Announcement

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

  • Marginal effects for count data models from interval data: intcount

    Hi,

    I am using intcount to estimate a count data model with interval data, i. e. the dependent variable is provided in intervals, and I would like to calculate the average marginal effects of all the covariates on the expected number of counts.

    However, after running

    margins, dydx(_all)

    I got the following error message

    default prediction is a function of possibly stochastic quantities other than e(b)
    r(498);

    And the hep file is not very useful.

    Is the command intcount supposed to generate marginal effects?

    Any help would be greatly appreciated.

    Simone

  • #2
    here is a quote from the help file for the user-written (SJ) -intcount- command:
    The predict command can be used postestimation to calculate various
    types of prediction. Other postestimation analyses (such as margins)
    are not currently supported.

    Comment


    • #3
      Hi Rich,

      Many thanks.

      Simone

      Comment

      Working...
      X