This paper examines to what extent, and in what ways, voters are willing to go beyond the information given--when that information consists of either a candidate's personal traits or his issue positions--to make broader candidate assessments. Based on an experiment in which students were given either personal trait or issue information about candidates for president, we find strong evidence for voter inferences from traits to issues and vice versa (although there is greater inference from issues than from traits). Finally, we find that, although inferences are frequently made, they are often idiosyncratic. Only in particular cases (i.e., relating candidate's compassion with his support of government providing jobs), do different types of respondents make the same inferences (a finding which is also borne out by the 1984 NES).
No takes yet. Share an insight, caveat, or question.
Rapoport et al. (1989) studied this question.
Synapse has enriched one closely related paper. Consider it for comparative context: