Test records often report converted raw test scores as percentiles orpercentile ranks, because this statistic is relatively easy to interpret andprovides a way to rank all examinees on a large scale from 1% to 100%.

Apercentile rank describes the percentage of people in the comparison group whoscored below a particular score.

For example, if a student's raw score of 62 corresponds to a percentile rankof 98, that student performed better than 98% of the other test-takers.

Referring back to the normal curve, about 98% of the area of the curve would beto the left of this score and about 2% would be to the right.

If a rawscore of 42 is the mean of this distribution of scores, then 42 is the 50thpercentile. The raw mean score is always the 50th percentile.

Educators can determine which scores correspond to a particularpercentile by relating percentile ranks to the normal curve.

If a testhas a mean of 42, and a SD of 10, a score of 52 (+1 SD) is at the 84.13 percentile (50% + 34.13% =84.13%). Similarly, a score of 22 (-2 SD) is at the 2.28 percentile (50% -34.13% - 13.59% = 2.28%). These are sometimes called cumulativepercentages with -3 SD equal to 0.09% and +3 SD equal to 99.91% of thedistribution.