Parametric, non-parametric, robust, and Bayesian one-way ANOVA.
Usage
oneway_anova(
data,
x,
y,
subject.id = NULL,
type = "parametric",
paired = FALSE,
digits = 2L,
conf.level = 0.95,
effsize.type = "omega",
var.equal = FALSE,
bf.prior = 0.707,
tr = 0.2,
nboot = 100L,
...
)Arguments
- data
A data frame (or a tibble) from which variables specified are to be taken. Other data types (e.g., matrix, table, array, etc.) will not be accepted. Additionally, grouped data frames from
{dplyr}should be ungrouped before they are entered asdata.- x
The grouping (or independent) variable from
data. For repeated measures designs, see the note on ordering insubject.id.- y
The response (or outcome or dependent) variable from
data.- subject.id
Relevant in case of a repeated measures or within-subjects design (i.e.,
paired = TRUE), it specifies the subject or repeated measures identifier. Important: If this argument isNULL(which is the default), observations are paired by their row order within each level ofx(i.e., the data is assumed to be sorted in a subject-1, subject-2, ... pattern within every level). If the data is not sorted this way, the paired results will be silently incorrect, so it is safest to always specifysubject.id. For paired designs, each subject must have at most one row per level ofx(an error is thrown otherwise), and rows with a missingsubject.idare dropped because they cannot be paired (seecomplete_subjects()).- type
A character specifying the type of statistical approach:
"parametric""nonparametric""robust""bayes"
You can specify just the initial letter (e.g.
"np"or"bf"). Matching is on the initial lowercase letter only, so any other value (including capitalized values such as"Bayes") falls back to"parametric"without a warning.- paired
Logical that decides whether the experimental design is repeated measures/within-subjects or between-subjects. The default is
FALSE.- digits
Number of digits for rounding or significant figures. May also be
"signif"to return significant figures or"scientific"to return scientific notation. Control the number of digits by adding the value as suffix, e.g.digits = "scientific4"to have scientific notation with 4 decimal places, ordigits = "signif5"for 5 significant figures (see alsosignif()).- conf.level
Scalar between
0and1(default:95%confidence/credible intervals,0.95).- effsize.type
Type of effect size needed for parametric tests. The argument can be
"eta"(partial eta-squared) or"omega"(partial omega-squared).- var.equal
a logical variable indicating whether to treat the two variances as being equal. If
TRUEthen the pooled variance is used to estimate the variance otherwise the Welch (or Satterthwaite) approximation to the degrees of freedom is used.- bf.prior
A number between
0.5and2(default0.707), the prior width to use in calculating Bayes factors and posterior estimates. In addition to numeric arguments, several named values are also recognized:"medium","wide", and"ultrawide", corresponding to r scale values of1/2,sqrt(2)/2, and1, respectively. In case of an ANOVA, this value corresponds to scale for fixed effects.- tr
Trim level for the mean when carrying out
robusttests. In case of an error, try reducing the value oftr, which is by default set to0.2. Lowering the value might help.- nboot
Number of bootstrap samples for computing confidence interval for the effect size (Default:
100L).- ...
Additional arguments (currently ignored).
Value
The returned object is a tibble data frame with the additional class
"statsExpressions". The exact set of columns depends on the test and, for
functions that accept a type argument, on the chosen analysis (parametric,
non-parametric, robust, or Bayesian). Any given call therefore returns some
(not all) of the columns below.
Hypothesis testing
statistic: the numeric value of a statisticdfand/ordf.error: degrees of freedom for the test. A t-test usesdf.error, a chi-squared test usesdf, and a parametric ANOVA uses both (numeratordfand denominatordf.error)p.value: the p-value associated with the observed statistic (two-sided unless a one-sidedalternativeis requested, where supported)method: the name of the inferential statistical test
Effect size estimation
effectsize: the name of the effect sizeestimate: estimated value of the effect sizeconf.level: the coverage level of the confidence/credible interval (e.g.0.95); the interval itself spansconf.lowtoconf.highconf.low: lower bound for the effect size estimateconf.high: upper bound for the effect size estimateconf.method: method used to compute the confidence/credible intervalconf.distribution: statistical distribution for the effect
Bayesian analysis (only when type = "bayes")
bf10: Bayes factor for the alternative hypothesis relative to the nulllog_e_bf10: natural logarithm of the Bayes factor (present for most, but not all, Bayesian analyses)prior.distribution,prior.scale,prior.location: prior specification used to compute the Bayes factor and posterior estimates
Pairwise comparisons (for pairwise_comparisons() and
pairwise_contingency_table())
group1,group2: the two levels being comparedp.adjust.method: the adjustment method used for multiple comparisonsp.value.adj: the adjusted p-value; returned bypairwise_contingency_table(). Note thatpairwise_comparisons()instead folds the adjusted value intop.value(and does not return a separatep.value.adjcolumn)
Common columns
n.obs: sample size used for the analysis: observations for unpaired tests, complete subjects/pairs for paired two-sample tests and ANOVAs, or studies for meta-analysisexpression: a list-column of pre-formatted plotmath expressions; each element is alanguageobject (not a character string) containing the statistical details, ready to be used in{ggplot2}(e.g. inlabs()orannotate())
For a per-function, column-by-column breakdown of the output (and an explanation
of the internal add_expression_col() engine that builds the expression
column), see the Return value schema
article. For more examples, see the Data frame outputs article.
One-way ANOVA
Below is a summary of:
the statistical test carried out for inferential statistics
the type of effect size estimate and a measure of uncertainty for this estimate
the functions used internally to compute these details
Between-subjects
Hypothesis testing
| Type | No. of groups | Test | Function used |
| Parametric | > 2 | Fisher's or Welch's one-way ANOVA | stats::oneway.test() |
| Non-parametric | > 2 | Kruskal-Wallis one-way ANOVA | stats::kruskal.test() |
| Robust | > 2 | Heteroscedastic one-way ANOVA for trimmed means | WRS2::t1way() |
| Bayesian | > 2 | Fisher's ANOVA | BayesFactor::anovaBF() |
Effect size estimation
| Type | No. of groups | Effect size | CI available? | Function used |
| Parametric | > 2 | partial eta-squared, partial omega-squared | Yes | effectsize::omega_squared(), effectsize::eta_squared() |
| Non-parametric | > 2 | rank epsilon squared | Yes | effectsize::rank_epsilon_squared() |
| Robust | > 2 | Explanatory measure of effect size | Yes | WRS2::t1way() |
| Bayesian | > 2 | Bayesian R-squared | Yes | performance::r2_bayes() |
Within-subjects
Paired data: See complete_subjects() for subject matching, duplicate
observations, and removal of incomplete subjects.
Hypothesis testing
| Type | No. of groups | Test | Function used |
| Parametric | > 2 | One-way repeated measures ANOVA | afex::aov_ez() |
| Non-parametric | > 2 | Friedman rank sum test | stats::friedman.test() |
| Robust | > 2 | Heteroscedastic one-way repeated measures ANOVA for trimmed means | WRS2::rmanova() |
| Bayesian | > 2 | One-way repeated measures ANOVA | BayesFactor::anovaBF() |
Effect size estimation
| Type | No. of groups | Effect size | CI available? | Function used |
| Parametric | > 2 | partial eta-squared, partial omega-squared | Yes | effectsize::omega_squared(), effectsize::eta_squared() |
| Non-parametric | > 2 | Kendall's coefficient of concordance | Yes | effectsize::kendalls_w() |
| Robust | > 2 | Algina-Keselman-Penfield robust standardized difference average | Yes | WRS2::wmcpAKP() |
| Bayesian | > 2 | Bayesian R-squared | Yes | performance::r2_bayes() |
Citation
Patil, I., (2021). statsExpressions: R Package for Tidy Dataframes and Expressions with Statistical Details. Journal of Open Source Software, 6(61), 3236, https://doi.org/10.21105/joss.03236
Examples
# for reproducibility
set.seed(123)
# ----------------------- parametric -------------------------------------
# between-subjects
oneway_anova(mtcars, cyl, wt)
#> # A tibble: 1 × 14
#> statistic df df.error p.value
#> <dbl> <dbl> <dbl> <dbl>
#> 1 20.2 2 19.0 0.0000196
#> method effectsize estimate
#> <chr> <chr> <dbl>
#> 1 One-way analysis of means (not assuming equal variances) Omega2 0.637
#> conf.level conf.low conf.high conf.method conf.distribution n.obs expression
#> <dbl> <dbl> <dbl> <chr> <chr> <int> <list>
#> 1 0.95 0.370 1 ncp F 32 <language>
# biased (partial eta-squared) effect size
oneway_anova(mtcars, cyl, wt, effsize.type = "eta")
#> # A tibble: 1 × 14
#> statistic df df.error p.value
#> <dbl> <dbl> <dbl> <dbl>
#> 1 20.2 2 19.0 0.0000196
#> method effectsize estimate
#> <chr> <chr> <dbl>
#> 1 One-way analysis of means (not assuming equal variances) Eta2 0.681
#> conf.level conf.low conf.high conf.method conf.distribution n.obs expression
#> <dbl> <dbl> <dbl> <chr> <chr> <int> <list>
#> 1 0.95 0.437 1 ncp F 32 <language>
# within-subjects design
oneway_anova(bugs_long, condition, desire, subject.id = subject, paired = TRUE)
#> Registered S3 method overwritten by 'car':
#> method from
#> na.action.merMod lme4
#> # A tibble: 1 × 18
#> term sumsq sum.squares.error df df.error meansq statistic p.value
#> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 condition 233. 984. 2.63 229. 4.30 20.6 8.27e-11
#> method effectsize estimate
#> <chr> <chr> <dbl>
#> 1 ANOVA estimation for factorial designs using 'afex' Omega2 (partial) 0.0783
#> conf.level conf.low conf.high conf.method conf.distribution n.obs expression
#> <dbl> <dbl> <dbl> <chr> <chr> <int> <list>
#> 1 0.95 0.0280 1 ncp F 88 <language>
# ----------------------- non-parametric ----------------------------------
# between-subjects
oneway_anova(mtcars, cyl, wt, type = "np")
#> # A tibble: 1 × 15
#> parameter1 parameter2 statistic df.error p.value
#> <chr> <chr> <dbl> <int> <dbl>
#> 1 wt cyl 22.8 2 0.0000112
#> method effectsize estimate conf.level conf.low
#> <chr> <chr> <dbl> <dbl> <dbl>
#> 1 Kruskal-Wallis rank sum test Epsilon2 (rank) 0.736 0.95 0.624
#> conf.high conf.method conf.iterations n.obs expression
#> <dbl> <chr> <int> <int> <list>
#> 1 1 percentile bootstrap 100 32 <language>
# within-subjects design
oneway_anova(
bugs_long,
condition,
desire,
subject.id = subject,
paired = TRUE,
type = "np"
)
#> # A tibble: 1 × 15
#> parameter1 parameter2 statistic df.error p.value method
#> <chr> <chr> <dbl> <dbl> <dbl> <chr>
#> 1 desire condition 55.8 3 4.56e-12 Friedman rank sum test
#> effectsize estimate conf.level conf.low conf.high conf.method
#> <chr> <dbl> <dbl> <dbl> <dbl> <chr>
#> 1 Kendall's W 0.211 0.95 0.154 1 percentile bootstrap
#> conf.iterations n.obs expression
#> <int> <int> <list>
#> 1 100 88 <language>
# ----------------------- robust -------------------------------------
# between-subjects
oneway_anova(mtcars, cyl, wt, type = "r")
#> # A tibble: 1 × 12
#> statistic df df.error p.value
#> <dbl> <dbl> <dbl> <dbl>
#> 1 12.7 2 12.2 0.00102
#> method
#> <chr>
#> 1 A heteroscedastic one-way ANOVA for trimmed means
#> effectsize estimate conf.level conf.low conf.high
#> <chr> <dbl> <dbl> <dbl> <dbl>
#> 1 Explanatory measure of effect size 1.02 0.95 0.824 1.62
#> n.obs expression
#> <int> <list>
#> 1 32 <language>
# within-subjects design
oneway_anova(
bugs_long,
condition,
desire,
subject.id = subject,
paired = TRUE,
type = "r"
)
#> # A tibble: 1 × 12
#> statistic df df.error p.value
#> <dbl> <dbl> <dbl> <dbl>
#> 1 21.0 2.73 145. 1.15e-10
#> method
#> <chr>
#> 1 A heteroscedastic one-way repeated measures ANOVA for trimmed means
#> effectsize estimate
#> <chr> <dbl>
#> 1 Algina-Keselman-Penfield robust standardized difference average 0.325
#> conf.level conf.low conf.high n.obs expression
#> <dbl> <dbl> <dbl> <int> <list>
#> 1 0.95 0.110 0.572 88 <language>
# ----------------------- Bayesian -------------------------------------
# between-subjects
oneway_anova(mtcars, cyl, wt, type = "bayes")
#> # A tibble: 6 × 17
#> term pd prior.distribution prior.location prior.scale bf10
#> <chr> <dbl> <chr> <dbl> <dbl> <dbl>
#> 1 mu 1 cauchy 0 0.707 20968.
#> 2 cyl-4 1 cauchy 0 0.707 20968.
#> 3 cyl-6 0.542 cauchy 0 0.707 20968.
#> 4 cyl-8 1 cauchy 0 0.707 20968.
#> 5 sig2 1 cauchy 0 0.707 20968.
#> 6 g_cyl 1 cauchy 0 0.707 20968.
#> method log_e_bf10 effectsize estimate std.dev
#> <chr> <dbl> <chr> <dbl> <dbl>
#> 1 Bayes factors for linear models 9.95 Bayesian R-squared 0.581 0.0815
#> 2 Bayes factors for linear models 9.95 Bayesian R-squared 0.581 0.0815
#> 3 Bayes factors for linear models 9.95 Bayesian R-squared 0.581 0.0815
#> 4 Bayes factors for linear models 9.95 Bayesian R-squared 0.581 0.0815
#> 5 Bayes factors for linear models 9.95 Bayesian R-squared 0.581 0.0815
#> 6 Bayes factors for linear models 9.95 Bayesian R-squared 0.581 0.0815
#> conf.level conf.low conf.high conf.method n.obs expression
#> <dbl> <dbl> <dbl> <chr> <int> <list>
#> 1 0.95 0.373 0.702 HDI 32 <language>
#> 2 0.95 0.373 0.702 HDI 32 <language>
#> 3 0.95 0.373 0.702 HDI 32 <language>
#> 4 0.95 0.373 0.702 HDI 32 <language>
#> 5 0.95 0.373 0.702 HDI 32 <language>
#> 6 0.95 0.373 0.702 HDI 32 <language>
# within-subjects design
oneway_anova(
bugs_long,
condition,
desire,
subject.id = subject,
paired = TRUE,
type = "bayes"
)
#> Multiple `BFBayesFactor` models detected - posteriors are extracted from
#> the first numerator model.
#> See help("get_parameters", package = "insight").
#> # A tibble: 8 × 19
#> term pd prior.distribution prior.location prior.scale effect
#> <chr> <dbl> <chr> <dbl> <dbl> <chr>
#> 1 mu 1 cauchy 0 0.707 fixed
#> 2 condition-HDHF 1 cauchy 0 0.707 fixed
#> 3 condition-HDLF 0.868 cauchy 0 0.707 fixed
#> 4 condition-LDHF 0.994 cauchy 0 0.707 fixed
#> 5 condition-LDLF 1 cauchy 0 0.707 fixed
#> 6 sig2 1 cauchy 0 1 fixed
#> 7 g_condition 1 cauchy 0 1 fixed
#> 8 g_.rowid 1 cauchy 0 1 fixed
#> bf10 method log_e_bf10 effectsize
#> <dbl> <chr> <dbl> <chr>
#> 1 1360982789. Bayes factors for linear models 21.0 Bayesian R-squared
#> 2 1360982789. Bayes factors for linear models 21.0 Bayesian R-squared
#> 3 1360982789. Bayes factors for linear models 21.0 Bayesian R-squared
#> 4 1360982789. Bayes factors for linear models 21.0 Bayesian R-squared
#> 5 1360982789. Bayes factors for linear models 21.0 Bayesian R-squared
#> 6 1360982789. Bayes factors for linear models 21.0 Bayesian R-squared
#> 7 1360982789. Bayes factors for linear models 21.0 Bayesian R-squared
#> 8 1360982789. Bayes factors for linear models 21.0 Bayesian R-squared
#> estimate std.dev conf.level conf.low conf.high conf.method component n.obs
#> <dbl> <dbl> <dbl> <dbl> <dbl> <chr> <chr> <int>
#> 1 0.531 0.0323 0.95 0.461 0.587 HDI conditional 88
#> 2 0.531 0.0323 0.95 0.461 0.587 HDI conditional 88
#> 3 0.531 0.0323 0.95 0.461 0.587 HDI conditional 88
#> 4 0.531 0.0323 0.95 0.461 0.587 HDI conditional 88
#> 5 0.531 0.0323 0.95 0.461 0.587 HDI conditional 88
#> 6 0.531 0.0323 0.95 0.461 0.587 HDI conditional 88
#> 7 0.531 0.0323 0.95 0.461 0.587 HDI conditional 88
#> 8 0.531 0.0323 0.95 0.461 0.587 HDI conditional 88
#> expression
#> <list>
#> 1 <language>
#> 2 <language>
#> 3 <language>
#> 4 <language>
#> 5 <language>
#> 6 <language>
#> 7 <language>
#> 8 <language>
