Explains how to select a statistical test suitable for your hypothesis. Suggests points to consider before deciding about a test. Gives a list of commonly used parametric and non-parametric tests with their purposes of use.
1. How to Choose
a Right Statistical Test
Prof. Dr. Khalid Mahmood
University of the Punjab
2. Points to
Consider
Purpose of the study
Testing against a value
Relationship between variables
Difference between groups
Independent (predictor) vs. dependent
(outcome) variables
Level of measurement of the data
Nominal, Ordinal, Interval, Ratio
Number of samples
One, Two, More
3. Parametric vs.
Non-parametric
Tests
Parametric tests have more
statistical power
They require to meet
assumptions
Random selection of sample
Normality of data
Equal variance of data from
multiple groups
Interval or ratio level of outcome
variable
5. LIST OF COMMONLY USED TESTS
Parametric test Use Non-parametric test
One Sample t Test Compares 1 Mean to a specified value Wilcoxon Signed Rank
Paired Samples t Test Compares 2 Paired Means Wilcoxon Rank Sum
Independent Samples t
Test
Compares 2 Independent Means Mann-Whitney U
One-Way ANOVA Compares 3 or more Means Kruskal-Wallis H
Pearson’s r Correlates 2 variables Spearman’s rho
Simple Linear Regression Predicts value of 1 variable given value of
another variable
Not applicable
Not applicable Associates 2 categorical variables Chi-Square Test of
Independence