EViews.com

Hi all!

Currently I'm struggling with my model. It's a logit model which computes if you're going to work or not.
It's for me the first time I use such a (logit) model.

I looked up several things and I found that I can just enter my dependent variables into Eviews (not making logarithms of it, right?).
After running my regression however, I don't know how I should interpret those results.

I searched and found the numbers don't mean anything directly, but i should make log odds or marginal effects of them. The last one should be the best.
For continious variables I can do the following:

forecast (index) -> workingf , xb=@mean(workingf), l_xb=@dlogistic(-xb) and multiply each coefficient with l_xb, which gives the marginal effect. Is that correct?

However, I have several dummies in my model as well. Because those are not contrinious, I dont now how I can compute the marginal effects.

Can someone help me with this? Many thanks in advance!

Cheers,

Mitch.

p.s. Is my way of computing the marginal effects for continous variables correct?

Hi,

I have a similar query. I was trying to compute the marginal effects in a logit model and wanted to calculate the odds ratio, by keeping all other variables at their mean. I was wondering if there was a command that this automatically, and if not how could I go about computing this? Any help would be much appreciated.

Many thanks,
Ash