I am examining the relationship between a categorical variable and a quantitative one. For the test of linearity, I was told that I must use scatter plots. However, this is no longer applicable if one of the variables is categorical. So I resorted to box plots. Now, I have difficulty identifying/checking if there is a linear relationship among males and females (gender is the categorical variable) in terms of their side by side box plots. Can anyone help me with this?

  • $\begingroup$ Relationship between a quantitative variable and a dichotomous variable (such as sex) is but linear. The shortest way to connect the two levels (midlines in the boxes) is a straight line $\endgroup$ – ttnphns Jul 1 '12 at 8:07
  • 1
    $\begingroup$ is it always linear ? so if that's the case, I can move on to another test (assumption of normality) $\endgroup$ – Deah Jul 1 '12 at 8:23
  • 3
    $\begingroup$ "Linear" and "nonlinear" make no sense when discussing the relationship between a categorical and a continuous variable. As @ttnphns said, you can always draw a straight line, but so what? $\endgroup$ – Peter Flom Jul 1 '12 at 11:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.