Beliefs and Data on the Relationship between Consistency and Accuracy of Eyewitness Testimony

Two studies concerned with consistency and accuracy of eyewitness testimony were conducted. In Study 1 potential jurors indicated the degree to which they considered that various witness on-stand behaviours indicated testimonial accuracy. Witness statements that were inconsistent with previous statements were considered to be the strongest indicators of inaccuracy. Study 2 examined the relationship between consistency and accuracy of testimony. Witnesses viewed a film of a robbery and were interviewed twice (2 weeks apart) about the crime in a 4 (interview format) × 2 (interview occasion) design. Regardless of whether consistency was operationalised in terms of direct contradictions between interviews, or degree of agreement on detail across interviews, no more than 10% of the variance in overall accuracy rate was explained by any individual measure. Number of contradictions and overall agreement between interviews did, however, make additive contributions to prediction of overall accuracy. Also, higher correlations between contradiction-based consistency measures and interview two accuracy rate were detected. Neither consistency nor accuracy for specific testimonial dimensions were predictive of accuracy on the other dimensions, or overall accuracy.