You probably heard of percentiles, or percentile rank, before. Percentile is the same as a relative cumulative frequency, or the percent of data points in, or below, some other bin of data.
Often, large data sets are given in frequency tables, frequently with rounded values.
EXAMPLEHere is a table showing heights (in inches) of 333 sixth-grade students, along with the frequency, relative frequency, and relative cumulative frequencies.
|Height||Frequency||Relative Freq||Rel. Cum. Freq|
How do we read this? Notice the first two rows have a relative frequency of 3% and 7%, respectively. Using these values, we can find the relative cumulative frequency of 10% in the second row by combining these two relative frequencies. You can also check this by dividing the cumulative amount of 11 and 23 students, which is 34, by the total amount of students, 333, you'll get a number close to 10%.
By the time we get to 65 inches, we will have accounted for all of the sixth graders in the data set.
|Which percentile will a student with a height of 62 inches fall into?||From the chart, you can see that 62 inches falls in the 85th percentile. That means that a 62-inch student is at least as tall as 85% of his/her classmates.|
|How tall is a student in the 94th percentile?||They would be 63 inches tall.|
|What is the median height for sixth graders?||
This question is a little tricky. By the time you finish counting up through all of the 59-inch students, you still haven't accounted for half the grade yet; only 47% of the students. However, by the time you're done counting all the 60-inch students, you've accounted for 62% of the grade.
That means that somewhere within that 60-inch range is the median height. So this tells us that half the students are at or above 60 inches, and half the students are at or below 60 inches.
Source: Adapted from Sophia tutorial by Jonathan Osters.