The Norm of Even-Handedness in Surveys as in Life

Results from four past split-ballot experiments are drawn on to show that a norm of even-handedness operates in surveys much as it does in the rest of life: contextual linkage leads to greater consistency in the treatment of two competing parties than would occur if each were considered in isolation. A fifth experiment, developed to test this theory, provides partial confirmation but also evidence that the norm can create effects in one direction only. In addition, a sixth experiment is reviewed which emphasizes the fact that a presumed needfor psychological consistency is too vague a description of the Jorces at work. Only where consistency receives normative definition and direction does the context effect occur. Finally, an attempt is made to specify types of persons most susceptible to the context effect, but further evidence will be needed in order to achieve more complete understanding on this point.