The cognitive interview: A meta-analysis

TLDR

The study meta‑analyses the impact of the cognitive interview on correct and incorrect recall. The analysis included 42 studies (55 comparisons) with about 2,500 participants. The cognitive interview significantly increased correct recall (d = 0.87) and, to a lesser extent, incorrect recall (d = 0.28), with accuracy rates nearly equal to standard interviews; effect sizes were larger for staged events and active participation, higher in adults, and the enhanced version produced more errors.

Abstract

Abstract A meta-analysis was performed on the effects of the cognitive interview on correct and incorrect recall. The database comprised 42 studies with 55 individual comparisons involving nearly 2500 interviewees. A strong overall effect size was found for the increase of correctly recalled details with the cognitive interview compared to a control interview (d = 0.87). The overall effect size for the increase in incorrect details, although considerably smaller, was also significant toward the cognitive interview (d = 0.28). However, the accuracy rates (proportion of correct details relative to the total amount of details reported) were almost identical in both types of interview (85% for the cognitive interview and 82% for standard interviews, respectively). Taking methodological factors into consideration it was found that effect sizes for correct details were larger if staged events were used as the to-be-remembered episode (as compared to video films) and if the interviewees actively participated in the event. With regard to incorrect details effect sizes were larger for adults as compared to children. Furthermore, the enhanced version of the cognitive interview produced more errors than the original version.