statistics - Obtaining the Standard Error of Weighted Data in SPSS -
i'm trying find confidence intervals means of various variables in database using spss, , i've run spot of trouble.
the data weighted, because each of people surveyed represents different portion of overall population. example, 1 young man in our sample might represent 28000 young men in general population. problem spss seems think young man's database entries each represent 28000 measurements when represent one, , makes spss think have more data do. result spss giving very low standard error estimates , very narrow confidence intervals.
i've tried fixing dividing every weight value mean weight. gives plausible figures , average weight of 1, i'm not sure resulting numbers correct.
is approach sound? if not, should try?
i've been using explore command find mean , standard error (among other things), in case matters.
you need scale weights actual sample size, procedures in complex samples option designed account sampling weights properly. regular weight variable in statistics treated frequency weight.
Comments
Post a Comment