analysis by region, how to?

[marcelo ribeiro]

Dear all,

I have a cross-section database. My dependent variable is the share of Female in the governing boards by region (Africa, Americas, Asia Pacific, Europe, MENA).
I regress it with other variables and I found the effects. However, I want to check how the results change according to the region.
How can I do it? Which command should I use? I tried to use cluster but it does not seems right.

Thank you

[Clyde Schechter]

I doubt anybody can answer your question without seeing an example of your data and the code you used for your regression. Please read the Forum FAQ, with particular attention to #12, for guidance on how best to show those things.

[Carlo Lazzaro]

Marcelo:
I do share Clyde's take.
My only aside would be to add -i.region- as a further predictor in the right-hand side of your regression equation (see -fvvarlist- for more details).
For the future, please note that positive replies to any query are heavily conditional on informative details provided by the original poster.

[marcelo ribeiro]

 Country  Region  %WomenBoard %WomenStaff Pres(M:0/W:1) CFO (0/1) HDI
              Andorra Europe .2 .25 0 1 .78
              Ecuador Americas .3 .31 1 1 .65
              Jordan Mena .1 .14 1 0 .6  
end

[marcelo ribeiro]

glm %WomenBoard %WomenStaff i.Pres##i.CFP HDI , link(logit) family(binomial) vce(robust) nolog

[marcelo ribeiro]

sorry for not sharing the dataset, I'm new in statalist. I tried to share an example of my variables and the commands I'm using. I used glm because I'm regressing proportions. I just wanted to see the differences of the results by region and I was wondering which command should I add to my model.

[Carlo Lazzaro]

Marcelo:
read (and act on) the FAQ about how to share your code and Stata output via CODE delimiters (is easier done than said).
As far as your code is concerned, you may want to try something alondg the following lines (see added predictor in red):

glm %WomenBoard %WomenStaff i.Pres##i.CFP HDI i.region, link(logit) family(binomial) vce(robust) nolog

[marcelo ribeiro]

Thank you Carlo, I could use the command you suggest me. It was exactly what I was looking for. Thanks, I appreciate it.
Have a good day.

Best regards,
Marcelo

[Marcos Almeida]

After the regression analysis, you may think about "checking how the results change by region" by using - margins - followed by - marginsplot - commands.