Ryu, Gangseog; PhD

                         UNIVERSITY OF PITTSBURGH, 1998

                         BUSINESS ADMINISTRATION, MARKETING (0338); PSYCHOLOGY, SOCIAL (0451); SPEECH
                         COMMUNICATION (0459)
 

                         Although there has been a long tradition of research on antecedents and consequences of WOM in the
                         marketing literature, surprisingly few systematic attempts have been made to explore the process of
                         word-of-mouth communication (hereafter WOM), especially what and how information is transmitted to
                         other consumers. This thesis proposes various ways to examine patterns of information transmission and
                         change. First, information is classified according to (1)&nbsp;level of abstractness (<italic>factual details -
                         elaborations abstractions - global evaluations</italic>) and (2)&nbsp;consistency with the theme of WOM
                         (<italic> consistent and inconsistent</italic>). Second, types of changes are analyzed based upon
                         (1)&nbsp;how the informational content of information changes (<italic> complete omission - partial
                         omission - error - reproduction</italic>) and (2)&nbsp;how the persuasive strength of information
                         changes (<italic>stronger - same - weaker</italic>). The thesis investigates three primary factors that
                         affect consumer information transmission behavior. These variables are (1)&nbsp;valence of WOM
                         (positive WOM vs. negative WOM), (2)&nbsp;type of social relationship between the consumer and the
                         listener (strong ties vs. weak ties), and (3)&nbsp;presentation format of the information (an overall rating
                         is included vs. excluded). Hypotheses are developed based upon previous research on rumor,
                         dissatisfaction, negativity bias, the NUM effect, confirmation biases, and social networks. The results
                         show that different types of information are transmitted and that information undergoes systematic
                         changes in WOM as a function of the experimental factors. Moreover, these patterns of information
                         transmission are shown to have differential effect on recipients&rsquo; evaluations of the product.
                         Several key findings emerge. First, more factual details and less global evaluations are transmitted in
                         negative than positive WOM, and to strong than weak ties. Second, the study emphasizes the
                         asymmetric impact of negative information on various dimensions of information transmission. For
                         instance, negative-inconsistent information is less likely to be completely omitted, less prone to error,
                         and more likely to be reproduced than positive-inconsistent information. Third, the effect of valence of
                         information is moderated by overall rating. In a number of instances, the confirmation biases induced by
                         the overall rating reduce or reverse the differential impact of negative information. Fourth, consistent with
                         the previous findings of social network research, subjects with strong and weak ties behave differently in
                         WOM. Moreover, it is shown that communicators modify the content of WOM by considering the tie
                         strength that they share with the recipient. For example, more inconsistent information is transmitted
                         when the recipient is a strong tie than a weak tie. In addition, the different behavior across tie strength is
                         moderated by the valence of WOM, indicating that, for instance, the difference between positive and
                         negative WOM is greater for strong ties than weak ties. Fifth, the patterns of information transmission and
                         changes lead to the finding that the difference between the communicator attitudes and the recipient
                         attitudes is larger in positive than negative WOM, although the recipient attitudes become less extreme
                         in general than those of communicators. (Abstract shortened by UMI.)

Social Systems Simulation Group P.O. Box 6904 San Diego, CA  92166-0904 Roland Werner, Principal Phone/FAX  (619) 660-1603