Abstract
In an earlier study of voting behavior of U.S. appeals courts judges, attitudinal patterns were investigated along with an analysis of the relationship of judges' backgrounds to their decisions. In this revisit, the earlier findings were treated as hypotheses and tested with a new case population covering a subsequent and longer time period. In all, 2,115 cases decided nonunanimously were coded on one or more issues. Most cases could be classified under ten broad issue categories which were then utilized for most of the analyses. Although the research design was similar to that of the earlier study, a wider variety of methods was employed including nonparametric and parametric intercorrelations of voting behavior on the ten issues and stepwise multiple regression and partial correla-tion analyses of seven background variables and their relationships to voting behavior on the issues. The principal findings were similar to those found earlier but it was possible to map voting behavior with some-what more precision and to uncover some unexpected relationships such as those concerning the potency of the age variable particularly for voting on political liberalism issues.