Non-Parametric Tests

Non Parametric Tests

Hi All, 

     It is interesting how perceptions can lead us astray when applications of statistical principles are encountered. Usually, we perceive data that is entirely quantitative in nature to be subject to statistical analysis. However, there is an entire area of statistics that deals with data which though is quantitative, emerges from qualitative considerations. Usually such data is at the most ordinal in nature. 

     The area of statistics that deals with such data is Non - Parametric statistics. Interestingly I was asked to throw some light on Non-Parametric Tests, especially Friedman ANOVA test. Well, Non-Parametric tests are to start with what the name suggests - non based on Parameters. Now what do we exactly mean by Parameters here? Why do the Non-Parametric tests not consider parameters as their name suggests, what is their theoretical basis? What are the uses? How can they be applied. For these questions and more, just follow the link below for a small paper that I had written for this blog itself a few days back. It covers Parametric Tests in general and discusses Friedman test - a kind of Parametric test which is popular as a test of different treatments or judgments expressed as ranking on a given set of subjects. The null hypothesis in such cases is that all treatments have identical effects or all judgments have similar rankings, or that the samples differ in some way. The alternate hypothesis is that the treatments do have different effects or the judges do rank the subjects in different ways. So just  follow the link below and enjoy the paper. 

⏬

Link to Paper 

     Once, you are through the paper, just come back with questions and comments, if any. We will cover more about other Non-Parametric Tests and Friedman test as well as other different types of Non-Parametric Tests in the next post.

Till then,

Happy Stat-ing 😊