Logistic regression and marinal effects

[Ana Guerrero]

Hello,

I'm Ana, and I'm currently working with a logit model that includes both continuous and categorical variables. Recently, I had a question regarding the calculation of marginal effects and the interpretation of coefficients.

After doing some research, I found out that the mfx command is no longer used, and instead, the margins command is employed. However, I'm still unsure if the command I'm using to calculate the marginal effects, considering both types of variables, is correct, as I haven't found enough information about it.

The command I'm currently using is as follows:

margins, dydx(*) atmeans

I hope you can help me, thank you very much

Ana

[Paul Obermann]

Yes, it sounds like margins is what you want as it's a simple way to convert your logit coefficients to interpretable economic effects. I think this article will give you a bit more of a start in understanding what it does and how it's used on a basic level:

https://stats.oarc.ucla.edu/stata/da...probabilities/

It's short and sweet, promise.

For anybody to give you more help than this, I think you will need to refine your questions and provide a sample of your data and what you are trying to calculate.