I am trying to find correlation/association between two categorical vectors. I tried using Chi2, but it does not have a predefined upper limit or lower limit. This is when I found Tschuprow's T and Cramer's V. Now, I have many questions in this regard, and I would appreciate all help
I am looking for some metric that is as usable and informative as Pearson correlation or Spearman correlation and having 0-1 limits for the score.
1- To some extent yes they are suitable.
2- Yes, they provide the value & Approximate Significance (P-value).
3- Yes. The method and formula for calculating them provide skewed measurements and have no symmetry like Pearson & Spearman.
4- Honestly, I did not see.
5- I think using a Nominal Logistic Regression with Minitab software is useful and better.
6- The same number 5.
good luck. very much
Why did you not use the chi-square analysis.please give more details.. Best wishes, David Booth
Aha. I remembered another point now.
You can also use Pairwise Z-Tests in SPSS software.
Cramer's V will range from 0 to 1. You can play with some toy data to see how it reacts in different cases. The interpretation for a "large" effect changes depending on how many categories there are (in the dimension with the smaller number of categories). These interpretations are addressed in Cohen, 1988, Statistical Power Analysis for the Behavioral Sciences, 2nd Edition. There is a bias correction available for Cramer's V. Some other effect size statistics to consider are Tschuprow's T and Goodman Kruskal lambda. One limitation to Tschuprow's T is that it can reach a value of 1 only in the case of square tables ( at least according to this: https://en.wikipedia.org/wiki/Tschuprow%27s_T )
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Statistical Software
- SPSS