In testing the contractor data on selection practices to determine whether they are race-neutral, COs must analyze the data by individual racial or ethnic categories. Usually, COs analyze the categories specified in the EEO-1 report or the racial and ethnic categories specified at 41 CFR 60-3.4. However, if the contractor provides more detailed data, the CO may conduct an analysis based on that case-specific grouping. As explained above, the CO must examine the availability of racial and ethnic groups in the geographic area or applicant pools and use statistical analysis to compare the selection rates of racial and ethnic groups to the rates that would be expected if the selection practices were neutral. COs must consider whether selection rates or identified shortfalls for each individual racial or ethnic group indicate discrimination, based on statistical or other analysis, or any other case-specific evidence.
If, during the desk audit or at a later point in the evaluation, the CO determines that multiple groups are affected by the same discriminatory policy or practice, or that multiple groups are favored by selection practices, the CO must consider whether to combine individual groups to determine the affected class or comparators. However, a CO must never combine racial and ethnic groups when the effect is to dilute the statistical indicator of discrimination for a particular racial or ethnic group. COs must also consult with their RSOL and DPO before determining whether to combine groups to ensure that doing so is consistent with applicable legal standards and best practices for statistical analysis.