Protocol for the development of a multiple-choice questionnaire for examining recall accuracy in eyewitnesses

Multiple-choice questions (MCQs) are a valuable tool for assessing eyewitness memory, offering a structured approach to evaluate recall accuracy and susceptibility to misinformation. However, prior research utilising MCQs highlights limitations in calibration and sensitivity, potentially reducing their effectiveness. To address these issues, we here describe and implement a four-stage process to develop a refined set of MCQs. Following a content analysis of qualitative witness statements from a previously used mock crime video, we identified key details across four categories (Action, Object, Person, and Setting). Using these findings, 100 MCQs were developed, tested, and refined to calibrate for difficulty. Additionally, confidence ratings were included to explore their influence on response behaviours. Results found that confidence ratings had no effect on accuracy but potentially provide additional information for the relationship between recall accuracy, confidence ratings, and response patterns. By establishing a clear protocol for MCQ development, we ensure consistency and comparability in future research. These MCQs will next be used to explore recall accuracy and misinformation susceptibility among young and older adults. Keywords: Self-Administered Interview, Multiple-choice questions (MCQs), eyewitness memory, question development protocol. accuracy-confidence relationship, investigative interviewing.