Table 3.

The table of assumptions and tests for their validation

Assumptions	Definitions	Tests
Independence	The selection of a participant must be random and independent of the selection of the other participants in the group.	Wald-Wolfowitz run test
Normality	The response variable is normally distributed.	Graphical: p-p# plot, q-q* plot, histogram, and boxplot
		Test: Shapiro-Wilk test (N ≤50) and D'Agostino skewness test (N >50)‡
Homoscedasticity	The variances are similar in all the groups.	Graphical: Boxplot
		Test: Levene test and Brown-Forsythe test
Group ratio	The number of participants in any of the group should not exceed the 1:4 ratio.	Check the sample size in each group