K H Vincent Lau1, Pue Farooque2, Gary Leydon3, Michael L Schwartz4,5, R Mark Sadler6, Jeremy J Moeller2. 1. a Department of Neurology , Massachusetts General Hospital , Boston , MA , USA. 2. b Department of Neurology , Yale School of Medicine , New Haven , CT , USA. 3. c Department of Teaching and Learning Center , Yale School of Medicine , New Haven , CT , USA. 4. d Department of Neurobiology , Yale School of Medicine , New Haven , CT , USA. 5. e Office of Education , Yale School of Medicine , New Haven , CT , USA. 6. f Department of Medicine, Division of Neurology , Dalhousie University , Halifax , Canada.
Abstract
BACKGROUND: The video-based lecture (VBL), an important component of the flipped classroom (FC) and massive open online course (MOOC) approaches to medical education, has primarily been evaluated through direct learner feedback. Evaluation may be enhanced through learner analytics (LA) - analysis of quantitative audience usage data generated by video-sharing platforms. METHODS AND RESULTS: We applied LA to an experimental series of ten VBLs on electroencephalography (EEG) interpretation, uploaded to YouTube in the model of a publicly accessible MOOC. Trends in view count; total percentage of video viewed and audience retention (AR) (percentage of viewers watching at a time point compared to the initial total) were examined. The pattern of average AR decline was characterized using regression analysis, revealing a uniform linear decline in viewership for each video, with no evidence of an optimal VBL length. Segments with transient increases in AR corresponded to those focused on core concepts, indicative of content requiring more detailed evaluation. We propose a model for applying LA at four levels: global, series, video, and feedback. DISCUSSION AND CONCLUSIONS: LA may be a useful tool in evaluating a VBL series. Our proposed model combines analytics data and learner self-report for comprehensive evaluation.
BACKGROUND: The video-based lecture (VBL), an important component of the flipped classroom (FC) and massive open online course (MOOC) approaches to medical education, has primarily been evaluated through direct learner feedback. Evaluation may be enhanced through learner analytics (LA) - analysis of quantitative audience usage data generated by video-sharing platforms. METHODS AND RESULTS: We applied LA to an experimental series of ten VBLs on electroencephalography (EEG) interpretation, uploaded to YouTube in the model of a publicly accessible MOOC. Trends in view count; total percentage of video viewed and audience retention (AR) (percentage of viewers watching at a time point compared to the initial total) were examined. The pattern of average AR decline was characterized using regression analysis, revealing a uniform linear decline in viewership for each video, with no evidence of an optimal VBL length. Segments with transient increases in AR corresponded to those focused on core concepts, indicative of content requiring more detailed evaluation. We propose a model for applying LA at four levels: global, series, video, and feedback. DISCUSSION AND CONCLUSIONS: LA may be a useful tool in evaluating a VBL series. Our proposed model combines analytics data and learner self-report for comprehensive evaluation.
Authors: Amr Salama; Janet Panoch; Elhaam Bandali; Aaron Carroll; Sarah Wiehe; Stephen Downs; Mark P Cain; Richard Frankel; Katherine H Chan Journal: J Pediatr Urol Date: 2019-12-04 Impact factor: 1.830