Dear Stata forum members,

Would you please help with interpretation of my obtained results. I have one continuous variable and one binary variable in my model. When I estimate their individual effects on DV, their coefficients are positive and significant which suport the hypotheses. However, when I take their interaction, the main effects become insignificant ( the coefficient of the continuous variable even became negative) though the interaction term is significant. Should I focus on the interaction term without looking at the main effects? Or is something wrong with my data and or model?

Thanks in advance!