test_equality_paired: test_equality_paired

In vinisousa04/AdvancedDataAnalysis: AdvancedDataAnalysis

Description

Tests if the mean difference of the same group in different situations, e.g. before and after a training, is statiscally significant In order to do so, first it is tested if the differen between the situations, e.g. again the time before the training and the time after the training is normally distibuted. If it is then a paired t test takes place, otherwise it is computed the Wilcoxon Signed-Rank Test.

Usage

test_equality_paired(df, time1, time2, confi = 0.95,
  alternative = "two.sided")

Arguments

time1	The quoted column name of one of the situations.
time2	The quoted column name of the other situation.
confi	The confidence level.
alternative	The alternative hypothesis for the t test and wilcox test. The default value is "two.sided", others options are "greater" and "less".

Details

@param df A data frame containing the values of the different situations in different columns, example in one column the time before the training and in other column the time after the training.

Value

A list containg the shapiro test of normality and the appropriate test of mean/median equality. Note: If the appropriate test if the t test, then the result of levene test for equal variance is also returned.

Examples

test_equality_paired(df = data, time1=score_before, time2=score_after,confi = 0.95, alternative = "two.sided")

vinisousa04/AdvancedDataAnalysis documentation built on May 14, 2019, 6:02 p.m.