-------- Original Message -------- Subject: Re: Mahalanobis distance Date: Mon, 14 Apr 2008 04:46:49 -0700 (PDT) From: Joseph Kunkel <[EMAIL PROTECTED]> To: morphmet@morphometrics.org CC: Joseph Kunkel <[EMAIL PROTECTED]> References: <[EMAIL PROTECTED]> Mahalanobis D squared is essentially a significance test. One can compute it by doing a principle components analysis and dividing the eigenvector lengths by their eigenvalues and summing them. This represents a sum of independent F-tests. It is like using a squared t- test to measure a difference. As with such measures of significance, they explode when there is a substantial difference between two groups. Because it is a sum of ratios it suffers from the problems with ratios particularly when the denominator is unstable such as when the underlying variances are estimated by small numbers. Joe Kunkel [EMAIL PROTECTED] On Apr 14, 2008, at 3:05 AM, morphmet wrote:
-------- Original Message -------- Subject: Mahalanobis distance Date: Thu, 10 Apr 2008 04:21:04 -0700 (PDT) From: Saber Sadeghi <[EMAIL PROTECTED]> To: morphmet@morphometrics.org Dear morphometriceans I am a student and work on population isolation. now, I'm confusing about using Squared Mahalanobis distance in geometric morphometrics analysis. Although some researchers used it in papers, some others forbid it. I would appreciate it if you could explain me about that. Best Regards Hamid __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com -- Replies will be sent to the list. For more information visit http://www.morphometrics.org
-·. .· ·. .><((((º>·. .· ·. .><((((º>·. .· ·. .><((((º> .··.·
=- =º}}}}}><
Joseph G. Kunkel, Professor Biology Department University of Massachusetts Amherst Amherst MA 01003 http://www.bio.umass.edu/biology/kunkel/ -- Replies will be sent to the list. For more information visit http://www.morphometrics.org
