3
$\begingroup$

I have data from two different ethnic groups (not same sample size...) ranking different traits. Each group rank ordered the 10 different traits, and I want to test to see if there is a difference between the way they ranked the data. Is there a test I can use to test whether the rankings are different overall?

$\endgroup$

1 Answer 1

3
$\begingroup$

I think you may be looking for what has sometimes been called ANACONDA. See this paper by Li and Schucany entitled "Some properties of a test for concordance of two groups of rankings" and then a subsequent one by Hollander and Sethuraman entitled "Testing for agreement between two groups of judges"

Suppose that one group has $m$ rankings with ranks $R_{ij}$ $(i = 1, ...,m; j = 1, ..., k)$ and that the other group has $n$ rankings with ranks $R^{'}_{ij}$ $(i = 1, ..., n; j = 1, .. ., k)$. Let $$S_j = \sum_{i=1}^{\infty} R_{ij},\quad T_j = \sum_{i=1}^{\infty} R^{'}_{ij}$$ the statistic $\mathcal{L}$ is defined to be $\sum_{i=1}^k S_j T_j$.

Li and Schucany then discuss how to derive its distribution.

$\endgroup$
1
  • $\begingroup$ Awesome, that's exactly what I was looking for. Thanks! $\endgroup$
    – Mrspud
    Dec 15, 2016 at 19:43

Your Answer

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

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