Displaying 1 result from an estimated 1 matches for "ospi25632".
2008 Aug 15
2
Design-consistent variance estimate
...end data to anyone interested
### It is the same code as above, just applied to a data
> gg <- svydesign(ids = ~CSRSSchoolCode, data = qq)
Warning message:
In svydesign.default(ids = ~CSRSSchoolCode, data = qq) :
No weights or probabilities supplied, assuming equal probability
> svymean(~OSPI25632, gg, deff='replace', na.rm=TRUE)
mean SE DEff
OSPI25632 0.84250 0.01072 1.0384
>
> ### My code which replicates the formulae in the LaTeX code below
> ## Cluster totals
> xx <- with(qq, tapply(OSPI25632, CSRSSchoolCode, sum, na.rm=TRUE))
> ## Mean of...