Wilcoxon Test

[amh]

1. What it is?

Non-parametric statistic for comparing two paired groups of independent samples. It calculates the difference between a set of pairs and evaluates their differences. The outcome determines if both independent samples are significantly different from one another or not.

2. Type of Wilcoxon test

- Wilcoxon rank sum test
- Wilcoxon signed rank test

Reference:

https://www.statstest.com/wilcoxon-signed-rank-test/

[amh]

1. Wilcoxon signed-rank test

- To determine whether two measurements taken from a single group are significantly different from each other (interest variable).
- The variable of interest should be continuous and your group randomly sampled to meet assumption.

2. Wilcoxon signed-sign rank test assumptions

- Continous
- Skew distribution
- Random sample

3. Assumptions

Continuous

- Means the variable can take any reasonable value
- Examples; age, weight , height , test score

Skewed distribution

- The interest variable does not need to be bell shaped
- Skew distribution can be used with Wilcoxon Signed Rank Test

Random Samples

- Data points selected must be from random sample (randomly selected)
- If not selected randomly, the analysis will be incorrect - due to bias/tendency for inccorect result due to bad data