Summarizes one continuous outcome across the levels of several
categorical variables, one block of rows per variable. It is the
inverse layout of table_continuous(), which puts several outcomes
in rows and one grouping in columns.
Each block reports the outcome's statistics level by level, plus its
own group comparison on the block's header row, and an Overall row
gives the marginal summary of the whole analytic sample.
Usage
table_outcome(
data,
outcome,
select,
labels = NULL,
overall = TRUE,
drop_na = FALSE,
weights = NULL,
rescale = FALSE,
test = c("welch", "student", "nonparametric"),
p_value = NULL,
statistic = FALSE,
show_n = TRUE,
show_columns = NULL,
effect_size = c("none", "auto", "hedges_g", "eta_sq", "r_rb", "epsilon_sq"),
effect_size_ci = FALSE,
ci = TRUE,
ci_level = 0.95,
digits = 2,
effect_size_digits = 2,
p_digits = 3,
decimal_mark = ".",
align = c("decimal", "center", "right"),
output = c("default", "data.frame", "long", "tinytable", "gt", "flextable", "excel",
"clipboard", "word"),
indent_text = " ",
indent_text_excel_clipboard = strrep(" ", 6),
excel_path = NULL,
excel_sheet = NULL,
clipboard_delim = "\t",
word_path = NULL,
user_na = TRUE,
style = NULL,
by
)Arguments
- data
A data frame.
- outcome
The continuous outcome, unquoted or as a string. Exactly one column.
- select
The grouping characteristics to describe the outcome across, as a tidyselect expression or a character vector of column names. One block of rows per variable, in the order given. As everywhere in the family,
selectis what structures the rows.- labels
Named character vector of display labels, for the outcome and for the
selectvariables alike.- overall
Show the marginal
Overallrow (defaultTRUE).- drop_na
Drop rows with a missing
selectvalue from that block (defaultFALSE: they are shown as a(Missing)level and excluded from the comparison).- weights, rescale
Frequency weights and whether to rescale them to sum to the sample size, as in
table_continuous().- test
Group comparison for every block:
"welch"(default),"student"or"nonparametric".- p_value
Show the p-value column (default
TRUE).- statistic
Show the test statistic column.
- show_n
Show the count column.
- show_columns
Character vector of statistic tokens;
NULLkeeps the historical display.- effect_size, effect_size_ci
Effect size per block and its confidence interval, as in
table_continuous().- ci, ci_level
The mean's confidence interval and its level.
- digits, effect_size_digits, p_digits, decimal_mark
Number formatting.
- align
Numeric-cell alignment:
"decimal","center"or"right".- output
One of
"default","data.frame","long", or a rendering engine:"tinytable","gt","flextable","excel","clipboard","word"."data.frame"and"long"are synonyms and return the same object: the compute frame is already long (one row per level, plus the block header and the overall row). Both names are kept so the argument reads the same here as intable_continuous(), where the pair is also synonymous.- indent_text, indent_text_excel_clipboard
Level-row indentation, for the console and for the plain-text engines.
- excel_path, excel_sheet, clipboard_delim, word_path
Output destinations, as in
table_continuous().- user_na
Honor declared missing values (see
?freq).- style
A journal style; see
spicy_style().- by
Defunct. The grouping characteristics are selected with
selectsince spicy 0.13.0; supplyingbyis an error.
Value
A spicy_outcome_table: the compute frame, with the display
frame and the typed view attached. output = "data.frame" /
"long" returns the compute frame unclassed – the two tokens are
synonyms and return identical objects.
Which shape do I need?
Several continuous variables across one grouping is
table_continuous() (select = , by = ). One continuous variable
across one or several groupings is this function. A single select is
legitimate here – it is the natural way in when you know more
groupings are coming – but with several outcomes and one grouping,
the sibling is the table you want.
Choosing the statistics
show_columns takes the same tokens as table_continuous(), with
the same meanings; see the show_columns section of
?table_continuous for the vocabulary. Only the character-vector
form is accepted here: there is one outcome, so a per-variable list
would name nothing.
Weights
weights applies the frequency-expansion convention of the family:
all weights 1 reproduces the unweighted table, and integer weights
reproduce the STATISTICS of the data duplicated that many times –
n stays the raw count of rows that carried the weights. Rows
with a missing or zero weight leave the analytic sample; the note
counts the missing ones.
rescale is the switch between the two readings of a weight: the
frequency reading above, and the sampling-weight reading, where the
weights are normalized to sum to the sample size. See the Weights
section of table_continuous() for the choice in full.
rescale = TRUE normalizes the weights over the outcome's whole
surviving sample, once, never per level – a per-level rescale
would destroy the relative weights across levels, which is the
entire information a sampling weight carries into this table. The
means are unchanged by it; the standard deviations move, because
their denominator is sum(w) - 1.
A weighted table refuses the group comparison. The estimates and
their interval have no weighted version here, and a p-value or an
effect size silently computed unweighted beside weighted
descriptives is the one thing that must not happen: set p_value = FALSE (and statistic = FALSE, effect_size = "none"), or use
table_continuous_lm() for a weighted comparison. The
order-statistic median interval is refused for the same reason.
Blocks and the group comparison
Every block is a separate one-way comparison of the outcome across the levels of that variable. Nothing in this table adjusts one block for another, and the table note says so. Read the blocks as a set of bivariate descriptions, not as a model.
Each block chooses its test independently: with two observed levels
test = "welch" is the Welch t-test, with three or more it is the
Welch one-way ANOVA, and test = "nonparametric" is the
Wilcoxon rank-sum or the Kruskal-Wallis test on the same rule. A
block with fewer than two observed levels, or with a level holding a
single observation, is not tested; its statistics stay empty and the
other blocks are unaffected.
The Overall row
overall = TRUE puts the marginal summary of the whole analytic
sample on the first row. Under the default drop_na = FALSE the
levels of every block partition that sample – the (Missing)
display level included – so each block's counts add up to the
Overall count exactly, which is what makes it a usable
denominator.
The row reads Overall, not Total, and the distinction is
deliberate. Total is the word of a COUNT margin: the column of
table_categorical() where frequencies add up. This row is the
whole analytic sample, where a mean is recomputed over every
observation and nothing is added. A mean is not a total.
Choosing the select columns
The canonical form is select = where(is.factor), or an explicit
enumeration. Negation (select = -c(x, y)) is not recommended: it
sweeps in every remaining column, and a numeric one becomes a block
with one LEVEL per distinct value, in order of first appearance. A
variable producing more than 20 levels raises a warning for that
reason – an arbitrary threshold, but a sixty-row block where a
reader expects a handful of categories is not a table.
A haven_labelled column used as select shows its numeric CODES, not
its value labels, as it does in table_continuous(). Convert it
first (haven::as_factor()) to get the labels in the stub.
The table note
One note sits under the table and states what left the analytic
sample, which group comparison ran in each block, what the displayed
columns mean, and how the blocks and the Overall row are to be
read. The rendering engines carry the same sentence as a table note.
On the "tinytable" route it is set one size down;
options(spicy.note_style) governs that (see table_regression()).
See also
table_continuous() for the transposed shape,
table_categorical() for categorical outcomes.
Examples
table_outcome(sochealth, bmi, select = c(sex, smoking))
#> Descriptive statistics of Body mass index
#>
#> Variable │ M SD Min Max 95% CI LL 95% CI UL n p
#> ────────────────┼─────────────────────────────────────────────────────────────
#> Overall │ 25.93 3.72 16.00 38.90 25.72 26.14 1188
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Sex │ .018
#> Female │ 25.69 3.78 16.00 38.90 25.39 25.98 616
#> Male │ 26.20 3.64 16.00 37.70 25.90 26.50 572
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Current smoker │ .903
#> No │ 25.96 3.76 16.00 38.90 25.72 26.21 915
#> Yes │ 25.93 3.58 16.80 35.30 25.48 26.38 248
#> (Missing) │ 24.74 3.63 17.60 32.50 23.24 26.23 25
#>
#> Missing values removed: bmi (12). Group comparison: Welch t-test. Each block compares Body mass index across the levels of one variable; blocks are not adjusted for one another. Overall = the whole analytic sample.
table_outcome(sochealth, wellbeing_score, select = where(is.factor))
#> Descriptive statistics of WHO-5 wellbeing index (0-100)
#>
#> Variable │ M SD Min Max 95% CI LL
#> ─────────────────────────────────┼────────────────────────────────────────
#> Overall │ 69.04 15.62 18.70 100.00 68.16
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Sex │
#> Female │ 67.16 14.80 19.60 100.00 65.99
#> Male │ 71.05 16.23 18.70 100.00 69.73
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Age group │
#> 25-34 │ 67.55 15.36 24.60 100.00 65.60
#> 35-49 │ 69.48 15.36 19.60 100.00 67.94
#> 50-64 │ 69.06 15.94 18.70 100.00 67.33
#> 65-75 │ 69.78 15.84 26.60 100.00 67.80
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Highest education level │
#> Lower secondary │ 57.22 15.44 18.70 97.90 55.33
#> Upper secondary │ 68.97 13.62 26.70 100.00 67.82
#> Tertiary │ 76.85 13.23 40.40 100.00 75.55
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Subjective social class │
#> Lower │ 64.26 14.61 21.10 100.00 61.85
#> Working │ 65.24 15.39 19.60 100.00 63.43
#> Lower middle │ 69.17 16.58 23.70 100.00 67.37
#> Middle │ 72.82 14.13 32.10 100.00 71.23
#> Upper middle │ 72.78 14.95 18.70 100.00 70.33
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Region of residence │
#> Central │ 68.59 15.64 23.70 100.00 66.62
#> East │ 69.69 16.00 30.40 100.00 67.07
#> North │ 69.70 16.04 21.10 100.00 67.48
#> Other │ 70.50 15.07 19.60 100.00 68.39
#> South │ 67.57 15.70 18.70 100.00 65.38
#> West │ 68.51 15.36 29.00 100.00 66.41
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Employment status │
#> Employed │ 69.42 15.07 19.60 100.00 68.35
#> Student │ 70.42 16.16 18.70 100.00 67.75
#> Unemployed │ 65.78 17.17 23.20 100.00 63.21
#> Inactive │ 69.72 15.64 21.10 100.00 66.90
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Household income group │
#> Low │ 67.66 15.78 21.10 100.00 65.68
#> Lower middle │ 68.78 15.21 23.70 100.00 67.26
#> Upper middle │ 69.18 16.16 19.60 100.00 67.42
#> High │ 70.95 15.51 18.70 100.00 68.89
#> (Missing) │ 67.94 12.41 41.00 93.20 61.77
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Current smoker │
#> No │ 69.36 15.62 18.70 100.00 68.36
#> Yes │ 67.65 15.66 21.10 100.00 65.69
#> (Missing) │ 70.99 15.05 42.00 100.00 64.77
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Regular physical activity │
#> No │ 67.71 15.92 18.70 100.00 66.48
#> Yes │ 70.61 15.13 21.10 100.00 69.35
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Dentist visit in last 12 months │
#> No │ 66.15 14.79 19.60 100.00 64.61
#> Yes │ 70.25 15.81 18.70 100.00 69.18
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Self-rated health │
#> Poor │ 43.60 11.15 19.60 70.70 40.74
#> Fair │ 58.54 13.49 18.70 94.30 56.92
#> Good │ 71.58 12.58 26.60 100.00 70.53
#> Very good │ 78.67 12.54 44.20 100.00 77.24
#> (Missing) │ 73.42 11.32 56.30 100.00 68.13
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> BMI category │
#> Normal weight │ 70.47 15.36 18.70 100.00 69.07
#> Overweight │ 68.78 15.65 19.60 100.00 67.49
#> Obesity │ 65.43 15.84 29.00 100.00 62.91
#> (Missing) │ 72.61 14.84 50.70 97.60 63.18
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Trust in institutions │
#> Very low │ 69.84 15.35 23.20 100.00 68.26
#> Low │ 69.35 15.40 24.60 100.00 67.95
#> High │ 67.92 16.17 19.60 100.00 65.83
#> Very high │ 67.69 16.15 18.70 100.00 64.87
#>
#> Variable │ 95% CI UL n p
#> ─────────────────────────────────┼────────────────────────
#> Overall │ 69.93 1200
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Sex │ <.001
#> Female │ 68.33 620
#> Male │ 72.37 580
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Age group │ .372
#> 25-34 │ 69.49 242
#> 35-49 │ 71.03 383
#> 50-64 │ 70.80 327
#> 65-75 │ 71.76 248
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Highest education level │ <.001
#> Lower secondary │ 59.10 261
#> Upper secondary │ 70.12 539
#> Tertiary │ 78.15 400
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Subjective social class │ <.001
#> Lower │ 66.66 144
#> Working │ 67.06 278
#> Lower middle │ 70.98 326
#> Middle │ 74.41 306
#> Upper middle │ 75.22 146
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Region of residence │ .461
#> Central │ 70.56 245
#> East │ 72.32 145
#> North │ 71.92 203
#> Other │ 72.62 198
#> South │ 69.75 201
#> West │ 70.61 208
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Employment status │ .049
#> Employed │ 70.49 762
#> Student │ 73.09 143
#> Unemployed │ 68.35 174
#> Inactive │ 72.54 121
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Household income group │ .149
#> Low │ 69.64 247
#> Lower middle │ 70.29 388
#> Upper middle │ 70.93 328
#> High │ 73.02 219
#> (Missing) │ 74.11 18
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Current smoker │ .125
#> No │ 70.37 926
#> Yes │ 69.60 249
#> (Missing) │ 77.20 25
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Regular physical activity │ .001
#> No │ 68.94 650
#> Yes │ 71.88 550
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Dentist visit in last 12 months │ <.001
#> No │ 67.70 354
#> Yes │ 71.32 846
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Self-rated health │ <.001
#> Poor │ 46.45 61
#> Fair │ 60.17 266
#> Good │ 72.62 558
#> Very good │ 80.11 295
#> (Missing) │ 78.72 20
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> BMI category │ .002
#> Normal weight │ 71.87 465
#> Overweight │ 70.06 569
#> Obesity │ 67.95 154
#> (Missing) │ 82.04 12
#> ╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌┼╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌╌
#> Trust in institutions │ .366
#> Very low │ 71.42 366
#> Low │ 70.74 472
#> High │ 70.00 234
#> Very high │ 70.51 128
#>
#> Group comparison: Welch one-way ANOVA (age_group, education, social_class, region, employment_status, income_group, self_rated_health, bmi_category, institutional_trust); Welch t-test (sex, smoking, physical_activity, dentist_12m). Each block compares WHO-5 wellbeing index (0-100) across the levels of one variable; blocks are not adjusted for one another. Overall = the whole analytic sample.