Dear Statalist community:
I have a logit model predicting DV (0 or 1) with two continuous independent variables, X1 and X2 where I also include an interaction term between X1 and X2.
I find thatX1*X2 is statistically significant. I would like to plot the predicted probabilities of DV on the y-axis, the full range of values of X1 on the x-axis, and two predicted probabilities conditional on two specified values of X2.
How can I know and then plot the range of X1 where the predicted probabilities of DV at the lower value of X2 is statistically significant from predicted probabilities of DV at the higher value of X2?
The lower and higher values of X2 are my own specifications based on the distribution of X2 in my sample.
I have studies the margins command but cannot find the proper sub command.
Thank you very much!
Related Posts with statistically significant difference between predicted probabilities in continuous * continuous interactions
Is it valid to apply fixed effects or GMM method to this model?The model is attached. This model is special as it includes macro environment variables (FB and X) w…
Endogenous Treatment Effects PostestimationHello all, Is there any reason, technical or otherwise, that there is no test of overlap or balance…
ivreg2 and Partialling-out exogenous regressorsDear STATA Community, i am estimating following model using panel data with dummy control variables …
Graphing two categorical variablesHello everyone, I'm new to the Forum and relatively beginner at Stata, so sorry if the answer for w…
Issue in estimating non-linear seamingly unrealated regressionHi, I am working on a research project in finding market power of the rice milling industry in Sri L…
Subscribe to:
Post Comments (Atom)
0 Response to statistically significant difference between predicted probabilities in continuous * continuous interactions
Post a Comment