Effects of affective-semantic mode of item presentation in balanced self-report scales: biased construct validity of the Zung Self-rating Depression Scale

Synopsis The widely applied procedure of balancing self-report instruments by including positively and negatively keyed items is exemplified by the Zung Self-rating Depression Scale (SDS). Investigation of the influence of the symptom-positive and symptom-negative item modes on the SDS in a depressed population resulted in two major findings. First, the reversed scoring of the symptom-negative items resulted in higher mean item scores. Secondly, factor analyses of the SDS in the present study and in previous research revealed that the semantic modes of item presentation were represented in the factor structure of the SDS. These findings were confirmed by analyses with the State–Trait Anxiety Inventory (STAI) and by previous factor analytical research with balanced instruments and were interpreted within the framework of the theory of Positive and Negative Affect. The present data cast doubts on the construct validity of the SDS as a measure of depressive symptomatology due to the presence of the negatively keyed items and suggest reconsideration of the use of balanced instruments for minimization of the acquiescence response set.