Transcript Accuracy in Preclinical Medical School Lectures

IntroductionMedical students acquire knowledge in the preclinical phase primarily through didactic lectures. Lectures are commonly delivered via video recordings and include transcripts and/or closed captions. Transcript accuracy may improve or worsen student perceptions of equity and accessibility in education, but data quantifying or otherwise exploring transcript errors in medical student education is scarce.Materials and MethodsFirst-year preclinical medical students were surveyed. The semi-structured survey consisted of yes/no questions, slider type questions, 5-point Likert scales, and an open-ended question to assess usage of transcripts while watching lectures and perceptions of transcript errors. Survey data were interpreted using univariate analyses and inductive open coding. The authors also audited lectures for accuracy of transcripts. Two-sample t-tests were performed to compare transcript error rates.ResultsQuantitative survey results from 85 students revealed that 94% of participants used transcripts when consuming lectures, 96% noticed errors, and 59% felt that errors negatively impacted their learning. An audit of 36 lectures from two curricular blocks revealed that transcript accuracy failed to meet industry standards (i.e., >= 99% accuracy) in 35% of clinical science lectures (e.g., the diagnosis/management of diseases) and 5% of basic science lectures (e.g., biochemistry).ConclusionAlthough medical students rely on accurate transcripts of didactic lectures for both accessibility and facilitation of learning, current lecture transcripts fail to consistently meet industry standards for accuracy and correlate with student dissatisfaction. Given medical schools deliver much of their lectures as video recordings, these findings represent an important gap in need of future study.