Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    The R-squared from the "dummy variable regression" is typically very high. But you shouldn't really get credit for including the dummy variables. As Carlo said, use the within R-squared, as it nets out the dummies. The usual R-squared can be used for pooled OLS without the unit-specific dummies.

    There is a way to net out the time effects, too, but that's often less of an issue.

    Comment

    Working...
    X