Methods For Twins Analysis

[Leon Edelman]

Greetings All,

I have posted about Twins Studies and after much searching have not quite found a solution. Say I have a sample of twins raised together and a sample of twins raised apart. One outcome of interest is smoking and one is income. How do I seek out the effects of environment and genetics in this type of analysis? As I mentioned previously I have examined intraclass correlations but I am afraid I have not found ways to use these data to the full capacity largely because unfortunately no one in my program is familiar with these techniques and I haven't quite found sufficient resources on the web to address it.

So if you had a sample of twins raised together and apart and you wanted to examine the role of environment and genetics on smoking (as one income) and income (as a separate outcome) what would you do?

[Phil Bromiley]

You didn't get a quick answer. You'll increase your chances of a helpful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output, and sample data using dataex. Also, do not include attached files - many of us will not open them.

There is a substantial literature on twins and much of it deals with estimation issues. Some is associated with the Minnesota Twins study. My off-the-cuff answer is I'd be inclined to do separate analyses with fixed effects for each twin, but you really need to look at the literature.