This study reports a portion of a series of investigations designed to determine to what extent credibility scales can be generalized. The investigation data was collected in six phases, representing six subject populations. All of the subjects were either college students or non-student adults. Subjects were asked to respond to one of four political figures, each representing a wide range of political viewpoints. Semantic differential-type scales, representing dimensions of source credibility reported by various researchers, and Likert-type statements on an eleven-step continuum bound by bipolar adjectives were used to measure potential communication behavior and response to communication behavior. The results suggested that researchers should not expect exactly the same dimensionality of source credibility for all subject populations. (Author/LG)