ABSTRACT
The general hypothesis in this research is that to predict how a voter will process and respond to a persuasive message it is necessary to account for the moderating effect of an attitudinal bond or relationship between a voter and a candidate. We argue that the bond between a person and an attitude object may be multidimensional, rooted in one or more of several independent psychological processes. Recent work on attitude strength (Abelson, 1988; Raden, 1985) has conceived of the construct as multidimensional. If that is so, then it is possible that each component of a person–object attitudinal bond may moderate differentially the processing of messages that relate to an underlying attitude.
