I am running a regression with a continous DP (scale of self-assesement of a person´s willingness to accept pay cuts to protect the climate) and two continous IVs (1=age 2=value-index) and their interaction.
My hypothesis is that age has a negative effect on a person´s willingness to accept said paycut and want to control it with IV2 for a person´s political values. In short: When holding values constant, does age still have the negative effect?
In Model 1, the Interaction and IV 2 are significant, while the IV 1 is not. In Model 2 only IV 2 is signficant. I have created marginplots but need some input in regard of the interpretation of the signficance in both Models.
1. Can I say that IV 1 has no significant effect on the DP in Model 1 and 2 and therefore I cannot accept my hypothesis?
2. How do I interpret an non-siginifacnt Interaction Term?
3. Is it still okay to show the plots of average marginal effects and predicitve margins for illustrative purposes and add a "non-significant but still relevant - mark"?
Here for more visual explanation
|Model 1||Model 2|
Thank you very much for any input