Building a Workforce for Data-Driven Research and Health: The National Library of Medicine Data Science Training Program

Background: The National Library of Medicine’s (NLM) 2017 – 2027 Strategic Plan outlines three broad goals, one of which is to “build a workforce for data-driven research and health.” Realizing this vision requires NLM staff to also develop their own data science skills and expertise. The Data Science @NLM (DS @NLM) Training Program was a year-long initiative to provide staff opportunities to enhance their skills and knowledge related to data science. The program’s main goal was not to turn all NLM staff into data scientists, but to provide a common vocabulary and understanding of data science principles regardless of job title.

Description: The DS @NLM Training Program included NLM-wide activities, including a Data Science Basics training session, individual training plans (ITP) for all staff, and a pilot intensive training course for selected staff. Staff members selected one of eight personas developed to cover the skills needed in various roles. The ITPs were developed by identifying gaps between the staff member’s knowledge as self-reported in a Data Science Readiness Survey and the skill level of the selected persona. Staff members were then provided a list of courses tailored to their unique needs from a catalog of nearly 250 courses. The year-long initiative culminated with a Data Science Open House, where staff shared how they had used new data science skills and considered how they could continue applying data science to NLM’s work.

Conclusion: The DS@NLM Training Program helped to increase staff’s understanding of data science and why it is important to NLM’s mission. NLM leadership and supervisor engagement played a role in staff success; staff who reported receiving support from their supervisor in selecting their persona and courses were almost seven times more likely to have made progress on or completed their ITP. Feedback suggested ongoing interest, with nearly half of respondents to the post-program survey indicating interest in further intensive, hands-on data science training. Future potential activities include mentoring to help connect NLM staff with their colleagues to learn about data science.