ttest_cohens_d: ttest_cohens_d

In selfea: Select Features Reliably with Cohen's Effect Sizes

Description

Fulfill Welch Two Sample t-test (t.test) and calculate Cohen's d as well as determine significance by p-value and effect size threshold.

Usage

ttest_cohens_d(values, groups, alpha = 0.05, power = 0.9,
  alternative = "two.sided", paired = FALSE, var.equal = FALSE)

Arguments

values	A scalar vector. Length of both of two vectors, values and groups, should be same.
groups	Experiment groups for the vector 'values'. Length of both of two vectors, values and groups, should be same. The number of groups is not limited to two, such as group <- c('A','A','A','B','B').
alpha	P-value threshold
power	Give the statistical power you desired for output significant list
alternative	Choose one of these c("two.sided", "less", "greater"). Default is "two.sided".
paired	if two groups are paired, set it to TRUE. Default is FALSE.
var.equal	if two groups are assumed to have same variance, set it to TRUE. Default is FALSE.

Value

A list containing the followings:

observed_pvalue	Calculated P-value from T-test
observed_cohens_d	Calculated Cohen's f
threshold_cohens_d	Cohen's d threshold at the desired power
threshold_pvalue	Desired p-value threshold
flag_pvalue	TRUE=passed the pvalue threshold, FALSE=not
flag_cohens_d	TRUE=passed the Cohen's d threshold, FALSE=not
power_desired	Statistical power in you input parameters
method	'Welch Two Sample t-test'
alternative	alternative option in you input parameters
paired	paired option in you input parameters
var.equal	var.equal option in you input parameters

Examples

library(selfea)

values <- c(8,10,8,8,11,29,26,22,27,26)
groups <- c("U200","U200","U200","U200","U200","U600","U600","U600","U600","U600")
list_result <- ttest_cohens_d (values, groups, 0.05, 0.90)

selfea documentation built on May 2, 2019, 5:08 a.m.