Percentile: Meaning in test results

What a percentile means

A percentile describes a position. It shows where an individual result falls within a comparison group. If a result corresponds to percentile rank 90, about 90 percent of the comparison group scored the same or lower, and about 10 percent scored higher. Percentile rank 50 is the middle of the distribution, the value that divides the group into a lower half and an upper half.

The advantage of this format is that you can understand it without technical background knowledge. You do not need to know how a raw score is calculated to understand the statement "scores higher than most of the comparison group." This is why percentiles are often used when results need to remain readable for a general audience. The recognized test standards from AERA, APA and NCME describe percentile ranks as a common way to relate raw scores to a reference group so that an individual result can be assessed relative to that group.

Percentile and percentage score are often confused

The most common confusion is between a percentile and a percentage score. Both sound similar and use a number between 0 and 100, but they mean very different things.

A percentage score refers to the task itself. If you answer 18 out of 20 questions correctly on a test, you answered 90 percent of the questions correctly. This is a proportion of correct answers and says nothing about how other people performed. A percentile, by contrast, refers to other people. Percentile rank 90 means the result is higher than the results of most people in the comparison group, regardless of how many questions were answered correctly. You could answer 90 percent of the questions correctly and still be at percentile rank 40 if most of the comparison group performed even better. Conversely, a low percentage of correct answers can correspond to a high percentile rank if the questions were difficult for everyone.

The percentile depends on the comparison group

A percentile rank cannot be assessed without its comparison group. This group is called the norm sample or reference group. It defines who the result is compared with. The same raw score can produce very different percentile ranks depending on the comparison group.

A result may be compared, for example, with a specific age group, a specific region or a specific population. If a score is compared with a high-performing selected group, the percentile rank will be lower than when it is compared with the general population. A properly reported percentile should therefore state what it refers to. When was the norm sample collected, how large was it, and which age group and region does it apply to. Without this information, a percentile rank is a number without a meaningful reference. For the same reason, percentiles from different reference groups cannot be compared directly without further information.

A simple example

The following example is hypothetical and is only for illustration. It does not come from a study. Suppose 100 people take the same test. One person gets a raw score that is higher than the scores of 80 of those 100 people and the same or lower compared with the remaining people. This result corresponds approximately to percentile rank 80. In this comparison group, the result falls in the upper range, and about four out of five people did not score higher.

What the example does not say is just as important. It does not mean that 80 percent of the questions were answered correctly. It also says nothing about another comparison group. If the same person were compared with a group that scored much higher on average, the same raw score could correspond to a lower percentile rank.

Percentiles in IQ tests

IQ results are often also reported as percentile ranks because this makes comparisons easier. An IQ score is compared with a norm sample, and the percentile rank shows where the result falls relative to that group. We explain in detail how IQ scores, standard deviation and percentiles are related, and what a result in the upper range means, on our page about giftedness. We also discuss there why a single test score should always be read with caution. For a more detailed explanation, see Am I gifted.

What this means for self-tests on medtests.net

Not every test on medtests.net uses percentiles. Many screening questionnaires use a cutoff instead. These tests are not about your position within a comparison group. They ask whether a total score reaches a specific threshold at which further assessment may be useful. A cutoff is different from a percentile rank and should not be confused with one.

Comparison values relative to a group are more common in personality tests. With the Big Five Test , for example, there are no right or wrong answers. Instead, the test describes how strongly particular traits are expressed, often compared with a reference group. The same applies to descriptions of Personality types . There is no pass or fail result. Instead, the result describes how strongly different characteristics are expressed. When we provide comparison values, we state the reference group and use the result as a point of reference, not as a ranking of people.

What the term does not mean

A percentile rank is not a percentage score and is not a judgment of a person's worth. It only describes a position within a specific comparison group and cannot be assessed without that group. A high or low percentile rank on a self-test is not a diagnosis. A self-test provides orientation and does not replace medical, psychotherapeutic or diagnostic assessment.