Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Communication Studies

First Advisor

Renee Edwards


The purpose of the present dissertation was to examine the influence of reference group membership on message interpretation and whether one's interpretation then influences one's agreement with the message source and subsequent willingness to speak with the message source about the topic. In particular, this study looked at conservative Christian reference group membership and message interpretations about the topic of abortion. The independent variables of level of Christian fundamentalism, involvement in a Christian reference group, and socio-political conservatism were averaged to create a conservative fundamentalist composite score which was then used to split the sample at the median. Findings indicated that persons with a high conservative fundamentalist score interpreted the intent of an abortion message as conservative if from a conservative source and as liberal if from a liberal source. Liberal respondents interpreted messages in the same way, however, which indicates that message source is a strong mediating variable. Structural equation modeling (path analysis) was used to determine whether conservatives had greater agreement with a perceived conservative message source than with a perceived liberal message source, and subsequently, whether agreement predicted willingness to speak to the source about the topic of abortion. These hypotheses were supported. Results indicated that both conservative and liberal persons were willing to speak to the source they agreed with. The final hypothesis was based on Elisabeth Noelle-Neumann's Spiral of Silence theory of public opinion and predicted that conservative men would be more willing than conservative women to speak to a perceived liberal source about abortion. While no direct support for this hypothesis was found, results indicated that liberal respondents were more willing to speak to a liberal source about abortion than were conservative respondents.