Create a tbl_svy survey object using two phase design

Create a survey object by specifying the survey's two phase design. It is a wrapper around twophase. All survey variables must be included in the data.frame itself. Variables are selected by using bare column names, or convenience functions described in select.

as_survey_twophase(.data, ...)

# S3 method for data.frame
as_survey_twophase(
  .data,
  id,
  strata = NULL,
  probs = NULL,
  weights = NULL,
  fpc = NULL,
  subset,
  method = c("full", "approx", "simple"),
  ...
)

# S3 method for twophase2
as_survey_twophase(.data, ...)

Arguments

.data

A data frame (which contains the variables specified below)

...

ignored

id

list of two sets of variable names for sampling unit identifiers

strata

list of two sets of variable names (or NULLs) for stratum identifiers

probs

list of two sets of variable names (or NULLs) for sampling probabilities

weights

Only for method = "approx", list of two sets of variable names (or NULLs) for sampling weights

fpc

list of two sets of variables (or NULLs for finite population corrections

subset

bare name of a variable which specifies which observations are selected in phase 2

method

"full" requires (much) more memory, but gives unbiased variance estimates for general multistage designs at both phases. "simple" or "approx" use less memory, and is correct for designs with simple random sampling at phase one and stratified randoms sampling at phase two. See twophase for more details.

Value

An object of class tbl_svy

Examples

# Examples from ?survey::twophase
# two-phase simple random sampling.
data(pbc, package="survival")
library(dplyr)

pbc <- pbc %>%
  mutate(randomized = !is.na(trt) & trt > 0,
         id = row_number())
d2pbc <- pbc %>%
  as_survey_twophase(id = list(id, id), subset = randomized)

d2pbc %>% summarize(mean = survey_mean(bili))
#> # A tibble: 1 × 2
#>    mean mean_se
#>   <dbl>   <dbl>
#> 1  3.26   0.256

# two-stage sampling as two-phase
library(survey)
data(mu284)

mu284_1 <- mu284 %>%
  dplyr::slice(c(1:15, rep(1:5, n2[1:5] - 3))) %>%
  mutate(id = row_number(),
         sub = rep(c(TRUE, FALSE), c(15, 34-15)))

dmu284 <- mu284 %>%
  as_survey_design(ids = c(id1, id2), fpc = c(n1, n2))
# first phase cluster sample, second phase stratified within cluster
d2mu284 <- mu284_1 %>%
  as_survey_twophase(id = list(id1, id), strata = list(NULL, id1),
                  fpc = list(n1, NULL), subset = sub)
dmu284 %>%
  summarize(total = survey_total(y1),
            mean = survey_mean(y1))
#> # A tibble: 1 × 4
#>   total total_se  mean mean_se
#>   <dbl>    <dbl> <dbl>   <dbl>
#> 1 15080    2274.  44.4    2.27
d2mu284 %>%
  summarize(total = survey_total(y1),
            mean = survey_mean(y1))
#> # A tibble: 1 × 4
#>   total total_se  mean mean_se
#>   <dbl>    <dbl> <dbl>   <dbl>
#> 1 15080    2274.  44.4    2.27

# dplyr 0.7 introduced new style of NSE called quosures
# See `vignette("programming", package = "dplyr")` for details
ids <- quo(list(id, id))
d2pbc <- pbc %>%
  as_survey_twophase(id = !!ids, subset = "randomized")