Sample sampling-estimator strategies across studies for meta-analysis

Source: R/meta_sample.R

Sample sampling-estimator strategies across studies for meta-analysis

get_meta_sample(
  data,
  drop_nonsampled = FALSE,
  sampling_variable = "meta",
  selection_variables = c("sample", "estimator"),
  samples_per_study = 2,
  estimator_per_sample = 3,
  force = list(sample = "pps", estimator = "ht")
)

Arguments

data

pass-through meta population

drop_nonsampled

logical indicating whether to drop units that are not sampled. Default is FALSE

sampling_variable

character string that is used as prefix for all variables generated by TLS sampling. Default is 'meta'

selection_variables

length 2 character vector of names of sample and estimator variables in population data frame (in this order). Default is 'sample' and 'estimator'

samples_per_study

number of sampling strategies per study to draw

estimator_per_sample

number of estimation strategies to draw per sampling strategy

force

named list of length two giving the sample and estimator identifiers that are forced to be sampled across studies