0
$\begingroup$

My dataset consists of a numeric variable (called "N4") and several categorical variables that affect the numeric variable. For example there is a categorical variable called "die" that if it equals "alpha" then N4 takes values around 100, if it equals "beta" then N4 takes values around 300.

My goal is to figure out which of the categorical variables most affects my numeric variable.

Can it make sense to turn categorical variables into numerical variables and calculate correlation? Is there any other more effective analysis?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
0
$\begingroup$

You could use the following different methods:

  1. Point-Biserial correlation: measures strength and association between a continuous and dichotomous variable.
  2. ANCOVA: to check which categorical variable has significant relation with the continuous variable (Variable with least p-value is most significant). Then use pairwise comparisons tests and confidence interval to check which pairs of groups differ
  3. Kruskal-Wallis H test: non-parametric similar to ANOVA, does not assume normality.
Improve this answer
$\endgroup$

Your Answer

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

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