Development of a Punishment Severity Scale: The Item Displacement Phenomenon

Abstract

A number of studies suggest that crimes and punishments can be scaled. Some evidence indicates, however, that development of such scales may be hampered by the number of items which appropriately can be evaluated by respondents. In the present study two survey instruments were administered to 90 respondents. The results suggest that although the rankings of identical items derived from two survey instruments (one containing a large number of items and the other containing a subset of the larger set) were highly correlated, the instrument containing the larger number of items resulted in several instances of scale item displacement.