Date:         Thu, 14 Dec 2006 14:59:01 -0800
Reply-To:     David L Cassell <davidlcassell@MSN.COM>
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         David L Cassell <davidlcassell@MSN.COM>
Subject:      Re: Computing the value KS statistic
In-Reply-To:  <1166040969.309896.188640@j72g2000cwa.googlegroups.com>
Content-Type: text/plain; format=flowed

mtamboli@GMAIL.COM wrote:
>
>Hi,
>
>I am using the KS statistic to determine how well the model scores
>(credit card world).
>
>I am confused on the way the KS value is computed.
>
>I read in one of the earlier SAs post that "The actual KS statistic is
>the maximum difference between the two disitribution values multiplied
>by sqrt(N1*N2/(N1+N2)), where N1 and N2 are the
>sizes of the two population". However, I also read an internal document
>where the KS value is
>determined to be the maximum difference between the two distribution
>values (edf).
>
>Any help in this matter is really appreciated.
>
>Thanks,
>
>Mahesh

Both are called 'the Kolmogorov-Smirnov statistic'.  It depends which
book you are reading.

If the reference is to the Kolmogorov-Smirnov D, then that is
typically the unscaled max absolute difference.  Then that is
typically multiplied by a constant - the constant you are talking
about - in order to get the classical asymptotic approximation.

HTH,
David
--
David L. Cassell
mathematical statistician
Design Pathways
3115 NW Norwood Pl.
Corvallis OR 97330

_________________________________________________________________
Get the latest Windows Live Messenger 8.1 Beta version. Join now.
http://ideas.live.com