1

I have been trying to calculate the association in R between two categorical variables—level of education and political party voted. I have 8 choices for education and 8 choices for political parties.

How can I determine if there is an association between these two variables and the choices within them? I have been recommended to use Cramer's V, Lambda, and Uncertainty measures. Which method would be appropriate for my question?

Todd Burus
  • 632
  • 2
  • 12
MilanWoland
  • 11
  • 2
  • 1
    Do you have truly categorical variables or ordinal variables? Ordinal variables would be like high school/associate’s/bachelor’s/master’s/doctorate and far left/left/center/right/far right. True categorical variables would be like choice of major for education and political party (democrat, republican, libertarian, green), where there isn’t a natural way to order the categories. – Dave Dec 19 '20 at 22:19
  • I am just a beginner in both statistics and R, so if I misuse a term, sorry in advance. I think that my variables are ordinal according to your explanation. To make it clear, in education variable, I have choices like primary school, high school, university degree etc, and for political parties, I literally have the names of the parties in a selected country. – MilanWoland Dec 19 '20 at 22:32
  • I had a hunch that your education variable was ordinal, but what seems ordinal about the political party variable? – Dave Dec 19 '20 at 22:33
  • oh, sorry. Political party variable seems categorical since it has directly the names of the parties. So, what is the the best solution to understand the relation between these two variables? – MilanWoland Dec 19 '20 at 22:43
  • See https://stats.stackexchange.com/questions/73065/correlation-coefficient-between-a-non-dichotomous-nominal-variable-and-a-numer – kjetil b halvorsen Dec 19 '20 at 23:33
  • thank you. the link seems illuminating. – MilanWoland Dec 20 '20 at 08:49

0 Answers0