Thus far, we have considered experiments having only one independent variable (IV). However, in the real world it is rare to nd an IV that in«uences criterion measures in isolation from other IVs. Usually two or more IVs contribute to changes in dependent measures. Factorial designs permit examination of the effects of two or more IVs separately (main effects) or in combinations (interactions). An interaction occurs when the effects of one IV vary according to the level of another IV.