What is the four-fifths rule?

It is the EEOC's rule of thumb: if a protected group's selection rate is less than four-fifths (80%) of the highest group's rate, that is treated as evidence of adverse impact warranting scrutiny. It is a screening heuristic, not a legal verdict.

Does a passing ratio mean the system is unbiased?

No. The four-fifths rule can miss real disparities in small samples and pass systems that are still biased in other ways. A passing result means no adverse impact was detected on this metric in this sample — nothing more.

What sample size do I need for this to be meaningful?

Selection-rate ratios are unstable with small groups. With fewer than roughly 30 decisions per group, treat any flag as a prompt to gather more data rather than a conclusion. The tool warns when groups are small.

Is the data sent anywhere?

No. All parsing and calculation happen in your browser. Nothing is uploaded, and the data never leaves the page.

Can it analyse ratings instead of yes/no decisions?

This version analyses binary outcomes (selected/not selected, promoted/not). For continuous ratings, threshold them into a decision first — for example, treat ratings above the median as the favourable outcome — then paste that.

AI HR Decision Output Bias Checker

AI HR decision output bias checker

When an AI screens CVs, ranks candidates, or recommends promotions, the same disparity tests that apply to human decisions apply to it — and regulators increasingly expect employers to run them. This checker takes your AI-generated decisions, computes the selection rate for each demographic group, and applies the standard four-fifths (80%) rule to flag groups that may be experiencing adverse impact. Everything runs locally in your browser.

How it works

You paste one row per candidate: a group label (gender, age band, or any protected characteristic) and a decision (selected or not). The tool counts favourable outcomes per group, divides by group size to get each group’s selection rate, then divides every rate by the highest group’s rate to get the adverse impact ratio. A ratio below 0.80 is the EEOC’s threshold for potential adverse impact and gets flagged. Because small groups produce unstable ratios, the tool also warns when any group has too few decisions to trust the result.

Tips and notes

The four-fifths rule is a screen, not a verdict. A flag means investigate; a pass means no impact detected on this metric in this sample.
Mind sample size. Under ~30 decisions per group, ratios swing wildly — collect more data before drawing conclusions.
One characteristic at a time. Run gender, then age, then intersections separately; pooling hides group-specific patterns.
Document your testing. Keeping a record of these checks is itself part of defensible, transparent AI hiring practice.