Examining the Self-Development Test for Race and Gender Fairness
Abstract
The Self-Development Test (SDT) was examined for gender and race fairness Three SDT versions with the largest male score advantage and three SDT versions with the largest White score advantage were selected for analysis. Potentially biased items were identified and analyzed. Item correlations with target construct, subject matter expert reviews, and the impact of removing all items showing large performance differences between subgroups were considered. The percentage of items showing large difference in subgroup performance ranged from 14 to 62 percent across the six SDT examined. However, few of these items showed a differential relationship with the target construct across subgroups and the subject matter experts could not identify the items that were more difficult for minority subgroups. Scoring the SDT after removing items with statistically significant differences did not generally eliminate the subgroup differences at the test score level. Although no support was found for race or gender bias in the SDT, differential assignments based on gender, along with SDT emphasis on material covered in some duty positions may give a performance advantage to males in some Military Occupational Specialties (MOS). An examination of assignment procedures for MOS showing large gender performance differences in SDT scores is recommended
Document Details
- Document Type
- Technical Report
- Publication Date
- Jul 01, 1994
- Accession Number
- ADA283528
Entities
People
- Jay M. Silva
Organizations
- U.S. Army Research Institute for the Behavioral and Social Sciences