Are there cracks in our foundation? An integrative review of diversity issues in job analysis
Citation. Strah, N., & Rupp, D. E. (2022). Are there cracks in our foundation? An integrative review of diversity issues in job analysis. Journal of Applied Psychology, 107(7), 1031–1051. https://doi.org/10.1037/apl0000989
Summary. Strah and Rupp review how diversity issues can enter the job-analysis process that underlies selection, performance management, compensation, and other HR systems. Their central argument is that job analysis is only a fair foundation when it captures work as experienced by all incumbents, not only majority-group norms or the perspectives easiest to elicit. The paper develops a framework for when demographic groups may have distinct work experiences and how job-analysis methods can obscure or distort those differences.
Key claims
- Job analysis is treated as a prerequisite for accurate, fair, and legally appropriate HR systems, but that inference depends on whether the analysis accurately captures the full performance domain for all employees.
- Employees from different demographic groups may have predictable differences in work experience because of job-relevant individual differences, stereotype- and backlash-shaped behavior, differential work assignments, and environmental or societal restrictions.
- Diversity-related inaccuracies can enter job analysis through organizational norms and stereotypes, impression management, unequal voice opportunities, demand effects, overreliance on a single source, language bias in archival materials, and majority effects when aggregating SME input.
- Representative sampling alone is not enough; averaging across all job-analysis responses can still reproduce majority perspectives and hide meaningful subgroup differences.
- Biased or incomplete job analysis can propagate into recruitment, selection, performance evaluation, promotion, and compensation because those systems inherit the job definition created upstream.
- The authors recommend more inclusive job-analysis practices, including bias-aware review of archival job materials, bottom-up data collection, rapport and confidentiality with SMEs, attention to voice and tokenism in panels, and analysis of demographic patterns in job-analysis data.
Evidence & limitations
- The article is an integrative conceptual review, not a new empirical test; its value is in organizing job-analysis and diversity literatures into a framework for future research and practice.
- The review focuses primarily on sex, race/ethnicity, disability status, and age, with less evidence on other identities and intersectional combinations.
- Many proposed mechanisms and recommendations require direct empirical testing in job-analysis settings before their effect sizes, boundary conditions, and practical tradeoffs are known.
- The paper is especially useful for construct validity because it moves fairness concerns upstream: the problem may not only be biased predictors or biased models, but an incomplete definition of the job and criterion domain.