thr3ads.net - R help - [R] is there anything wrong with doing a multinomial logistic like it is a binary logistic regression? [Aug 2013]

If this information is useful, please help other people find it:
Share via:

Eunice Chou

2013-Aug-17 03:13 UTC

[R] is there anything wrong with doing a multinomial logistic like it is a binary logistic regression?

My outcome variable (y) is 3 categories. Is there anything bad about using
the following code to get a parameter estimate for my bivariate model?

publicfit = glm(y ~ public, data=dataSPSS.vmj, family=binomial)

	[[alternative HTML version deleted]]

Michael Dewey

2013-Aug-17 12:49 UTC

head link

[R] is there anything wrong with doing a multinomial logistic like it is a binary logistic regression?

At 04:13 17/08/2013, Eunice Chou wrote:>My outcome variable (y) is 3 categories. Is there anything bad about using
>the following code to get a parameter estimate for my bivariate model?
>
>publicfit = glm(y ~ public, data=dataSPSS.vmj, family=binomial)
Have you tried it? What did R tell you?

>         [[alternative HTML version deleted]]
Michael Dewey
info at aghmed.fsnet.co.uk
http://www.aghmed.fsnet.co.uk/home.html

R help - Aug 2013 - is there anything wrong with doing a multinomial logistic like it is a binary logistic regression?

[R] is there anything wrong with doing a multinomial logistic like it is a binary logistic regression?

[R] is there anything wrong with doing a multinomial logistic like it is a binary logistic regression?