In: Statistics and Probability
An interaction term is the multiplication of two or more explanatory variables
true or false
TRUE
explanation ----
Consider this model:
Y=β0+β1X1+β2X2+ε.
Now, the marginal effect of X1on Y(Y's derivative w.r.t X1) is simply β1. I can replace β1 with anything, that will be X1 's effect (assuming it doesn't contain X1):
Y=β0+(<what we write here will be X1 's effect>)X1+β2X2+ε
In the first case, we wrote β1, which is a constant, so the effect of X1is independent of X2 That's why there is no interaction. We want to introduce an interaction, i.e., that X1 's effect depends on X2? Let's write something there which depends on X2 ? For simplicity, be linear here as well, i.e., assume that the dependence itself is also linear:
Y=β0+(β1+ɣ12X2)X1+β2X2+ε..
That is, the effect of X1 will be β1+ɣ12X2 so we have the interaction as we aimed for: the effect of X1 depends on the value of X2. (In this simplest case, linearly.)
Now we expand it:
β0+β1X1+ɣ12X2X1+β2X2+ε
and voilá, we have the product.
interaction term is the multiplication of two or more explanatory variables.