The use of percentiles and standard deviations for cutoff values

Percentiles refer to the position of a child among a group of normal children ranked by size. For example, if 100 children of a given age and sex are lined up by height (stature), the one at the 10th percentile is among the smaller children, tenth from the bottom. Clinicians usually use percentiles because their meaning is straightforward. Many cutoff values are based on percentiles.

Some cutoff values are based on standard deviation. Standard deviation is a number that tells how far the data are from the average (mean). If the standard deviation of a distribution curve is large, then the data, in general, are far from the average; if the standard deviation is small, then the data are close to the average.

This graph shows a normal distribution of data. The mean (average) is in the middle.

The graph shows that the 5th percentile and 2 standard deviations below the man are close but not the same. The 5th percentile corresponds to 1.65 standard deviations below the mean; the 2.3 percentile corresponds to 2 standard deviations below the mean.

For more information about standard-deviation scores (Z-scores), see the module, Describing the Growth of Groups of Children. (This module is under development; the link will be posted when the module is available.)