Which statistical method you use to compare data sets depends on two main factors: your overall goal and the type of data you have. Parametric data means that you know the underlying distribution (for example, your data might be normally distributed). Non parametric tests are an option if you don’t know the data’s distribution.
Each method has their quirks and pitfalls: see the links below for more information about each test.
Further Reading
References
Statistical Methods for Comparison of Data Sets of Construction Met…
