ttest without grouped variable

[Raphael Moritz]

Dear community,

I'm a stata newbie, so I really hope my post make sense to you and this kind of beginner question is okay. For a data project (n>100) I want to analyze the statistical significance of differences between two groups. But, I do not have a grouped variable. The data looks like this (as an example):

Share Men	Share Women
0,3	0,2
0,9	0,1
2,5	8,9
21,4	10,1

Thus, I'm missing a dummy that gives me the grouped variable. How can I then test for differences?

Thank you very much!

Kind regards,
Raphael

[Rich Goldstein]

this cannot be what your data actually look like as Stata does not allow spaces in variable names; please read the FAQ and post as requested there

if you look at

help ttest

you will see that there is a syntax that is basically testing equality (paired or unpaired) between two variable names - at a guess, you can use that syntax

[Raphael Moritz]

Thanks. Of course, this was an example, these are not the actual variable names. So, I guess there is no possibility to test unpaired data with this kind of format?

[Richard Williams]

Like Rich G. says, it would be much easier to advise you if you provided a real example with your own data. See pt 12 in the FAQ, especially the part about dataex.

I will be quite surprised if Stata can't handle this. Like Rich, My guess is you want something like

ttest mvar=wvar

or maybe

ttest mvar=wvar, unpaired.

For a brief discussion of the various types of ttest commands, see

https://www3.nd.edu/~rwilliam/stats1...mple-Stata.pdf