How Serious are Methodological Issues in Surveys? A Reexamination of the Clarence Thomas Polls

Opinion polling procedures allow for reasonable inferences about attitude changes. We examined this contention using surveys about the nomination of Clarence Thomas. In this situation, prior theory allowed us to predict the direction of changes, surveys had been conducted by a number of organizations, and substantial information was available about the methodology used in the surveys. As a result we concluded that the deteriorating opinions of Thomas were real.