The Kolmogorov-Smirnov test is a non-parametric approach for comparing an empirical CDF derived from observed data with that of a known distribution by computing the greatest distance between the two. A common application of the test is to verify the normality assumption, as in that case, it compares the observed distribution to the quantiles of the theoretical normal distribution. However, it is very easy to reject the null hypothesis, or conclude the distributions are different, when the sample size is large, and thus it is often not as useful as residual plots and other more qualitative diagnostics produced after a model is fit
What is Kolmogorov–Smirnov statistic?
Help us improve this post by suggesting in comments below:
– modifications to the text, and infographics
– video resources that offer clear explanations for this question
– code snippets and case studies relevant to this concept
– online blogs, and research publications that are a “must read” on this topic
Partner Ad
