2

I'm trying to analyse a multiple response question from a weighted survey dataset. I like the srvyr package because it allows me to use the dplyr pipes, but I can't find the reference material on how to handle multiple response questions.

I have a simple dataset looking at different sources of income. Here's an example of how the data looks like

ID <- c(1,2,3,4,5,6,7,8,9,10)
rent_income <- c("Yes", "Yes", "No", "Yes", "No", "Yes", "No", "Yes", "No", "No")
salary_income <- c( "No", "Yes", "No", "Yes", "No", "Yes", "Yes", "No", "Yes", "No")
other_income <- c( "No", "Yes", "No", "No", "No", "No", "Yes", "No", "No", "No")
survey_weights <- c(0.6, 1.2 , 1.1 , 0.7 , 2.4 , 1.1 , 0.3 , 0.6 , 1.0 , 0.8)
df<-data.frame(ID, rent_income, salary_income, other_income, survey_weights)

Note that the data is entirely made up. With srvyr if first have to create a survey object

weighted_dataset <- df %>% as_survey_design(ids=ID, weights=survey_weights)

Now I would want to calculate the weighted percentage of the sample that has different types of incomes. Any ideas on how to do that? In Stata there is a function called mr_tab . But I can't find a similar one in R

Gianzen
  • 73
  • 4

2 Answers2

1

You could use the convenient group_by() and variable selection syntax available through the dplyr and srvyr R packages.

weighted_dataset %>%
  # Organize the data into groups defined by each combination of the income variables
    group_by_at(vars(ends_with("_income"))) %>%
  # For categorical variables, this calculates estimates of percentages
    summarize(Percent = survey_mean())

> # A tibble: 6 x 5
>  rent_income salary_income other_income Percent Percent_se
>  <fct>       <fct>         <fct>          <dbl>      <dbl>
> 1 No          No            No             1          0    
> 2 No          Yes           No             0.769      0.265
> 3 No          Yes           Yes            0.231      0.265
> 4 Yes         No            No             1          0    
> 5 Yes         Yes           No             0.6        0.312
> 6 Yes         Yes           Yes            0.40       0.312
bschneidr
  • 6,014
  • 1
  • 37
  • 52
0

see the proportions by group block of https://cran.r-project.org/web/packages/srvyr/vignettes/srvyr-vs-survey.html

Anthony Damico
  • 5,779
  • 7
  • 46
  • 77