**Online RDM Education During a Global Pandemic: Trends and Lessons Learned**
The Harvard Medical School Countway Library’s Massive Open Online Course (MOOC) Best Practices for Biomedical Research Data Management (RDM) launched on Canvas in January 2018 to provide free RDM training to a broad audience. When the COVID-19 pandemic hit in March 2020, student registration almost doubled in less than a month. Previous reports analyzed enrollment periods pre-pandemic and through July 8, 2020. This analysis looks deeper into subsets of student data during the pandemic to understand global interest in and knowledge of biomedical research data management.
**Objectives**: The Harvard Medical School Countway Library’s Massive Open Online Course (MOOC) Best Practices for Biomedical Research Data Management (RDM) launched on Canvas in January 2018 to provide free RDM training to a broad audience. When the COVID-19 pandemic hit in March 2020, student registration almost doubled in less than a month. Previous reports analyzed enrollment periods pre-pandemic (Trepanowski 2019) and through July 8, 2020 (Fay & Goldman 2021). This analysis looks deeper into subsets of student data during the pandemic to understand global interest in and knowledge of biomedical research data management. Research questions include:
1. Where are the learners in the course coming from? How do the demographics of students compare to global COVID numbers?
2. How did students find out about the course?
3. What percentage of students have completed the course? How long did it take students to complete the course?
4. What are student's motivations for enrolling in the course?
5. What type of professionals and nonprofessionals participate in the course, (i.e. students, researchers, instructors, librarians, etc.)?
**Methods**: Raw data from the Welcome Survey, Course Assessment Survey, and User Experience Survey covering students who enrolled in the course Best Practices in Biomedical Research Data Management over the period from March 2020 through June 2021, was exported from the Canvas Network site and deidentified. Data were further subdivided into three periods, March 16 – May 9, 2020 (“Spring 2020”), November 30, 2020 – February 1, 2021 (“Winter 2020”), and March 28 – June 14, 2021 (“Spring 2021”). Data contain unique participant identifiers allowing for response and assessment data for individual students to be linked across course modules and surveys. For qualitative analysis to evaluate learner’s personal and professional goals, open ended responses were categorized into motivations and characteristics. Categories were identified and standardized for manual tagging. All analysis was performed in Microsoft Excel.
**Results**: Pre-pandemic, students enrolling in the Best Practices in Biomedical Research Data Management course were primarily from North America. In Spring 2020, the percentage of students from Europe and Asia/Pacific both increased relative to pre-pandemic trends. The increase in students from Asia/Pacific peaked in Winter 2020 at 68.5% and remained a majority through Spring 2021. Students from Latin America hovered around 5% through Winter 2020 but jumped to 16.5% in Spring 2021. Comparing enrollment trends with worldwide COVID case data from the New York Times shows a correlation between COVID cases and course enrollment numbers and student locations. Over all three periods, at least 46% of students self-reported a language other than English being their primary spoken language.
The completion rate for the course increased steadily from 17% in Spring 2020 to 45% in Spring 2021. The mean time to completion also decreased from 17.5 days in Spring 2020 to 2.4 days in Spring 2021. Over the periods analyzed, the method through which students heard about the course shifted from word of mouth or other direct communication to web searches with the latter increasing from 17.07% in Spring 2020 to 29.73% in Spring 2021.
Across all three periods, learners consistently enrolled in the course to improve their skills for a future career, current career, or current course work. In Spring 2020 and Spring 2021, many learners also mentioned a personal interest for enrolling in the course. Learners consistently identified as students, with other top professions being Researchers/Scientists, Health Care Professionals, Instructors, and Doctors. Spring 2020 saw more geographic diversity in students across these motivational categories. In Winter 2020 and Spring 2021, the geographic distribution of students across motivational categories aligned with the overall geographic distribution of students.
**Conclusions**: This course was designed primarily for students in the United States and teaches data management guidelines within the US regulatory framework. Over the period analyzed, the student demographics shifted to primarily students outside of the United States. As a result, some students did not have their information needs met. A lesson learned for future course design is to explicitly state the scope and perspective of the course at the start so that students can make informed enrollment decisions, which is especially important when courses are easily discoverable via web searches and spread beyond the original audience.
In addition to the course reaching a broader geographic audience than originally intended, the course also drew in a wider variety of learners. The course description did outline the recommended educational and professional background for students, so this student population knew at the point of enrollment that they were not the target audience. Motivations for this group of learners included topics such as improving English-language vocabulary related to research data management and studying online pedagogy. This group of students serve as a valuable reminder that any time content is made free and widely available, it can be used and reused in ways the developers did not originally intend, including to meet information needs of populations who were left out of the original scope.
The analysis that could be performed on the student data from this course was necessarily limited by the limitations of the survey design. All the surveys used to generate the data sets were created by the Canvas Network platform and could not be modified by instructors. As a result, some data is overly broad and combines too many populations under a single umbrella. For example, since the survey question about geographic location included only "Asia/Pacific", it was not possible to separate out trends in enrollment from East Asia vs South Asia despite a hypothesis that enrollment trends between the two regions differed.
