Hi colleages,
I am correcting a points in my manuscript in revision process.
i have a doubt about the n that i need to report in my table.
The data is the National health survey from Chile, but one of the problems with this data is strata is not specificed in the data and it is impossible to obtain.
For this reason, i am using this sintaxis in STATA to consider survey design in the results:
svyset _n [pweight=FEXP1], vce(linearized) singleunit(missing)
svy linearized : tabulate stress_financ Depression_1, count column
the results i obtaind with this syntaxsis are expanded, but i was thinking is better to report weighted but not expanded (how i can do it in STATA?) .
Or should i simply report raw "n"? I think this last option could be the best because readers could estimated power analysis or consider the real validity of result. For example, in this data two men are houseworkers and they had depressive episode.
Thanks for advice, whatever recommendations or reference about this issue.
Hi sir
If u cant get the the exact data you can made exploratory resarch works from different strata in your survey in order to approximate in your results
Try to find exact data for exploratory research. Otherwise, share sample data and project detail here for understanding . Best of luck for your research.
- Badges
- Science topic
- Similar topics
- Linguistics
- Philology
- Classical Philology
- Manuscripts