ez.anova: Anova

In NicolasStefaniak/easieR: easieR: A GUI R metapackage.

Anova

Description

GUI for analysis of variance and covariance, with contrasts

Usage

ez.anova<-function(data=NULL, DV=NULL, between=NULL, within=NULL,id=NULL, cov=NULL, RML=NULL, 
                   RML.factor=NULL, param=c("param","bayes"),outlier=c("complete","id", "removed"), 
                   ES="ges", SumS="3", save=F , html=T, contrasts="none",p.adjust="none", n.boot=1000, rscaleFixed = 0.5,
                   rscaleRandom = 1 )

Arguments

data	name of dataframe in which to find variables
DV	character or vector. Dependant variable - the measured variable name
between	character or vector. Between group variable. Must be a factor variable
within	character or vector. Within group variable. Must be a factor variable
id	character or vector. Variable used to identify subjects. Must be a factor variable
cov	character or vector. Variable used as covariable. Must be a numeric/integer variable
RML	character with length >= 2. Repeated measure levels. All the columns that corresponds to repeated measures in the wide format
RML.factor	list. The names in the list corresponds to the names of the factors and the values at each level correspond to the levels of each factor. The product of the number of levels must equal the length of RML. If RML is not NULL and RML.factor is NULL, it is assumed that there is only one factor and the name of the factor is "variable.1"
param	character. One or several among "param", "bayes", "non param", and "robust"
outlier	character. One or several among "complete", "id", or "removed"
ges	one among "ges" or "pes"
SumS	Type of sum of squares, one among "2" or "3"
save	logical. Do you want to save the results
html	logical. Do you want easieR to output the results in nice html document ?
contrast	list. The names in the list corresponds to the names of the factors and the values is a matrix of coefficients for the contrasts. "pairs" or "none" are also possible
p.adjust	adjust p values for multiples comparisons. see <code>p.adjust</code>
n.boot	number of bootstrap for bayes and robust analyses. Default is 1000
rscaleFixed	rscale for the fixed effect in Bayes analyses. Default = 0.5. It is not possible to state it in dialog box
rscaleRandom	rscale for the fixed effect in Bayes analyses. Default = 1. It is not possible to state it in dialog box

Details

ez.anova relies on the packages 'afex', 'emmeans', 'WRS2', 'WRS', 'BayesFactor' and 'PMCMRplus' for performing parmatric and non parametric analyses, as well as Bayes and robust analyses

Value

Model	The model that has been tested
Descriptive	Descriptive statistics
normality	Shapiro test, Lilliefors test, graphical distribution of residuals and QQplot
Homogenity of variance	Levene test testing the homogenity of variance
Plot	The ggplot that summarizes the analysis
Mauchly test	Mauchly test that allows to assess the sphericity of the covariance matrix
principal analysis	The main anova table
Contrast	Give the matrix of contrasts for each variable and the contrasts computed by emmeans
Bayes factor	Bayes factor for the analysis
Non paramtric test	Non parametric test. The test of Kruskall-Wallis or anova of Friedman depending on the nature of the variable
Robust analyses	Anova on trimmed means or on medians
Grubbs test	Grubbs' test allowing to determine whether there are outlier in the analysis

Author(s)

Nicolas Stefaniak

Examples

ez.anova() # open GUI

NicolasStefaniak/easieR documentation built on Jan. 31, 2025, 2:59 p.m.