Adverse impact is something that comes up a lot in hiring, but it's one of those terms that can feel a bit abstract until you see it in action. Put simply, it's what happens when a selection process ends up rejecting candidates from a particular group at a noticeably higher rate than others.
In hiring, we're usually most concerned about this when it affects legally protected groups, things like ethnicity, gender, age, or sexuality, because discrimination on those grounds isn't just unfair, it's illegal.
The easiest way to get your head around it is with an example. Say you're hiring for a graduate scheme and you interview 100 candidates, split evenly between men and women. If all 50 men get offers and all 50 women are rejected, that's a pretty clear-cut case of adverse impact on the basis of gender. But adverse impact isn't always that obvious.
Take ethnicity in the same process. If 80 candidates are white and 20 are from ethnic minority backgrounds, and you make 40 offers to white candidates and 10 to ethnic minority candidates, that's actually proportionally equal. Both groups had a 50% selection rate, so on paper, there's no adverse impact. So it's not just the end numbers that matter, but the proportion.
The four-fifths rule
So how do you actually measure it? The most widely used starting point is the four-fifths rule (sometimes called the 80% rule). It says that if your selection rate for a minority group drops below four-fifths of the selection rate for your majority group, that's a flag worth taking seriously. So if you're selecting 50% of white candidates, you'd want to be selecting at least 40% of ethnic minority candidates to stay within that threshold.
It's a useful rule of thumb, but it's not the whole picture. Statistical tests like chi-square analysis can tell you whether a gap in selection rates is large enough to be meaningful, rather than just the result of small sample sizes. And for psychometric assessments specifically, Cohen's d is often used to measure the size of score differences between groups, which gives you a more precise sense of whether a test is contributing to unequal outcomes.
None of these methods will give you all the answers on their own. Adverse impact analysis works best when you're looking at multiple angles, including the numbers, the statistical significance, and the broader context of your hiring process.
Conclusion and next steps
If you're not already keeping an eye on selection rates across different groups, that's the place to start. That's something we take seriously at Test Partnership. Every assessment we develop goes through adverse impact analysis as part of our R&D process, so by the time a test reaches you, we already have a clear picture of how it performs across different groups.
Ben Schwencke
Chief psychologist at Test Partnership. MSc in Organisational Psychology with over ten years experience in psychometric testing.
