Adjusted Mean and Adjusted Proportions

[Claudia Orlas]

Hello everyone.

I am working with a database to calculate some models to study the association between a categorical variable called "Social support", it has three categories: strong, average and weak. AND my outcome number 1, which is a binary variable called PTSD ("Yes" or "No") and my outcome number 2, which is a continuous variables called SF-12 ranges from 0 to 100 (a composed physical functioning score) adjusting for sociodemographic variables, such as age, sex, education level, trauma mechanism, etc.

I can run my models with no problem, but I need to calculate ADJUSTED proportions and means of my outcomes.
So let's say my outcomes variables are: ptsd (categorical) AND sf12 (continuous), I need to know the adjusted proportion of PTSD among individuals who reported "Strong Social Support", "Average Social Support" and "Weak Social Support". Same for my outcome sf12, I need to know the adjusted mean of sf12 for individuals who reported each one of these categories of social support.

I don't know what the commands are to run adjusted means and proportions.

Can anyone help me out?

Thanks,

C

[William Lisowski]

You will want to make use of the margins command.

If you are not already familiar with these papers, you may find the intruduction you need, and more, in the nice overview of margins prepared by Richard Williams, a frequent contributor here, at https://www3.nd.edu/~rwilliam/xsoc73994/Margins01.pdf with a more detailed paper in the Stata Journal at http://www.stata-journal.com/article...article=st0260. I'll also note that Margins01.pdf is followed by Margins02.pdf ... Margins05.pdf covering more specialized topics.

And all five of these PDFs, and plenty more of use to someone learning about the analysis of categorical data, are linked to from https://www3.nd.edu/~rwilliam/xsoc73994/index.html - the material mentioned above is found in the section headed Interpreting results: Adjusted Predictions and Marginal effects.

[Claudia Orlas]

Thank you very much for addressing my question. I have been working based on the information that I found in those links and it was very useful. But I was wondering how to do the interpretation of these results? I thought I would get sort of an only "output" but instead I got an entire output from a logit model.

For example, let's say my outcome is a binary variable, PTSD, 0=No, 1=Yes, my variable of interest is social support network with its three categories, adjusted for all the variables I said above. At the end I got a table with this output (keeping in mind that I had to specify which category of Social Support I would take as reference), so for this example I picked "Average" social support as my reference category.

PTSD	dy/dx	Delta-method, Std. Err.	z	P> (z)	(95% CI)
Social support
Strong/Very Strong	-0.0999494	0.0576192	-1.73	0.083	-0.212881, 0.0129822
Weak/Inexistent	0.1268225	0.1034084	1.23	0.220	-0.0758542, 0.3294992

So I realized that by using this method I gotta run three times the model depending on what category from my independent variable I want to use as a reference.

My question at this point is:
1. Would it be "Adjusted comparisons" instead "Adjusted mean" or "Adjusted proportions"?
2. The codes typically work very well with categorical outcomes (logit) but when I tried to run it for my continuous outcomes (SF-12 composed score) this code did not work, is there any other way to do it? or is it a different code?

I hope my questions are understandable,

Thank you very much in advance,

Claudia