There are a number of different types of hypothesis tests, useful for different hypothesis scenarios and data samples. The most commonly used are:
- Normality: tests for normal distribution in a population sample.
- T-test: tests for a Student’s t-distribution – ie, in a normally distributed population where standard deviation in unknown and sample size is comparatively small. Paired t-tests compare two samples.
- Chi-Square Test for Independence: tests for an association of significance between two categorical variables in a population sample. Typically used with random sampling.
- Homogeneity of Variance (HOV): tests the similarity of dispersion parameters in two or more population samples.
- Analysis of Variance (ANOVA): tests for and analyzes differences between the means in several groups. Often used similarly to a t-test, but for more than two groups.
- Mood’s Median: compares the medians of two or more population samples.
- Welch’s T-test: tests for equality of means between two population samples. Also known as Welch’s unequal variances t-test.
- Kruskal-Wallis H Test: compares two or more groups with an independent variable, based on a dependent variable. Also known as one-way ANOVA on ranks.
- Box-Cox Power Transformation: transforms a data set into normal distribution.