tab_cross: Tabulate subsets or interactions

Description

Create subsets of the survey using one variable, and tabulate another variable within each of the subsets. Interact two variables and tabulate.

Usage

tab_cross(
  vr,
  vrby,
  max_levels = getOption("surveytable.max_levels"),
  csv = getOption("surveytable.csv")
)
tab_subset(
  vr,
  vrby,
  lvls = c(),
  test = FALSE,
  alpha = 0.05,
  p_adjust = FALSE,
  drop_na = getOption("surveytable.drop_na"),
  max_levels = getOption("surveytable.max_levels"),
  csv = getOption("surveytable.csv")
)

Value

A list of tables or a single table.

Arguments

vr: variable to tabulate
vrby: use this variable to subset the survey
max_levels: a categorical variable can have at most this many levels. Used to avoid printing huge tables.
csv: name of a CSV file
lvls: (optional) only show these levels of vrby
test: perform hypothesis tests?
alpha: significance level for tests
p_adjust: adjust p-values for multiple comparisons?
drop_na: drop missing values (NA)? Categorical variables only.

Details

tab_subset creates subsets using the levels of vrby, and tabulates vr in each subset. Optionally, only use the lvls levels of vrby. vr can be categorical (factor), logical, or numeric.

tab_cross crosses or interacts vr and vrby and tabulates the new variable. Tables created using tab_subset and tab_cross have the same counts but different percentages. With tab_subset, percentages within each subset add up to 100%. With tab_cross, percentages across the entire population add up to 100%. Also see var_cross().

test = TRUE performs a test of association between the two variables. Also performs t-tests for all possible pairs of levels of vr and vrby.

Examples

Run this code

set_survey(namcs2019sv)

# For each SEX, tabulate AGER
tab_subset("AGER", "SEX")

# Same counts as tab_subset(), but different percentages.
tab_cross("AGER", "SEX")

# Numeric variables
tab_subset("NUMMED", "AGER")

# Hypothesis testing
tab_subset("NUMMED", "AGER", test = TRUE)