ebp_test_means {povmap}

Perform test for difference between survey and census means

Description

This function computes weighted means of the same set of variables within the census and the survey. A test for difference of the means are performed for each variable with two-tailed p-values returned.

Usage

ebp_test_means(varlist, smp_data, pop_data, weights = NULL, pop_weights = NULL)

Arguments

varlist

character vector, the set of variables of interest

smp_data

the survey data

pop_data

the population data

weights

a character string containing the name of a variable that indicates weights in the sample data. If a character string is provided a weighted version of the ebp will be used. The variable has to be numeric. Defaults to NULL.

pop_weights

a character string containing the name of a variable that indicates population weights in the populatation data. If a character string is provided weighted indicators are estimated using population weights. The variable has to be numeric. Defaults to NULL. Please note that pop_weights should only be used if the samples and population data are at different levels (e.g.: smp_data at individual level and pop_data at household level, then pop_weights is needed for the comparison with a variable indicating household size).

Value

dataframe with census and survey means and test results for their difference.

Examples

data("eusilcA_pop")
data("eusilcA_smp")

variables <- c("gender", "eqsize", "cash", "self_empl",
               "unempl_ben", "age_ben", "surv_ben",
               "sick_ben", "dis_ben", "rent", "fam_allow",
               "house_allow", "cap_inv", "tax_adj")

ebp_test_means(varlist = variables,
               pop_data = eusilcA_pop,
               smp_data = eusilcA_smp,
               weights = "weight")