworthiness," and "confidence." The final version of the WCS used 20 adjectives with four subscales of five items, each subscale reflecting high loadings on the respective factors. The scale was then tested in five additional studies, in which the scale successfully differentiated between groups of videotaped experts testifying in manipulated conditions. The empirical data from these studies permit a foundation for comparing outcome data in future research investigations.
30