Design-based t-test

One-sample or two-sample t-test. This function is a wrapper for svymean in the one-sample case and for svyglm in the two-sample case. Degrees of freedom are degf(design)-1 for the one-sample test and degf(design)-2 for the two-sample case.

svyttest(formula, design,  ...)

Arguments

formula

Formula, outcome~group for two-sample, outcome~0 or outcome~1 for one-sample. The group variable must be a factor or character with two levels, or be coded 0/1 or 1/2

design

survey design object

...

for methods

Value

Object of class htest

See also

t.test

Examples

data(api)
dclus2<-svydesign(id=~dnum+snum, fpc=~fpc1+fpc2, data=apiclus2)
tt<-svyttest(enroll~comp.imp, dclus2)
tt
#> 
#> 	Design-based t-test
#> 
#> data:  enroll ~ comp.imp
#> t = -2.8882, df = 36, p-value = 0.006518
#> alternative hypothesis: true difference in mean is not equal to 0
#> 95 percent confidence interval:
#>  -384.24757  -67.22654
#> sample estimates:
#> difference in mean 
#>          -225.7371 
#> 
confint(tt, level=0.9)
#> [1] -357.68999  -93.78413
#> attr(,"conf.level")
#> [1] 0.9

svyttest(enroll~I(stype=="E"),dclus2)
#> 
#> 	Design-based t-test
#> 
#> data:  enroll ~ I(stype == "E")
#> t = -9.8291, df = 36, p-value = 9.826e-12
#> alternative hypothesis: true difference in mean is not equal to 0
#> 95 percent confidence interval:
#>  -706.306 -464.688
#> sample estimates:
#> difference in mean 
#>           -585.497 
#> 

svyttest(I(api00-api99)~0, dclus2)
#> 
#> 	Design-based one-sample t-test
#> 
#> data:  I(api00 - api99) ~ 0
#> t = 9.0704, df = 38, p-value = 4.782e-11
#> alternative hypothesis: true mean is not equal to 0
#> 95 percent confidence interval:
#>  20.02455 31.53117
#> sample estimates:
#>     mean 
#> 25.77786 
#>