Quartile Coefficient of Dispersion

[zipadeedodah]

Am hoping someone might be able to help. I have used the quartile coefficient of dispersion (Q1-Q3 divided by Q3+Q1) in a piece of work, but don't actually see what it means. Have searched the internet for some info, but have had no luck. Can anyone exaplin what it is?

[Gnu]

The interquartile range (Q3 - Q1) gives a measure of the spread of the data. For example, if the diameter of small trees were measured, the interquartile range could be 0.3m - 0.1m = 0.2m. If large trees were measured, the interquartile range could be 2.3m -2.1m = 0.2m. Each of these answers is the same, but the large trees would appear to be of a much more uniform diameter.

Using this to calculate the quartile coefficient of dispersion (Q1 - Q3)/(Q3 + Q1) takes the size of the trees into account, so for the small trees it is 0.2/0.4 = 0.5, and the larger trees 0.2/4.4 = 0.045.

Hope this makes sense - the only example I came across on the internet was to do with the quality of asparagus stalks, where a uniform size is considered good so a low coefficient of dispersion is needed.

[Gnu]

If the work is to do with shares, you could link the coefficient of dispersion to risk. A high coefficient of dispersion would indicate a large fluctuation in share prices relative to their 'average' value and would be suitable for a high return, high risk strategy.

[zipadeedodah]

Thanks Gnu, that has helped alot! I've just been looking at one set of data, and I guess its only really useful for comparisons, as the value I got (0.023) doesn't mean much on its own.