Statistics and Percentiles in Anthropometry

If you are more than 95th percentile height or less than 5th percentile, we predict that some everyday objects do not always seem quite the right size for you.

Just as you were able to compare your height with others in a given population, it is also possible to make such comparisons for other dimensions such as arm length, hand grip span, seated elbow height etc. Those whose build puts them towards the one extreme or the other will be familiar with the problems of using awkward sized objects. This is probably because the designers of these things have chosen limits outside your height range in their design. If you are more "average" in height, and yet still experience problems with some objects, you may ask yourself whether anthropometry has been used in their design at all. So what do these percentiles mean?

Suppose you measure the heights of all the people who live on your street. You would expect some variety in their heights, and perhaps the occasional very tall or short person - but by and large the people will tend to be a "normal" or "average" sort of height. By measuring these neighbours you would understand the distribution of their heights and you could use this information to make a reasonable guess about the height distribution of all the people in your area or town. This is how the principles of statistics are used in anthropometry.

If we plot a graph, with height along the X axis, and the number of people who are that height as the Y axis, we see a bell-shaped, or "Gaussian" curve. This shows that most people are "about average" height, but a few are very tall or very short in comparison to everyone else. We can split the X axis up into sections which show what percentage of the population are that height (or more/less). These divisions are called percentiles. For instance, 5% of the population are smaller than or exactly the 5th percentile height. 95% of the population are smaller than or exactly the 95th percentile height (so only 5 percent are taller).

By using this concept, designers can specify in advance what proportion of a population they want their design to satisfy or fit, and can then use the appropriate percentile dimension to use in the design. This choice will be influenced by the importance of getting the design right. For a safety feature, such as the size of an aircraft emergency exit, it is important that everyone be able to use it, and so the designer will choose very strict limits, ensuring that the vast majority of the passengers will be able to escape. However, accommodating the extremes of a population can be very time consuming and expensive, and there are, after all, only a very few people who will benefit from the consideration. Because of this many people will use the 5th and 95th percentiles to define the range which they "accommodate" with their design. This will ensure that at least 90% of the population will be suited by the dimensions.

Need more data? Find out where you can get it for free.

Ansgar Kupper 

First published 2004

Last reviewed Jan 2008