What is the primary characteristic of non-parametric statistical tests?

The primary characteristic of non-parametric statistical tests is that they can be used without assuming a specific distribution. This flexibility is particularly important when dealing with data that does not meet the assumptions of parametric tests, such as normality. Non-parametric tests are beneficial for analyzing ordinal data or non-normally distributed interval data because they rely on ranks or other order-based criteria rather than specific probability distributions.

In contrast, options that suggest requiring normally distributed data directly contradict the foundational principles of non-parametric tests. Furthermore, while non-parametric tests can yield p-values, the statement about providing exact p-values does not universally apply, as approximate methods might be used. Finally, while some non-parametric tests can be applied to small samples, they are not limited to larger sizes, making that option inaccurate as well. By not imposing restrictive assumptions about data distribution, non-parametric tests offer a versatile approach for various statistical analyses.