Percentile vs Percentile Rank

The $k$th percentile is the value at or below which about $k$ percent of the observations lie. The grammar of that sentence is worth slowing down for: $k$ is the percent, and the percentile is the value, carrying whatever units the data carry. If the 90th percentile of a test is 628 points, then 90 percent of scores are 628 or lower and about 10 percent are higher.

Run the definition backwards and you get the percentile rank, which starts from a value and returns a percent:

Take the ten scores 52, 58, 60, 64, 65, 69, 73, 79, 86, 91. Seven of them are 73 or lower, so the percentile rank of 73 is $\frac{7}{10} \times 100 = 70$, and 73 is the 70th percentile of this set. Move the same score into a different group and its percentile rank changes, because a percentile always describes a position within one particular set of data.

Percentiles work on a smooth distribution too, where they are read off the curve instead of counted. On a normal distribution with mean 500 and standard deviation 100, the 90th percentile sits about 1.28 standard deviations above the mean, at $500 + 1.28(100) = 628$ points. That is the same 628 from the opening paragraph, and it shows the two routes to a percentile: counting observations in a data set, or finding the area to the left under a density curve.

The misreading that costs the most marks is treating a percentile as a score. Standing at the 85th percentile is a claim about the other test-takers, not about how many questions you answered correctly. On a hard test you can get 40 percent of the items right and still sit at the 85th percentile of it.

Three percentiles have names of their own: the 25th is the first quartile $Q_1$, the 50th is the median, and the 75th is the third quartile $Q_3$. Percentile and z-score are the two standard ways of reporting relative standing, and they answer different questions. A percentile says how many values sit at or below yours; a z-score says how far above or below the mean yours is, measured in standard deviations.

A percentile rank runs from a value to a percent. On this site it counts the values at or below $x$ and divides by $n$, the number of observations, then multiplies by 100. The answer is a percent with no units, and it describes where a value sits inside one particular group rather than anything about the value itself.

Suppose 22 of the 25 students in a class scored at or below Maya's 84. Her percentile rank is $\frac{22}{25} \times 100 = 88$. Ties come along for the ride. In the five scores 70, 70, 70, 80, 80 each of the three 70s has three values at or below it, so all three have percentile rank $\frac{3}{5} \times 100 = 60$, and both 80s land at 100.

"I am at the 88th percentile, so I beat 88 percent of the class" is close, and not what the number says. The count is at or below, which includes Maya herself and anyone tied with her, so the honest reading is that 88 percent of the class scored 84 or lower. In a class of 25 the gap is one student. Where scores are heavily tied it is not small: the three students at 70 beat nobody and still carry a rank of 60.

Two consequences follow from counting at or below. The largest value always has a percentile rank of 100, and no value ever has a rank of 0, because every value counts itself: the lowest of 25 scores comes out at $\frac{1}{25} \times 100 = 4$. Other sources count only the values strictly below, or those below plus half the ties, and they give different numbers on the same data, so say which count you used.

A percentile rank is meaningful only against a stated reference group. Move that same 84 into a stronger class and the rank drops, though nothing about the score changed. On a smooth distribution the counting stops and the rank becomes an area under the curve, which is how a z-score converts into a percentile once the distribution is close enough to normal.