Interaction analysis

[umaganesh]

In my interaction analysis in smart pls, the beta value of the interaction term is negative (should be positive as per the model asssumption), though the beta values of the moderator and the independent variable are positive. The t-value for all the 3 variables are significant and the r-squared value of the dependent variable also increses due to the moderating effect. Should we consider the negative beta while intrepreting moderating effect or can we stick with change in r-squared value and ignore beta value? Please clarify. 

[bradley7680]

Hi,

See much of my work on this topic. at researchgate website link below.

Is there a conceptual possibility for why the interaction effect could be explained as negative?

brad