We investigated the use of rating, ranking, constant sum, and paired comparison tasks for evaluating both the importance and support for a wide range of political issues. In 8 waves of fielding (running from October, 2003 to November, 2004), we had U.S. respondents participate in a web-based survey. For respondents of the 7 pre-election waves, we recontacted them following the November elections to complete another evaluative task. This allowed us to determine both the test-retest reliability and predictive validity of the initial evaluations. We replicated many of our earlier findings concerning the differences between rating and comparative trade-off tasks and further extended our understanding of the relative comparability and utility of the different measures across time.