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

  • Probit model explanation

    I have a dataset that contains information on applications to any job vacancy in an organization. I want to know if my Y (Y=1 if the application is from a woman) depends on the nationality of the applicant, so I created a dummy for every nationality (there are 36 in my sample).
    Here are my questions:
    1) How can I interpret the relation between gender and nationality with all these dummies? Does it make sense to include all of them (or to include n-1) in the regression?
    2) How do I interpret the coefficients in Probit?
    3) How to calcolate the margins given all the dummies? Should I do it for all the variables?

    Thanks a lot for any clarification you might have on this!

    PS: My other independent variables are all categorical: whether the application is internal, type of job, year of the application (here I also created multiple dummies for 6 years), years of experience of the applicants.

  • #2
    You'll increase the chances of a useful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output, and sample data using dataex. Also, assume we're not from your area

    I would be inclined to include all the dummies - if you only include some (and they're not orthogonal), then you can suffer from omitted variables bias. If all the dummies are orthogonal, this may not be a problem.

    Use factor variable notation and margins after probit to help interpret the results. How many you do it for is really a question of your research substance.