Formerly non-significant main effect becomes significant after interaction is entered


New Member
Hi there,
I'm running a hierarchical regression with 3 IVs. First step: main effects, second step: two-ways, third step: three-ways. In the first two steps, one of the IVs was non-significant. Once the three-way interaction was entered in the third step (which was non-significant), the IV suddenly became significant. Why might that be the case?