Percentiles were originally developed in order to clarify the standing of large groups (more than 100 units). For instance, if there were a thousand people in the graduating class, this analysis allowed you to easily identify those highly capable individuals who scored in the top 10% (the 90th percentile).
A warning: Some percentile numbers are not absolute but based on approximate calculations. Here's why.
This system enables you to choose percentiles that are multiples of 10. That means there are exactly 11 percentile "slots" into which an individual may fit. These slots are: 0 (the lowest-rated person), 10, 20, 30, 40, 50, 60, 70, 80, 90, 100 (the highest-rated person).
If there are other than exactly 11 (or 110, or 1100) Subjects, the system must interpolate (calculate) an arbitrary percentile value for the individuals that do not fit exactly into a percentile slot.
If you have 6 Subjects in a group, and request the value for the 90 percentile, you will receive an estimate. It will be interpolated, based on the average of the ratings of the fifth and sixth highest-rated Subjects (the ones who exactly matched the 80 and 100 percentiles).
With a greater the number of Subjects, Subjects will fit more precisely into every percentile slot.
There are several widely-used methods of calculating percentiles. For large groups, all provide essentially the same results. For 360-degree feedback, we use the "Continuous" calculation method, because it provides the greatest degree of accuracy for small numbers of Subjects (less than 100).
