Abstract
Consumer and marketing researchers often measure opinion leadership with the seven-item, self-report scale developed by King and Summers (1970). They have devoted little effort, however, to evaluating the psychometric qualities of this scale. The present study suggests that the scale is not unidimensional. Factor analysis reveals a stable two-factor structure for the scale across three samples of consumers. This article presents and tests a revised version of the opinion leadership scale that has superior psychometric qualities as demonstrated on a sample of 185 adult consumers.