how to run a logit regression in excel

[PhantomOak]

Is there a way to run a logit regression in excel?

 

[shg]

There's an example at http://www.youtube.com/watch?v=NHOO7iceJrw

 

[PhantomOak]

There's an example at http://www.youtube.com/watch?v=NHOO7iceJrw

Yep... thats the best I could find searching google... it was one of the two questions that forced me to stop lurking. I was HOPING there was a better way.

Am I to assume that there is no way to run it without doing that?

 

[shg]

What's the nature of your data, and what are you trying to estimate?

 

[PhantomOak]

What's the nature of your data, and what are you trying to estimate?

The data is from a local hospital. Date admitted, Age, Sex (M/F), Blood Alcohol, and substance (non-alcohol) (Y/N).

This is for my girlfriends MSW project. She wanted to see what in the available data (that is all the feds want, so its all the hospitals track) was a significant predictor of substance being present. It is widely believed that all three would be.


I have done a lot of work with logit regressions in the past with STATA, and other packages, but I dont have access to them. She only has access to SPSS at school. I was HOPING to help her find a way to get some significance from her data (right now there is no significance, including "no significance").


I am thinking that it might be my best shot to do it at work, or with her at school...

 

[shg]

You want to know how age, gender, and blood alcohol level affect the likelihood of a positive blood test for other drugs? I assume date is disregarded?

Am I to assume that there is no way to run it without doing that?

I don't know of a better way, but think I understand that method. Do you discount that approach out of hand?