Tabulate a check-all-that-apply (multi-select) question as percentages

Multi-select questions arrive as a block of columns sharing a prefix, each holding the chosen option (or blank). This computes, per option, the share of respondents who selected it – so the percentages can (and usually do) sum to more than 100. The denominator is the number of distinct respondents who selected at least one option.

Usage

calc_percentage_multi(
  data = NULL,
  prefix,
  id = NULL,
  by = NULL,
  sort = c("none", "desc", "asc"),
  digits = 0,
  drop = NULL,
  clean_names = TRUE,
  split = "auto"
)

Arguments

data: A data frame.
prefix: Common column-name prefix identifying the option block, e.g. "motivations_".
id: Optional respondent identifier column (unquoted) used as the distinct-respondent denominator. If omitted, each row is treated as one respondent.
by: Optional grouping column(s); percentages are computed within group.
sort, digits: As in calc_percentage(). Sorting acts on the option labels.
drop: Optional character vector of option labels to remove before counting (matched against the cleaned labels, case-insensitive). Defaults to the drop_answers option. See drop_items().
clean_names: If TRUE (default), tidy the option labels by stripping prefix and un-mangling exporter artefacts ("A...B" -> "A / B").
split: How to handle a question that arrives packed into a single cell per respondent ("Speed; Drivers"). "auto" (default) detects ";", "|" or "," and splits on it; pass a string to force a delimiter, or FALSE never to split. Only ever applies when prefix selects exactly one column.

Value

A tibble with option, n (respondents choosing it) and pct.

Details

The crucial difference from calc_percentage() is the denominator: it is the number of distinct respondents who picked any option in the block, not the number of ticks. So if 700 of 1,000 respondents selected at least one motivation, each option's percentage is "out of 700", and the percentages can (and usually do) sum to more than 100. Supply id to count distinct respondents exactly; without it, each row counts as one respondent. With clean_names = TRUE the option labels are stripped of prefix and exporter artefacts like "A...B" are turned back into "A / B".

Two export shapes are handled. The usual one is a block of columns, one per option, named with a shared prefix; that is what prefix selects. The other is a single column holding every answer a respondent ticked, joined by a delimiter, which is what Google Forms and most spreadsheet exports produce. If prefix matches exactly one column and a delimiter is found in it, that column is unpacked automatically and the answer text itself becomes the option label (clean_names is not applied, since the answers are real wording rather than mangled column names). Behaviour on a column block is unchanged. Use split_multi() when you want those unpacked columns kept for other work.

Examples

calc_percentage_multi(podracing_survey, "motivations_",
                      id = respondent_id, sort = "desc")
#> # A tibble: 5 × 3
#>   option        n   pct
#>   <fct>     <int> <dbl>
#> 1 speed       772    79
#> 2 drivers     533    54
#> 3 social      437    45
#> 4 tradition   349    36
#> 5 betting     295    30

# a spreadsheet export that packs every answer into one cell
packed <- data.frame(
  respondent = 1:4,
  motivations = c("Speed; Drivers", "Speed", "", "Betting; Speed")
)
calc_percentage_multi(packed, "motivations", id = respondent, sort = "desc")
#> # A tibble: 3 × 3
#>   option      n   pct
#>   <fct>   <int> <dbl>
#> 1 Speed       3   100
#> 2 Betting     1    33
#> 3 Drivers     1    33