This might be a silly question, but I'm confused why ANOVA is different from a panel data model, with dummy variables for group, and inference on the coefficients on the dummy variables. Is this what ANOVA is? Am I missing something important about ANOVA? Thanks for your help.

Related to Linear Regression and ANOVA

Realized that this is a duplicate: Why is ANOVA taught / used as if it is a different research methodology compared to linear regression?

  • 2
    $\begingroup$ Depends more than a little on how you define panel data models. But's it's correct that simple versions of ANOVA can be recast as regression with indicator (you say dummy) variables. You would have a hard time translating all the extra apparatus surrounding ANOVA into regression terms; opinions vary on how much would be missed. $\endgroup$ – Nick Cox Feb 11 '15 at 9:54
  • $\begingroup$ Can you provide an example of something that would be missing / difficult to recast in a regression? $\endgroup$ – user103828 Feb 11 '15 at 10:22
  • 1
    $\begingroup$ Look at the later chapters of any text emphasising ANOVA and especially what you do afterwards. (In my previous comment "But's" is a bad typo.) $\endgroup$ – Nick Cox Feb 11 '15 at 10:30

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.