Sorry for yet another beginner question...

#1
Hello all,

I am conducting a retrospective cohort analysis and will need to compare the demographic data between the cohorts to check if they are the same or if any categories significantly differ. For age I plan to use the mean of each group and utilize a t-test (or z-test).

In order to assess whether the groups (call them HEMS and GEMS) have significantly different amounts of males vs females what's the best way to go? Is there a simple way to compare the two groups, with a null hypothesis that they are the same? Can I use a 2x2 table with HEMS and GEMS on the top and Male/Female on the side? I'm not sure how that gets me my expected value (i.e. one would expect M/F to be 0.5). Or can I simply compare HEMS with an 'expected' of the GEMS numbers to see if they differ enough to reject that they are the same?

Sorry if this is a bit obtuse, I'm stretching in statistics...
 

hlsmith

Less is more. Stay pure. Stay poor.
#2
Given the sample size, formal test may be an issue. And even if there is an issue, you are just looking at bivariate comparison and not controlling for the other covariates.

I would just recommend making a "Love Plot".