The Problem

It is often desirable to print variable labels above a summary table that shows the count of factor labels. The labels exported on choose all that apply questions, including the question and whichever response was chosen. This redundancy is often unwanted, and the results are not presented professionally.

For example, in the Nacho Craving Index data, the first ingredient is “Chips”. We see how R presents this information by simply printing the components of the ingredients___1 column.

redcap <- readRDS(file = "./redcap.rds")
redcap$ingredients___1
# What ingredients do you currently crave?: Chips
#  [1] Checked   Checked   Unchecked Unchecked Unchecked Unchecked Checked  
#  [8] Unchecked Unchecked Unchecked Unchecked Unchecked Checked   Unchecked
# [15] Unchecked Checked   Unchecked Unchecked Checked   Unchecked Unchecked
# [22] Unchecked Checked   Unchecked Checked   Unchecked Unchecked Unchecked
# [29] Unchecked Checked  
# attr(,"redcapLabels")
# [1] Unchecked Checked  
# attr(,"redcapLevels")
# [1] 0 1
# Levels: Unchecked Checked

As we can see, this information is quite ugly, so we want to tabulate the results instead. However, if we use the simple table() function to clean up this information, we lose the original question and the answer label for ingredients___1.

table(redcap$ingredients___1)
# 
# Unchecked   Checked 
#        21         9

We no longer know what the question was or which “select all” option this information represents.

The Solution

The make_choose_one_table() function can be used with a factor variable to tabulate the response while preserving the question and checked option context.

make_choose_one_table(redcap$ingredients___1) 
# What ingredients do you currently crave?: Chips
#   Response  n percent
#  Unchecked 21     70%
#    Checked  9     30%

Further, this output can be molded into a publication-ready table with a single additional function call.

make_choose_one_table(redcap$ingredients___1) %>% 
  knitr::kable()

What ingredients do you currently crave?: Chips

The subset option, if set to TRUE, will cause the function to remove the label’s text and only show the response option (i.e., not repeat the “What ingredients do you currently crave?” question).

make_choose_one_table(
  redcap$ingredients___2,
  subset = TRUE
) %>% 
  knitr::kable()

Yellow cheese

This function can also be used in an analysis pipeline with a data frame name and the name of the factor inside that data frame. For example:

redcap %>% 
  make_choose_one_table(ingredients___3) %>% 
  knitr::kable() 

What ingredients do you currently crave?: Orange cheese