Correlation between continous and categorical variable

[Diane Smith]

Dear all,

I would like to compute the correlations between several continous variables and a categorical variable.

Specifically, the continuous variables are scores (taking any value between 0 and 1), and the categorical variable is an industry classification (Healthcare, Tech, Consumer Goods, Other).

Which test can I use to compute the correlations between these types of variables? Or should I re-code the industry classification to dummy variables and run a Pearson or Spearman correlation?

Thank you very much in advance for your comments!

Best, Diane

[Nick Cox]

No standard correlation coefficient (at least of Pearson, Spearman, Kendall) makes sense here as your industry classification appears nominal, and not ordinal. You certainly can regress your continuous responses against a categorical variable treated as a series of dummies (indicators).

[Diane Smith]

Thanks, very helpful! That's what I've done now.