This R package provides a suite of functions for assessing adverse impact in employee selection and hiring procedures. Adverse impact occurs when a selection procedure disproportionately excludes members of a protected class, such as race or gender. The package includes functions to compute statistical tests, such as the Z-test and LMP test, as well as practical measures, like the h-statistic, Phi-coefficient, and adjusted shortfall.
By offering a comprehensive set of tools, this package helps organizations evaluate their selection procedures and make data-driven decisions to minimize adverse impact and promote fairness in the hiring process.
- Human resources professionals evaluating the fairness of their selection processes.
- Organizations aiming to reduce adverse impact and promote equal opportunity in their hiring practices.
- Researchers studying the impact of selection procedures on different demographic groups.
You can install the development version of advserseimpact from GitHub with:
# install.packages("devtools")
devtools::install_github("GGLuca/adverseimpact")
The data in this example stems from Morris (2017). In this selection scenario, we take into account the disparity between male and female hires, with women being designated as the minority group. Out of 35 applicants, 10 are women and 25 are men. Among the female candidates, 2 are chosen and 8 are not, while 15 of the male candidates are selected and 10 are not. Therefore, a total of 17 candidates are chosen and 18 are not.
Group | Fail | Pass | Total |
---|---|---|---|
Women | 8 (NF_min) | 2 (NP_min) | 10 |
Men | 10 (NF_maj) | 15 (NP_maj) | 25 |
Total | 18 (NN_t) | 17 (NP_t) | 35 |
library(adverseimpact)
## basic example code
# Z (Two Standard Deviation) Test (Formula 4.2)
ai_ztest(8, 2, 10, 15)
#> z_score p_value
#> 1 -2.138963 0.03243865
-
Morris, S.B. (2017). Statistical Significance Testing in Adverse Impact Analysis. In S.B. Morris & E.M. Dunleavy (Eds.), Adverse Impact Analysis: Understanding Data, Statistics, and Risk (pp. 71-91). Routledge.
-
Nottingham, A., Iliescu, D., & Rupp, D.E. (2023, May 24-27). Adverse Impact Statistics across Selection Stages: A Monte Carlo Simulation [Poster presentation]. European Association of Work and Organizational Psychology, Katowice, Poland.
-
Nottingham, A., Iliescu, D., & Rupp, D.E. (2023, April 19-22). Choosing the “Right” Adverse Impact across Selection Stages: A Monte Carlo Simulation and Recommended Best Practices [Poster presentation]. Society for Industrial and Organizational Psychology Conference, Boston, MA, United States.
-
Oswald, F.L., Dunleavy, E.M., & Shaw, A. (2017). Measuring practical significance in adverse impact analysis. In S.B. Morris & E.M. Dunleavy (Eds.), Adverse Impact Analysis: Understanding Data, Statistics, and Risk (pp. 92-112). Routledge.