Insight on Analyzing Small Groups from OFCCP Branch of Expert Services

The NILG virtual conference is underway and well-represented by OFCCP, law firms, industrial and organizational psychologists, labor economists, and federal contractor professionals. There was one session in which the OFCCP established a panel of representatives from the Branch of Expert Services (BES). The panelists included: Dr. Robert LaJeunesse, Bogong Li, Ryan Peterson, Andy Leu, Bret Phillips, and David Garber. These experts provided an exploratory stance on methodological techniques for reviewing small groups. They shared information and demonstrated several techniques for addressing small group analyses that contractors may adopt in evaluating their own selection and pay practices; however, it is important to note, the experts did not prescribe any particular method during their NILG session.

First, the experts provided examples of what constitutes a small group. Small groups may occur in several contexts: pay analysis groups (PAGs), demographic coverage (e.g., race/sex) within PAG, and hiring for niche positions or even for distinct hiring decisions (e.g., multiple requisitions) for the same job title. The concern is that typical statistical tests for a small group analysis may be less reliable and may not generalize as well as for large groups. Some techniques that were mentioned to combat this concern for small group analysis include, but are not limited to, the following:

  • Cohort review;
  • Fisher’s Exact Test (FET);
  • Rank Sum tests (e.g., Mann-Whitney U);
  • K-Nearest Neighbor (KNN); and
  • Multiattribute Utility Analysis (MAUT).

The presenters explored each technique and provided examples, advantages, and even potential problems with each approach.

Next, panelists pivoted to a discussion on systemic discrimination and whether “isolated indicators” can form the basis of an allegation. Scenarios were explored to demonstrate how a one-off indicator (e.g., 1 of 15 PAGs flagging with statistical significance) may be problematic, not be indicative of a systemic issue alone. This finding is exemplary of a “within” PAG finding. Given this, it was shared that a comprehensive pay equity approach is desired to allow for an evaluation with a broader lens (i.e., more than just a within PAG review). A means to this type of comprehensive review could mean the addition of a combination test (e.g. Fisher’s Combination Test).

As stated above, the OFCCP experts were providing this information as awareness and a means for contractors to expand their evaluation methods. It may lead to future policy change, but these considerations are not official policy at this time. This was an insightful orientation into various techniques that OFCCP is exploring to assess pay equity in a more holistic way. Please reach out to a DCI consultant for information on how we can support you in expanding efforts and looking at pay equity in a comprehensive manner.

By Keli Wilson, Principal Consultant at DCI Consulting Group

Stay up-to-date with DCI Alerts, sign up here:

Advice, articles, and the news you need, delivered right to your inbox.


Stay in the Know!