Dear All, my query is related to the interaction effect in the instrument variable probit model. When I running the ivprobit model with my main variables and their interaction (in addition to other control variables) the coefficient of the main variables are coming out to be significant but the coefficient of the interaction effect is insignificant. Whereas, in the case when I am running ivprobit with only the interaction effect variable (not using the main effect variable) the interaction effect coefficient is coming out to be significant. Kindly, guide me on what should I do in this regard, i.e. should I report the regression result with main variables and interaction effect or result with interaction effect only.
Looking forward to suggestions

Thanks