Teaching / education
Statistical software skills for graduate-level jobs in epidemiology: An analysis of U.S. job posting data Emily Goldmann* Emily Goldmann Ruby Barnard-Mayers M. Maria Glymour Megan A. Healey
Background. Developing competency in statistical software is a core component of graduate-level epidemiology training. The most useful software packages to introduce in coursework is debated. Identifying the software skills most desired by potential employers may help guide this selection.
Methods. This study used data from Lightcast, a data vendor that collects, deduplicates, and codes job posting data scraped from websites globally. We conducted a search of unique U.S. job postings from January 2010-December 2024 that included a preference or requirement for a master’s degree using “epidemiology” as a keyword (n=188,962). We examined yearly trends in the proportion of postings that mention specific software skills, overall and for the five most common job industries represented in these postings (academic 18%, pharmaceutical 11%, hospital 10%, government 9%, insurance 7%). We also analyzed job description text for 500 sample postings to evaluate overlap in demand for SAS Software and R Programming Language skills.
Results. Of the 15,950 unique postings from 2024, SAS was mentioned in the greatest proportion (28%), followed by R (23%), SQL (18%), Python (15%), and Microsoft Excel (14%). Of the 500 job descriptions that mentioned SAS or R (n=139, 28%), 58% mentioned both SAS and R, 29% mentioned SAS but not R, and 13% mentioned R but not SAS. Mentions of SAS decreased in proportion in recent years (Figure 1; from 42% in 2021 to 31% in 2023 and 28% in 2024), while mentions increased from 2010 to 2024 for R (3% to 23%), Python (1% to 15%), and SQL (9% to 18%). In 2024, SAS was the most commonly mentioned software skill in insurance (61%), government (28%), and hospital (22%) postings, while R was most commonly mentioned in pharmaceutical (23%) and academic (21%) postings.
Conclusion. While SAS skills remain frequently desired in epidemiology job postings, R has increased substantially in demand, particularly in academia and the pharmaceutical industry.
