What is a percentile in a selection test

A percentile places a result against a reference group: it tells you what percentage of that group scored equal to or lower than a given result. Being in the 75th percentile means the person matches or beats roughly 75% of that group—not that they answered 75% of the questions correctly or that they “know 75%.” It’s a relative position, not a measure of how much someone masters something, and that changes how it should be read.

A test report comes in and says “75th percentile.” Almost everyone reads it as a grade: “they scored 75 out of 100.” It’s the most natural interpretation, and also the wrong one. Understanding what a percentile really is changes the way you read any selection result, because it avoids the trap of treating a relative position as if it were a grade.

A percentile is not the percentage of correct answers

The key confusion: the percentage of correct answers measures how many questions a person got right out of the total; the percentile compares them with other people. They’re different things, and sometimes they point in opposite directions.

An example: on a hard test where almost no one answers much, someone can get relatively few questions right and still land in a high percentile, simply because everyone else did worse. And the reverse: on an easy test, getting plenty right can leave you in a low percentile if nearly everyone got more. The percentile doesn’t talk about the test; it talks about your place relative to the group.

The reference group is everything

A percentile means nothing without knowing who it’s calculated against. That comparison set is called the normative group, or norm. The same raw score can land in the 60th percentile against a general group and in the 80th against a more specific one.

That’s why, when interpreting a percentile, the first question is: which population is this being compared against? A result read against a norm that doesn’t fit the role or context can mislead. When the reference group is clear and relevant, the percentile becomes a useful, comparable way to place a candidate.

How to read it well in selection

• Treat it as a position, not a grade. Ask “where do they rank?”, not “what score did they get?”.
• Ask for the reference. A percentile without a clear normative group says little.
• Don’t turn it into an automatic cutoff. A high or low percentile is one input, not a verdict; a low percentile on one test doesn’t rule someone out on its own. For the same reason, it’s worth not using cognitive tests as an exclusionary filter.
• Read it in context. Combine it with competencies, experience, and the interview before drawing conclusions.

In summary

A percentile places a result against a reference group: the 75th percentile means matching or beating roughly 75% of that group, not “scoring 75 out of 100” or “knowing 75%.” It’s not the same as the percentage of correct answers, and its meaning depends entirely on the normative group it’s compared against. Read well, it’s a comparable way to place a candidate; read poorly, it becomes a false grade or an unfair cutoff. It’s an input for the team’s judgment, not a verdict. At Kokoro, results are delivered to support the decision, not to make it; browse the library to see how they’re reported, or learn how Kokoro supports that decision.