View Full Version : Re: Comparing Data for Closeness

G. Anthony Reina
03-27-2000, 09:01 AM
> Hello, I am examining the torque-time characteristics of bowlers in
> cricket. I wish to compare the torque curves of different fast bowlers by
> obtaining a measure of "closeness" or "similarity". How do I do this? For
> instance, if one is comparing two curves, a simple method would be to
> average the differences in magnitude between corresponding points, i.e.
> find the deviation between the curves. What are the standard ways of
> quantitatively comparing two or more curves in biomechanics?
> Rene Ferdinands

One standard way is to perform a two-sample Kolmogorov-Smirnov test
(2-sample KS test). It will use the maximum difference between the two
curves to tell you if they are from the same distribution or from
different distributions (i.e. are the curves the same or different). It
will also give you a p-value statistic (e.g. 99% confidence) if you want
more teeth in your quantitative analysis. Most standard stats packages
come with this test (e.g. SysStat).


> --
> ///////////////////////////////////////////////////
> // G. Anthony Reina, MD //
> // The Neurosciences Institute //
> // 10640 John Jay Hopkins Drive //
> // San Diego, CA 92121 //
> // Phone: (858) 626-2132 //
> // FAX: (858) 626-2199 //
> ////////////////////////////////////////////

To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l