Using Signal Detection Theory Indexes for the Experimental Manipulation of Emotional States

Abstract

Classic studies regarding emotional processing, either in normal subjects or in pathological groups, rely on the basic assumption of a linear correspondence between stimuli input and perceived emotional information. This neglects the evidence that different experimental groups may perceive the same stimuli features differently. Thus, when emotional input is to act as an independent variable, it becomes fundamental to assure equivalent emotional loading in all groups under comparison, even if using different stimuli. This is not a commonly used procedure when studying emotion or its influence on other cognitive processes. In order to address this issue, a signal detection theory based methodology is presented, which permits measuring sensitivity indices for emotional stimuli, while controlling participants' subjective criteria. This proposal may represent a useful methodological approach for the study of several emotion-related problems and albeit applicable to any group of subjects.