What type of variables does the Kruskal-Wallis test handle?

The Kruskal-Wallis test is designed to analyze differences among three or more independent groups when the assumptions necessary for parametric tests (like normal distribution) are not met. It is a non-parametric statistical test, meaning it does not rely on the distribution of the data. This makes it particularly useful for handling data that may not follow a normal distribution or when sample sizes are small.

The test operates on ranked data, allowing researchers to compare the median values of different groups instead of using means, which is characteristic of parametric tests. This aspect of the Kruskal-Wallis test means it is effective for ordinal data or continuous data that do not assume normal distributions.

While the test can process continuous variables that are not normally distributed, it categorically falls under non-parametric statistics, which specifically addresses situations with undefined or skewed distributions. Therefore, the correct characterization of the types of variables the Kruskal-Wallis test handles is that it is used for non-parametric variables.