Is it possible to know if a set of rankings are internally correlated? For example, if you ask to rank 5 cars as their favorite from


these cars are not completely independent of each other, chances are if you like offroad vehicles you won't be a fan of sports cars, vice versa. this will be reflected in the rankings. many real-world rankings are like this, however, most measures of ranked lists ignore any inter-item correlations.

so, how can we tell if the rankings of an individual (measured over multiple repeated surveys) or a group of individuals show correlation?

  • $\begingroup$ You start by asking about "correlation" and then refer to "cluster size," but these concepts appear to have little to do with each other. That casts doubt on what you are trying to ask. Could you clarify what you are looking for and what you mean by "determine this"? To what does "this" refer? $\endgroup$
    – whuber
    Mar 19, 2020 at 11:43
  • $\begingroup$ Thanks, I updated the question, please let me know if it is clearer. $\endgroup$
    – ha554an
    Mar 19, 2020 at 14:46
  • $\begingroup$ hi, I have edited the answer, but it is still closed? no feedback? $\endgroup$
    – ha554an
    Mar 20, 2020 at 16:44
  • $\begingroup$ Your post entered the review queue to bring it to the attention of community members. I saw it there but I still couldn't figure out what you are trying to ask, so I did not vote to reopen it. It looks like everybody else has a similar reaction. $\endgroup$
    – whuber
    Mar 20, 2020 at 16:52
  • 1
    $\begingroup$ I suggest editing it again. $\endgroup$
    – whuber
    Mar 20, 2020 at 16:57


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.