Analyse multiple imputations

Performs a survey analysis on each of the designs in a svyimputationList objects and returns a list of results suitable for MIcombine. The analysis may be specified as an expression or as a function.

# S3 method for svyimputationList
with(data, expr, fun, ...,multicore=getOption("survey.multicore"))
# S3 method for svyimputationList
subset(x, subset,...)

Arguments

data,x

A svyimputationList object

expr

An expression giving a survey analysis

fun

A function taking a survey design object as its argument

...

for future expansion

multicore

Use multicore package to distribute imputed data sets over multiple processors?

subset

An logical expression specifying the subset

Value

A list of the results from applying the analysis to each design object.

See also

MIcombine, in the mitools package

Examples

library(mitools)
data.dir<-system.file("dta",package="mitools")
files.men<-list.files(data.dir,pattern="m.\\.dta$",full=TRUE)
men<-imputationList(lapply(files.men, foreign::read.dta,
  warn.missing.labels=FALSE))
files.women<-list.files(data.dir,pattern="f.\\.dta$",full=TRUE)
women<-imputationList(lapply(files.women, foreign::read.dta,
  warn.missing.labels=FALSE))
men<-update(men, sex=1)
women<-update(women,sex=0)
all<-rbind(men,women)

designs<-svydesign(id=~id, strata=~sex, data=all)
designs
#> Multiple (5) imputations: svydesign(id = ~id, strata = ~sex, data = all)

results<-with(designs, svymean(~drkfre))

MIcombine(results)
#> Multiple imputation results:
#>       with(designs, svymean(~drkfre))
#>       MIcombine.default(results)
#>                           results         se
#> drkfreNon drinker      0.41692308 0.02705950
#> drkfrenot in last wk   0.33555556 0.02082624
#> drkfre<3 days last wk  0.20632479 0.01701647
#> drkfre>=3 days last wk 0.04119658 0.00776177

summary(MIcombine(results))
#> Multiple imputation results:
#>       with(designs, svymean(~drkfre))
#>       MIcombine.default(results)
#>                           results         se     (lower     upper) missInfo
#> drkfreNon drinker      0.41692308 0.02705950 0.36387940 0.46996676      2 %
#> drkfrenot in last wk   0.33555556 0.02082624 0.29469352 0.37641759      6 %
#> drkfre<3 days last wk  0.20632479 0.01701647 0.17297052 0.23967906      2 %
#> drkfre>=3 days last wk 0.04119658 0.00776177 0.02596223 0.05643093      7 %

repdesigns<-as.svrepdesign(designs, type="boot", replicates=50)
MIcombine(with(repdesigns, svymean(~drkfre)))
#> Multiple imputation results:
#>       with(repdesigns, svymean(~drkfre))
#>       MIcombine.default(with(repdesigns, svymean(~drkfre)))
#>                           results          se
#> drkfreNon drinker      0.41692308 0.025370783
#> drkfrenot in last wk   0.33555556 0.020675251
#> drkfre<3 days last wk  0.20632479 0.015791738
#> drkfre>=3 days last wk 0.04119658 0.007900786