Is there such a thing as similarities between parametric and nonparametric statistics?

I've been doing a research on the subject, spoiler alert: I'm a noob on this. So far, I've been able to find lots of information about the differences between the two, but nothing about the similarities, except for this:

Differences & Similarities between Parametric & Non-Parametric Statistics

.. but the article never touches the subject of similarities.

Can anyone throw me some light on the subject? Thanks.

EDIT: May I politely ask why was my question downvoted? I've done my research (as best as my abilities and understanding of the subject have allowed me to), I've searched on the site, I've found similarly written questions (and getting answered without any issues), I've read the tour and help pages, so I'd love a heads up so I can keep up the quality of the content on the StackExchange sites. Thanks.

  • 3
    $\begingroup$ Do not bother too much about a single downvote. If anything, your question is quite broad and difficult to tell, when it is answered. Everyone who votes down is asked to explain in a comment. If they do not and it is just one person, do not mind too much. $\endgroup$ – Bernhard Nov 16 '17 at 7:44
  • $\begingroup$ Two possible reasons for the downvote are that the question is very broad (there are all manner of similarities one might discuss) and that the question relies on an external link that you give no context for; if the pdf were removed it would be difficult for anyone to find out what you were asking about.. Can you give the an outline of the sorts of things it says are differences in your question (or perhaps a couple of pertinent quotes), for some context? I'm sure many people don't wish to sign up to a site just to find out what you're talking about there $\endgroup$ – Glen_b Nov 17 '17 at 4:08
  • $\begingroup$ I managed to get google to give me cached text of the pdf without signing up. That's very shallow -- it almost looks like an undergraduate assignment or something; I suggest you look elsewhere for learning -- that goes wrong in its first sentence, is wrong again in the second sentence, and again in the fourth sentence, ... and doesn't get much better after that. If the author is correctly reporting his sources, some of them seem to have errors as well. If you're trying to understand what a parametric test is or what a nonparametric test is, don't look there. $\endgroup$ – Glen_b Nov 17 '17 at 4:23
  • $\begingroup$ Thanks everyone, this has been enlightening. And thanks @Glen_b, I wouldn't have known it was not a good source before this. $\endgroup$ – herrmartell Nov 23 '17 at 3:49

Their general similarity is in their approach. Most non-parametric methods are rank methods in some form. Nonparametric methods are, generally, optimal methods of dealing with a sample reduced to ranks from raw data. The logic behind the testing is the same, but the information set is different. Their similarity is in the logic of their construction.


There's often the assumption of "randomness" or independence of the samples. My bet would be that most common similarities would result from common design-based assumptions.

It might be useful to look at the examples in which some of both parametric and non-parametric tests are not applicable, e.g. When t-tests or Wilcoxon-Mann-Whitney tests won't do.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.