Close

Complete Captcha

• 5 years - 43 posts - Latest - RSS
• Thread: 326 Goods vs 229 No Goods

1. Economist
   c30f
   

   If you need to include an interaction term, it shows that your model is wrong. Either your causal structure is wrong, or you're not in fact comparing the same treatment.

   Interaction terms are just band aids that save you from having to correct your earlier mistakes in model specification.

   5 years ago # QUOTE 13 Good 148 No Good !

2. Economist
   b848
   

   Please let this be a troll. Please let this be a troll. Please let this be a troll.

   5 years ago # QUOTE 101 Good 2 No Good !

3. Economist
   204c
   

   Explain why interaction terms imply lazy/mistaken theory like I'm a reg monkey.

   5 years ago # QUOTE 12 Good 2 No Good !

4. Economist
   405c
   

   statsbro here. you are a dumbass

   5 years ago # QUOTE 40 Good 2 No Good !

5. Economist
   5988
   

   OP has a problem of interaction with others.

   5 years ago # QUOTE 57 Good 1 No Good !

6. Economist
   c30f
   

   I suggest that rather than make unfounded criticisms of my position, you focus your energies on properly specifying your models.

   It would serve you very well indeed.

   5 years ago # QUOTE 3 Good 28 No Good !

7. Economist
   e2b1
   

   OP must be gt01. We all know he wasn't really a game theorist.

   5 years ago # QUOTE 3 Good 3 No Good !

8. Economist
   81a1
   

   @c30f, reference, please.

   5 years ago # QUOTE 2 Good 1 No Good !

9. Economist
   c30f
   

   Additivity is a central feature of our universe.

   If your model departs from this, then you should indeed have strong concerns.

   5 years ago # QUOTE 5 Good 28 No Good !

10. Economist
    e88c
    

    You should really read Hal white's book.

    5 years ago # QUOTE 1 Good 0 No Good !

11. Economist
    a510
    

    Additivity is a central feature of our universe.
    If your model departs from this, then you should indeed have strong concerns.

    Friday night. OP bored in front of computer.

    5 years ago # QUOTE 15 Good 0 No Good !

12. Economist
    d344
    

    OP. Regression discontinuity. Comment.

    5 years ago # QUOTE 8 Good 0 No Good !

13. Economist
    9c71
    

    OP, we already knew our model was wrong. All of yours are too.

    That doesn't mean interaction terms cannot help you summarize the data.

    Ever wondered if the correlation between wages and education depends upon where you were born? That is an interaction term, and it is interesting even without a causal interpretation.

    Hell, you could even say that big Raj Chetty paper on intergenerational mobility is one giant interaction term.

    5 years ago # QUOTE 6 Good 0 No Good !

14. Economist
    87a4
    

    The troll is strong with this one. But it's true that interaction terms often have substantive interest in their own right, and sometimes signal problems to do with specification.

    5 years ago # QUOTE 4 Good 0 No Good !

15. Economist
    c7ba
    

    So the diff-in-diff methodology is always no good?

    5 years ago # QUOTE 22 Good 0 No Good !

16. Economist
    ada9
    

    OP alpha as f**k.

    5 years ago # QUOTE 3 Good 4 No Good !

17. Economist
    0743
    

    So, OP, just where did you learn that reality is additive? Did God tell you this while you were wiping your behind after taking a crap?

    5 years ago # QUOTE 2 Good 0 No Good !

18. Economist
    a190
    

    OP does structural econometrics, confirmed.

    5 years ago # QUOTE 1 Good 3 No Good !

19. Economist
    c30f
    

    I see that there are no substantive responses to my position.

    I suspect that you also wail and scream to editors whenever you feel my wrath. But that only serves to paint you in a poor light.

    5 years ago # QUOTE 1 Good 3 No Good !

20. Economist
    8db6
    

    If in my model I have a p*q term means I should only regress on p data and q data and not on interaction. I would rather question if I don't have to regress only on interaction and not on the independent variables.

    5 years ago # QUOTE 0 Good 0 No Good !

Reply »