The 2 groups are parents and their children. The data is not normal. Would like the comparison without the effects of age.
Hello Kamlam,
The answer depends on several things, one of which is whether you intend to treat the groups as matched (e.g., parent of family 1 = case 1 in parent batch; child of family 1 = case 1 in child batch) or as independent groups. The second depends on whether your sample is really wide enough in terms of parent and child age to permit a traditional covariate treatment of the age variable (or if, instead, there's a huge gap in which no observed cases appear, as would likely be the case if you chose only families with pre-school aged children).
You could use a bootstrap/resampling approach in an ancova or regression framework in order to evaluate scores (on the DV) for differences due to group, and this would offer a viable comparison without having to worry about normality of model residuals. This option is available in a number of statistical programs.
Good luck with your work.
Go for correlation between children and mothers, there you will find what mothers are doing with children. Then you will come to know what to do.....
- Badges
- Science topic
- Similar topics
- Medicine
- Epi
- Age Groups
- Children