An alternative approach to analysis of mental states in experimental social cognition research

Establishing the mental states that affect human behavior is a primary goal of experiments on social cognitive processes. Such mental states can be manipulated only indirectly; therefore, after delivering a manipulation, researchers attempt to verify that the mental state of interest, the representation of a mental state, was in fact changed by the manipulation and that this change caused the observed effect. The usual procedure is to examine mean differences in a measure of the mental state of interest (a manipulation check) among experimental conditions and to infer whether the manipulation was effective. We describe a procedure that strengthens the construct validity of manipulations and, hence, causal inferences in experiments that focus on mental states using analyses familiar to most researchers. This procedure employs a traditional manipulation check that assesses the relationship between manipulations and mental states but, additionally, tests the relationship between the manipulation check and dependent measure.