Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Association meassure with 1 ordinal and 1 nominal variable and regression

    I´m working with 1 ordinal variable and 1 nominal variable from a probabilistic sample (1000 cases), as of now i have been doing descriptive analysis of tables, but i have been advised to use association meassures (specifically Cramer´s V). I wanted to know if Cramer´s V is the best option in this case and if there are other coefficients I should consider.

    Also, I got a Cramer´s V = 0,173

    Would this be considered a weak, to mild association ? (the field is Sociology)

    I´m also considering a regression model, the nominal variable has 3 categories and its the dependent variable, and the ordinal variable would be the independent (along with gender and age group for control), would a multinomial logit regression be the correct choice in this scenario?


    Thanks in advanced!

  • #2
    Sorry for bumping, I know it´s not actually a Stata question but i always get the best responses in this forum and I´ll the work i´ve been doing for this paper is with Stata.

    Comment

    Working...
    X