The aim of this project was to examine which factors are important in influencing communication between health professionals and patients. Communication Accommodation Theory (CAT) was the theoretical framework adopted in this project. CAT proposes that individuals are influenced by their personal and social identities and that, in many cases, it is an individual's group or social identity that is most salient in an interaction. The underlying theoretical assumption in this project was that communication between health professionals and patients is characterised by intergroup rather than interpersonal salience and convergent methodologies were used to test this proposition. In addition to CAT, the linguistic category model (LCM), which is also used to investigate individuals' perceptions of intergroup salience, was adopted to complement the findings derived from CAT. There were seven studies in this project. The first study described the methodology for obtaining the data set used in Part 1 of the thesis. Participants wrote retrospective descriptions of a satisfactory and unsatisfactory conversation with a hospital staff member which they had experienced as a hospital in-patient. Study 1 provided a profile of the participants. In Study 2 the stimuli were participants' written recollections of 69 unsatisfactory and 79 satisfactory conversations. The LCM was used to test for differences in participants' perceptions of differing levels of intergroup salience between the two types of descriptions. While intergroup bias was evident, the results were complex. In Study 3 a qualitative and quantitative analysis of the same data set was conducted. This study explored differences between patients' and health professionals' goals, sociolinguistic strategies, and the patients' descriptions of the health professionals across the two types of conversational descriptions. Results revealed differences in goals, strategies, and descriptions. In the descriptions of satisfactory conversations, participants reported goals concerning reassurance and developing relationships. These two goals were not evident in the descriptions of unsatisfactory conversations, where attending to role relations was of greater importance. Participants reported the use of different sociolinguistic strategies for themselves and for health professionals across the different conversation types, which indicated that, for reports of satisfactory conversations, participants viewed their interactions with health professionals as more personal and positive than in the unsatisfactory ones. Participants described the health professionals in their descriptions of satisfactory conversations in more undifferentiated terms than the health professionals in the descriptions of unsatisfactory ones. Study 4 comprised two parts. First, using the data set from Studies 2 and 3, a qualitative analysis was conducted which explored thematic differences between the descriptions of satisfactory and unsatisfactory conversations. This analysis indicated more interpersonal themes for the satisfactory interactions and more negative intergroup themes for the unsatisfactory ones. Second, 134 participants rated 16 exemplar descriptions from the data set on 13 items derived from CAT. In general, the exemplars of satisfactory descriptions were rated as containing accommodative use of discourse management, emotional expression, and interpersonal control strategies. Exemplars of the unsatisfactory interactions were rated as more overaccommodative or counteraccommodative on these strategies. Study 5 introduced Part 2 of the thesis and involved a second data set. The stimuli were real-time videotaped interactions between health professionals and patients. Participants rated 25 videotaped interactions on 28 questions developed to tap the intergroup and interpersonal salience of the interaction. Of these 25 interactions, participants rated seven as highly intergroup and seven as highly interpersonal. These 14 interactions formed the stimuli for the Study 6. In Study 6 an LCM analysis of these 14 videotaped interactions compared the intergroup with the interpersonally rated interactions for levels of intergroup bias. The interpersonally salient interactions suggested lower intergroup bias than did the intergroup ones. This and other findings from the LCM analysis are discussed. Study 7 used a CAT perspective to examine six of the 14 videotaped interactions used in Study 6. These six interactions represented three videotaped interactions that were rated as highly intergroup, and three that were rated as highly interpersonal.. Participants provided ratings on both interactants' strategies, and health professionals' goals. They also rated the interactions for outcome measures, including patient satisfaction, and effective communication. Results indicated that the interactions rated as interpersonally salient were perceived as attending to relationship needs and emotional needs. Generally the interpersonal interactions were also rated as providing a more satisfactory outcome than the intergroup interactions. Together the seven studies provide evidence that interactions between health professionals and patients are essentially intergroup interactions. However, such interactions have the potential to move through dimensions of high intergroup and high interpersonal to dimensions of low intergroup and low interpersonal. Each pair of combinations provides different outcomes of communication effectiveness and satisfaction for the patient. Further, this research project demonstrates the ability of CAT to pick up on the dynamics of health provider and patient communication. It also highlights the usefulness of convergent methodologies to understand the complexities of patient and health professional interactions.

Two studies were conducted to investigate how the use of different types of language affects attitudes. Participants scrutinized arguments supporting a hypothetical toothpaste that differed in terms of argument strength (strong versus weak) and linguistic abstractness (abstract versus concrete) and subsequently evaluated the toothpaste. In addition, half of the participants in the second study were subjected to a cognitive load manipulation (i.e., rehearsing a ten-digit number) in order to limit their level of cognitive elaboration. Results indicated that strong arguments and those containing concrete descriptions led to more positive attitudes about the toothpaste, whereas weak messages comprised of abstract terms gave rise to the least favorable evaluations. These findings represent the first demonstration of the effect of language type on attitudes and suggest that future research into the functions of differential linguistic abstractness in a persuasive context will broaden our understanding of attitude change.

For years, researchers have examined financial data in corporate earnings announcements and their influence on market participants. More recently, a body of research has been developing recognizing the impact of narrative disclosures and managers' deliberate language choices. However, no prior studies have investigated those language choices of managers which are likely unintentional in composing such narratives; language choices which, as previous research has revealed, escape conscious access. Using an empirically-grounded model which systematically classifies different predicates, I examined whether managers use systematic patterns of language when construing the earnings press release in a likely unintentional effort to channel or direct readers' attention. I found that managers write positive information using a more concrete construal than negative information. Additionally, I used experimental data to examine whether these systematic differences lead to different perceptions of the company and its value as an investment alternative. Nonprofessional investors performed an analysis of an earnings press release where I manipulated the valence of the narrative as positive or negative and the construal of the narrative as abstract or concrete. I found that these manipulations had an interactive influence on investment decisions. Specifically, investors were least likely to invest when a negatively valenced narrative was written concretely. I also found that the influence of the narrative on the investment decision was direct and not the result of the narrative influencing the investors' focus of attention on the accompanying financial statements. Additionally, I tested whether the investor judgments were due to intentional cognitive effects and found that the influence of the narrative on the investment decision was not conscious on the part of the investor. Lastly, I conducted an analysis of archival data to examine the relationship between managers' language use in forward-looking statements of the earnings press release and future firm performance and the extent to which the market responds to these linguistic clues. Results from the analysis suggest that construal is predictive of future firm performance and the market is incorporating this into pricing for firms that meet or beat earnings expectations.