Back

Disability disparities in STEM: Gaps in salaries and representation for doctorate recipients with disabilities in the U.S., 2019

Castro, F.; Stuart, E.; Deal, J.; Varadaraj, V.; Swenor, B. K.

2022-12-06 epidemiology
10.1101/2022.12.04.22283081 medRxiv
Show abstract

IntroductionThere is paucity of data examining disparities in salary and representation for disabled scientists, which is needed to advance inclusion and equity for people with disabilities in Science, Technology, Engineering and Mathematics (STEM). MethodsAnalyses used cross-sectional data from the 2019 Survey of Doctorate Recipients. We compared salaries between doctorate recipients with and without disabilities who were currently employed in STEM (N = 704,013), but who were otherwise similar on socioeconomic, degree and job-related characteristics, using propensity score weighting to carry out balanced comparisons, and further examined these differences in salary in the subset of doctorate recipients working in STEM at academic institutions (N = 219,413). In the subset of participants working in academia, we examined whether the representation of people with disabilities differed across categories of academic career milestones using chi-square tests ( = 0.05). ResultsDoctorate recipients working in STEM with early onset disabilities (identified <25 years of age) earned $10,580 less per year than non-disabled workers, and in the subset of academic workers this difference was larger (-$14,360). Salaries appeared lower for people with late onset disabilities as compared to those without, although these differences did not reach statistical significance. We observed an underrepresentation of academics with disabilities at higher faculty ranks (p<0.0001), among Deans/Presidents (p<0.0001) and among those with tenure (p: 0.0004). ConclusionThese findings support a need to expand efforts to foster inclusion, provide equal opportunities for career advancement, and improve working conditions for people with disabilities in STEM.

Matching journals

The top 10 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.