A denominator changes the picture
In a hypothetical program, 50 participants come from each of two groups. If one group has 100 eligible people and the other has 1,000, the enrollment proportions are 50% and 5%.
Equal counts conceal a substantial difference in reach.
Participation is only one stage
People may enroll but receive different amounts of service, face different burdens, or have different opportunities to benefit. Examine relevant stages rather than stopping at the registration form.
| Stage | Possible equity question |
|---|---|
| Invitation | Who can be contacted? |
| Enrollment | Who can complete the process? |
| Delivery | Who receives a usable service? |
| Outcome | Who benefits, and who may be harmed? |
Avoid assuming the reason from the gap
A difference can identify a concern without explaining its cause. Investigate scheduling, eligibility, trust, cost, communication, and other relevant conditions with affected groups.
Keep uncertainty visible
Small subgroup samples can limit precise comparisons, and categories may conceal important within-group differences. Report what the data support without labeling an uncertain finding as either proof of fairness or proof of discrimination.
The goal is a service people can access and use meaningfully, with differences understood well enough to guide a practical response.
