cut score for classifying in discriminant analysis

[Debbie Hahs-Vaughn]

Can someone direct me to the computation for the cut score for classifying
cases in discriminant analysis?  Thanks in advance!

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[Oleg Solovyev]

>Can someone direct me to the computation for the cut score for classifying
>cases in discriminant analysis?  Thanks in advance!

I am not sure. I just guess. Can you use one-level decision tree for that?

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD

[SR Millis-3]

I would recommend using logistic regression rather than discriminant function analysis.  Logistic regression has several advantages:

--less restrictive assumptions

--straight-forward interpretation of the individual covariates/predictor variables

--easily derived cutoff scores, ie, when you exponeniate the linear composite composite, you get the probability of group membership:

Pr = e^composite / 1 + e^composite

By default, cutoff for classifying someone is Group 1 is pr> .50, but that cutoff can be adjusted up or down by examining the sensitivity and specificity values associated with each cutoff.

In summary, I cannot think of a single reason why anyone would use DFA in this case.

Scott R Millis, PhD, MEd, ABPP (CN,CL,RP), CStat
Professor & Director of Research
Dept of Physical Medicine & Rehabilitation
Wayne State University School of Medicine
261 Mack Blvd
Detroit, MI 48201
Email:  [hidden email]
Tel: [phone]
Fax: [phone]


--- On Sun, 1/25/09, Oleg Solovyev <[hidden email]> wrote:

> From: Oleg Solovyev <[hidden email]>
> Subject: Re: cut score for classifying in discriminant analysis
> To: [hidden email]
> Date: Sunday, January 25, 2009, 2:56 PM
> >Can someone direct me to the computation for the cut
> score for classifying
> >cases in discriminant analysis?  Thanks in advance!
>
> I am not sure. I just guess. Can you use one-level decision
> tree for that?
>
> =====================
> To manage your subscription to SPSSX-L, send a message to
> [hidden email] (not to SPSSX-L), with no body
> text except the
> command. To leave the list, send the command
> SIGNOFF SPSSX-L
> For a list of commands to manage subscriptions, send the
> command
> INFO REFCARD

=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD