Probit regression - Interpretation of marginal effects of IVs in log/percentage form

[MSCBUS]

Dear talkstats-community,

I already spent hours on websearch for an answer on a question I have regarding the interpretation of marginal effects (at means) of probit regressions.
How can I explicitly interpret a marginal effect of a variable that enters in

1) logarithmized form (eg. monthly turnover (log))
2) in form of a percentage? (eg. profit share)

I would be very thankful for your help or an example.

Best
M.

 

[JesperHP]

Re: Probit regression - Interpretation of marginal effects of IVs in log/percentage f

I would think that you can interpret marginal effects (at means) as:
1) Change in probability of succes as a result a 1% change in monthly turnover...
2) Change in probability of succes as a result of a one percentage point change in profit share...