<%BANNER%>

A Public Relations Campaign of Corporate Social Responsibility: A Test of a Cognitive Processing Model of a CSR Message


PAGE 1

A PUBLIC RELATIONS CAMPAIGN OF CORPORATE SOCIAL RESPONSIBILITY: A TEST OF A COGN ITIVE PROCESSING MODEL OF A CSR MESSAGE By SHU YU LIN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS IN MASS COMMUNICATION UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Shu Yu Lin

PAGE 3

This thesis is dedicated to my father and my mother. Without their love and support, I could not have accomplished this study.

PAGE 4

ACKNOWLEDGMENTS The accomplishment of this thesis could not have realized without the help of many people. First of all, I want to express deep appreciation for my parents, Ji-Shun Lin and Pi-Hua Su, for their endless love and support. Their beliefs in me had given me strengths throughout the difficulties in completing the degree and the thesis. They have given me more than I could ever acknowledge. I would like to thank my committee chair, Dr. Spiro Kiousis, and my committee members, Dr. Margarete Hall and Dr. Mary Ann Ferguson, for their support of carrying out this research study. I also want to express my gratitude to Dr. Lynda Kaid, Dr. Amanda Ruth, and Dr. Tracy Irani for their generous help to accommodate the experiment in their class periods. I would like to attribute special thanks to Joon Soo Lim for his generous help and insightful suggestions. I am especially grateful for the opportunity to work together with him for a paper submitted to the conference of AEJMC. Without him, I could not have hoped to accomplish such a challenging and creative experimental study like this. I like to thank my friends at UF who have lifted my spirits during the stressful time of writing the thesis: Chin-Hsin Liu, Wan Ping Chao, Yimin Wung, Yang Hsin Hsu, Chen-Hsuan Chen (Judith), Ting Bing Wu, and Allan. In particular, I want to thank all the friends who participated in the pilot study and provided useful suggestions. In particular, I would like to show my appreciation to Camelia Baluta for her generous help of editing my thesis. iv

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES...........................................................................................................viii LIST OF FIGURES.............................................................................................................x ABSTRACT.......................................................................................................................xi CHAPTER 1 INTRODUCTION........................................................................................................1 Corporate Social Responsibility Communication........................................................1 CSR Communication from the Issues Management Perspectives................................2 Statement of the Purpose..............................................................................................4 Background...................................................................................................................4 Research Questions.......................................................................................................5 Theoretical Framework.................................................................................................5 2 LITERATURE REVIEW.............................................................................................8 Corporate Social Responsibility...................................................................................8 The Origin of CSR.................................................................................................9 CSR Definition....................................................................................................10 Perceived CSR and Public Relations...................................................................11 CSR and Issues Management..............................................................................12 Corporate Credibility...........................................................................................13 The Elaboration Likelihood Model............................................................................15 The Theoretical Framework of the ELM.............................................................16 Central Route and Peripheral Route....................................................................17 The Variable of Personal Relevance/Involvement..............................................18 Involvement and Environmental Advocacy........................................................19 Message Sidedness Effect...........................................................................................20 Definitions...........................................................................................................20 Origins.................................................................................................................21 Previous Studies..................................................................................................22 Advertising and Non-Advertising Messages.......................................................22 Prior Knowledge, Motivation, Personality..........................................................23 v

PAGE 6

Issue Involvement................................................................................................24 Three Theoretical Accounts for the Message Sidedness.....................................25 Inoculation Theory..............................................................................................26 Correspondence Theory.......................................................................................26 Cognitive Response Theory................................................................................28 Hypotheses..................................................................................................................30 3 METHODOLOGY.....................................................................................................32 An Experimental Design.............................................................................................33 Pilot Study..................................................................................................................35 Procedure.............................................................................................................35 Pilot Test Results.................................................................................................37 Main Study..................................................................................................................38 Sample.................................................................................................................38 Stimuli.................................................................................................................39 Procedure.............................................................................................................41 Operational Definitions: Independent Variables.................................................42 Involvement..................................................................................................42 Message sidedness........................................................................................43 Operational Definitions: Dependent Variables...................................................45 Cognitive elaboration...................................................................................45 Perceived CSR..............................................................................................45 Attitudes.......................................................................................................46 Manipulation Check............................................................................................47 Involvement..................................................................................................47 Message sidedness........................................................................................48 4 FINDINGS..................................................................................................................49 Profile of Participants.................................................................................................49 Results of Manipulation Checks.................................................................................50 Involvement.........................................................................................................50 Message Sidedness..............................................................................................51 Reliability Checks for Dependent Measures..............................................................52 Test of Hypotheses.....................................................................................................53 Cognitive Elaboration..........................................................................................53 Main Effects of Message Sidedness on Perceived CSR......................................54 Interaction Effect on Perceived CSR...................................................................56 Attitudes toward a Company...............................................................................58 Interaction Effect on Attitude toward a Company..............................................60 A sub Analysis between Two Sample Groups....................................................62 A Correlation between Perceived CSR and Attitudes toward a Company.........62 A Hierarchical Regression Analysis between Perceived CSR and Attitudes toward a Company...........................................................................................63 Path Analysis.......................................................................................................66 vi

PAGE 7

5 DISCUSSION.............................................................................................................71 Over View of the Present Study.................................................................................71 Overview of the Hypotheses and Research Question.................................................74 6 CONCLUSION...........................................................................................................81 Discussions about the Effect of Message Sidedness..................................................82 Credibility Perceptions........................................................................................83 Pre-Attitudes toward an Issue..............................................................................84 Fear Arousal........................................................................................................85 Application.................................................................................................................85 Limitations..................................................................................................................86 Suggestions for Future Research................................................................................87 Knowledge Level.................................................................................................87 Message Length...................................................................................................87 Source Credibility................................................................................................88 Attitudes toward Issues........................................................................................88 APPENDIX A CALCO ABOUT USUNITED STATES.............................................................90 B CALCO ABOUT USTAIWAN VERSION.........................................................91 C CALCO ENVIRONMENTAL COMMITMENT-UNITED STATES...................92 D CALCO ENVIRONMENTAL COMMITMENT-TAIWAN.................................93 E CALCO MYTH & FACT.......................................................................................94 F CALCO ENVIRONMENTAL COMMITMENT-NONREFUTATIONL.............95 G QUESTIONNAIRE FOR HIGH INVOLVEMENT GROUPS.................................96 H QUESTIONNAIRE FOR LOW INVOLVEMENT GROUPS................................101 LIST OF REFERENCES.................................................................................................106 BIOGRAPHICAL SKETCH...........................................................................................115 vii

PAGE 8

LIST OF TABLES Table page 2-1 Factorial design of the experiment...........................................................................31 3-1 Liker-type items of the manipulation check of involvement level for high involvement groups..................................................................................................36 3-2 Liker-type items for the manipulation check of message sidedness........................37 3-3 The conditions of the 2x3 experimental design.......................................................39 3-4 Liker-type items for perceived CSR.........................................................................46 3-5 Semantic differential type items for attitude toward the company..........................47 4-1 The number of participants in each cell...................................................................50 4-2 The manipulation check of involvement..................................................................51 4-3 The manipulation check of message sidedness........................................................52 4-4 A t-test of mean difference of message-related thoughts generated between the high and the low involvement group........................................................................53 4-5 A t-test of mean difference of message-related thoughts generated between three groups of sided messages.........................................................................................54 4-6 Perceived CSR by message sidedness......................................................................55 4-7 F-test of perceived CSR by message sidedness.......................................................55 4-8 Perceived CSR by message sidedness......................................................................56 4-9 Interaction of involvement and message sidedness on perceived CSR....................57 4-10 Tests of between-subject effects (dependent variable: perceived CSR)..................57 4-11 Attitudes toward CALCO by message sidedness.....................................................59 4-12 F-test for attitudes toward CALCO by message sidedness......................................59 viii

PAGE 9

4-13 Attitudes toward CALCO by message sidedness.....................................................60 4-14 Interaction of involvement and message sidedness on attitudes..............................61 4-15 Tests of between-subject effects (dependent variable: attitudes).............................62 4-16 A Pearson Bivariate Correlation between perceived CSR and attitudes..................63 4-17 Hierarchical regression analysis for proposed model..............................................64 ix

PAGE 10

LIST OF FIGURES Figure page 1-1 The cognitive processing model of a CSR message proposed in this study..............7 4-1 Interaction of involvement and message sidedness on perceived CSR....................58 4-2 The original cognitive processing model of a CSR message proposed in this research.....................................................................................................................69 4-3 The path diagram based on the path analysis...........................................................69 4-4 The revised cognitive processing model of a CSR message proposed in this research.....................................................................................................................70 x

PAGE 11

Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts in Communication A PUBLIC RELATIONS CAMPAIGN OF CORPORATE SOCIAL RESPONSIBILITY: A TEST OF A COGNITIVE PROCESSING MODEL OF A CSR MESSAGE By Shu Yu Lin August, 2005 Chair: Spiro Kiousis Major Department: Journalism and Communications Little scholarship has examined the effectiveness of CSR campaigns in influencing attitude toward the company. The purpose of this present study is to measure the impact of CSR communication by proposing a cognitive processing model of a CSR message. According to the Elaboration Likelihood Model (ELM), it is assumed in this study that recipients perceived CSR and attitudes toward a company can be influenced by CSR campaigns in terms of the effect of message sidedness (one-sided, refutational two-sided and nonrefutational two-sided) and message recipients involvement level (high or low). A mixed 2x3 factorial design is applied for this experiment in which involvement and message sidedness are independent variables, while attitude toward the company and perceived CSR are the dependent variables. Two hundred and five students from the University of Florida participate in this study and randomly assigned to one of the seven conditions. A CSR campaign advocating the corporate social performance of a factitious xi

PAGE 12

lumber company named CALCO is created and presented on the printed Web pages of CALCO. A controversy over CALCOs practice of clearcutting, which is fundamental to the survival of the timber industry, is discussed in the stimuli. Results of the present study show that the one-sided CSR message is significantly more persuasive than the refutational two-sided and the nonrefutational two-sided one. Two main effects are found between message sidedness on both perceived CSR and attitude toward the company. These results indicate that the one-sided CSR campaign has a significant advantage over the refutational two-sided and the nonrefutational two-sided one in influencing participants perceptions about the companys social performance and their attitudes toward the company. In addition, the path analysis indicates a significant direct effect between perceived CSR and attitude toward the company. The findings of this study show that the perceived CSR is highly associated with the attitude toward the company. In other words, when a companys social performance is highly perceived, it is likely that people will generate more positive attitudes toward the company. These results demonstrate the value of CSR communication as part of a business strategy. When a company attempts to communicate its corporate social performance, the results of this study suggest that the one-sided CSR campaign simply presenting supportive information related to the companys corporate social activities has more persuasive effects than the two-sided CSR campaigns in influencing public perceptions about the companys social commitment and generating more positive attitude toward the company. xii

PAGE 13

CHAPTER 1 INTRODUCTION Corporate Social Responsibility Communication Understanding the effects of communicating corporate social responsibility (CSR) on public perceptions and attitudes toward the company can create a good opportunity for public relations practitioners to demonstrate the importance of PR function in the corporate world. Communication research has looked specifically into how best to communicate the social responsibility of an organization (Wartick & Cochran, 1985; Wartick & Rude, 1986; Heath & Ryan, 1989; Carroll, 1991; Wood, 1991; Frederick, 1994; LEtant, 1994; Esrock & Leichy, 1998; Heath, 1998; Clark, 2000). However, little of previous research on public relations has empirically tested the effects of CSR campaigns on peoples cognitive responses and attitudes. The concept of CSR is not new. It has been widely discussed both in the business and marketing field. More and more CEOs of Corporate America started to talk about a business obligation to society that goes beyond making profits. In this emerging corporate philosophy, contemporary corporations are expected to be both doing good in terms of social performance and doing well in terms of business performance. With an increasing emphasis on corporate social performance, thus, todays business organizations tend to advocate CSR not only to gain measurable outcomes but also to gain intangible outcomes (Lim, 2001). Corporate executives now believe that socially responsible companies will outperform their peers by enhancing reputations, reducing risks, and seizing new opportunities. An emerging agenda for todays CEOs is to bring 1

PAGE 14

2 the economics together with the environmental and the social needs as part of a business strategy. There haven been a lot of discussions from marketing literature regarding the effects of CSR initiatives on achieving organizational goals such as increasing the organizational bottom-line or boosting shareholder value. However, any attempt to account for CSR as part of a business strategy needs to address a communication management perspective of corporate social responsibility. In this regard, an important research topic for public relations researchers is how to communicate CSR effectively. To achieve excellence in public relations campaign, a good communication strategy should engage a good defense as well as offense. Then, presenting only a CSR message should be considered to be a defense whereas managing an issue with the CSR message could be an offense. In other words, should a company present simply CSR messages or utilize the CSR campaign as an opportunity to address a relevant environmental or a social issue? This thesis attempted to investigate the best model for the CSR campaign in terms of issues management. CSR Communication from the Issues Management Perspectives The power of information technology allows todays public to scrutinize corporate performances in detail. Through the Internet, individuals are able to access all kinds of information from a variety of sources and perspectives. The convenient accessibilities of mass information of corporate performances allow people to actively discuss social issues or environmental concerns related to corporate operations around the world. Irresponsible corporate actions are likely to induce criticisms from external stakeholders such as the media, governmental authorities, communities or even activists groups that organize

PAGE 15

3 boycotts against the company. For example, Shell placed its 100 years of brand building and reputation in jeopardy when the company was accused of the environmental damages and the neglect of human rights in Nigeria in 1994 (Brand Strategy, 2002). Due to this changing business environment, more corporations have woken up to the environmental and social expectations of todays society. For instances, to reduce emissions created by daily business operations, FedEx converted all its trucks to hybrid electric-diesel engines while UPS also included 1, 800 alternative-fuel vehicles. To ensure biodiversity, Starbucks is buying more organic and shade-grown coffee that minimizes damages of rain forests (Fortune, 2003). In academia, a sustainable body of literature has discussed CSR from the issues management perspective (Wartick & Cochran, 1985; Wartick & Rude, 1986; Heath & Ryan, 1989; Carroll, 1991; LEtant, 1994; Esrock & Leichty, 1998; Heath, 1998; Clark, 2000;). According to Lim and Lin (2005), some researchers have specifically looked into how CSR could play an active role in dealing with issues that are critical to organizational survival (Heath & Ryan, 1989; Frederick, 1994; LEtant, 1994; Esrock & Leichty, 1998; Heath, 1998). An abundant number of studies have conceptualized CSR in terms of issues management. However, little research has empirically delved into the communication management perspective of a CSR campaign. Moreover, a few studies that could be identified as research on communication management have used the concept of CSR very narrowly as either cause-related marketing or advocacy advertising (Lim & Lin, 2005, for a full list of literature).

PAGE 16

4 Statement of the Purpose The purpose of the present study is to fill this research gap by proposing an effective CSR communication strategy from an issues management perspective. This study intends to contribute to a better understanding of theoretical and practical implications of CSR communication. In this study, two communication goals are included in the CSR campaigns. First of all, a CSR message aims to inform targeted audiences about the corporate good citizenship. Secondly, in response to external stakeholders claims, a CSR message also addresses corporate positions or statements of controversial issues that are important to the companys fundamental practices. It is proposed in this study that communicating corporate social performance will eventually influence peoples attitude toward the company. Moreover, it is assumed in this study that the persuasiveness of CSR campaigns depends on the effect of message sidedness. Some companies may choose to simply present positive messages that support their social and environmental initiatives, while others may mention additional counterarguments from environmental groups. Background In this study, the topic of the experimental stimuli is the clearcutting issue that is critical to the survival of the timber industry. Clearcutting is a management technique in which all of the trees in an area are cut at the same time. Clearcutting is commonly applied by many lumber companies because it is cost effective and efficient. Bliss (2000) has addressed a problem of clearcutting issue. He argued that there had been a perceptual gap between the public view and the industrys view. He emphasized that corporate communication should play an important role in restoring public trust on the clrearcutting issue. Especially, it was emphasized the social and

PAGE 17

5 environmental performance of the timber industry so that it influences the dominant beliefs and values among the public. It is proposed in this present study that CSR campaigns might help to reduce the perceptual gap of clearcutting issues between the lumber company and the general public. Research Questions In this study, a dilemma in communication strategy is discussed. In dealing with controversial issues, should companies refute the counterarguments or not? This study examines the effectiveness of communication strategies in influencing individuals perceptions of the companys social performance and their attitude toward the company. From the persuasion perspectives, should the companies present one-sided CSR campaigns that are simply describing the social performance of the company? Or, should corporations design refutational two-sided CSR campaigns that include opposite claims from activists groups but follow with refutational arguments emphasizing the advocated positions? Moreover, companies could also consider the nonrefutational two-sided CSR campaigns presenting both supportive and opposite arguments without refutational arguments. In order to address these questions, an experimental study of the CSR campaigns was conducted to test empirically which strategy is more effective in influencing peoples attitudes and perceptions Theoretical Framework Richard Petty and John Cacioppos (1981, 1986a, 1986b) Elaboration Likelihood Model (ELM) is an attitude formation and change model that can be used to explain how publics evaluate persuasive communication. Based on the ELM, two previously used

PAGE 18

6 variables were employed in this study: involvement (high and low) and message sidedness (one-sided, refutational two-sided and nonrefutational two-sided). A cognitive processing model (see Figure 1-1) of CSR messages is proposed in this study. In this model, the researcher assumes that CSR messages will influence the audiences perception of the companys social performance and attitude toward the company depending on the message sidedness effect (one-sided, refutational two-sided and nonrefutational two-sided) and the audiences involvement level (high and low) The level of subjects involvement is manipulated in the booklet, directing them to either browse through the presented message or to carefully read and think about the message. Consistent with previous research, it is assumed in this study that those who are highly involved will induce more cognitive elaboration. The hypotheses of the current research were posited to address the main effects and the interaction effects of message sidedness and involvement on perceived CSR and attitudes toward the company advocated. The main effects of both involvement and message sidedness are that high involvement and two-sided messages will induce more cognitive elaboration than low involvement and one-sided messages. The interaction effects predict that when highly involved people are exposed to refutational two-sided messages, they will be more influenced by the message and create more positive attitudes toward the company. On the other hand, when involvement is low, people will be more influenced by a one-sided message in their evaluation of perceived CSR and attitude toward a company. Nonrefutational two-sided messages are said to have the least persuasiveness in terms of influencing attitudes toward the company and perceived CSR in both high and low involvement conditions.

PAGE 19

7 Based on the ELM, this study empirically tests the persuasive effects of CSR messages on peoples perceptions of corporate social performance and attitude toward the company. This study intends to contribute to a better understanding of the theory and the implications of CSR communication. In essence, the studies of ELM and CSR can be strengthened and supported by connecting the persuasive effects of communicating CSR on publics attitude. Perceived CSR Message Sidedness Cognitive Elaboration Issue Involvement Attitude toward the Company Figure 1-1. The cognitive processing model of a CSR message proposed in this study.

PAGE 20

CHAPTER 2 LITERATURE REVIEW Corporate Social Responsibility Corporate social responsibility (CSR) has been widely discussed in terms of a managerial function in business organizations. CSR has long been referred to the managerial obligation to improve both the societal welfare and a companys enlightened self-interest (Davis & Blomsstrom, 1975). A study of how U.S. corporate entities make use of the Web to present themselves as socially responsible citizens showed that more than 80% of a random sample from Fortune 500 companies had Web pages that addressed at least one corporate social responsibility issue (Esrock & Leichty 1998). This trend has reflected the pervasive belief among business leaders that CSR is an economic imperative in todays global marketplace. CSR is important for the public relations function in organizations. James Grunig and Todd Hunt (1984), two well-known communication scholars, devoted a whole chapter in their book to public relations and public responsibility. They argued that "Public, or social responsibility has become a major reason for an organization to have a public relations function and that the two-way symmetrical communication is the best means by which to evaluate social responsibility. (p. 48) Through CSR programs, public relations not only gets involved in the advisory management role by building relationships with key stakeholders but also engages in communicational management by delivering messages to target publics (LEtang, 1994). 8

PAGE 21

9 The Origin of CSR The idea of CSR appeared around the turn of the 20th century (Post, Frederick, Lawrence and Weber, 1996).Clarks (2000) comprehensive review of CSR and corporate public relations found similarities of these two areas in the origins, theories, processes, and primary responsibilities. Both public relations and CSR struggled to define their disciplines beginning primarily in the 1920s (Clark, 2000, p.366). Although a large number of todays business still think of CSR as paternalistic in terms of charitable contributions, researchers argue that there is another form of CSR from the stewardship perspective (Post, Frederick, Lawrence and Weber, 1996). According to the stewardship principle, corporations become stewards or public trustees by using their resources to affect all people in society in fundamental ways. This principle led to modern stakeholder theory that argued that managers should recognize the need to interact meaningfully with all groups who have a stake in the organization's activities. In the 1960s and 1970s, CSR underwent some of its most important iterations. Wood (1991) pointed out the emerging importance of CSR in her conceptual development of corporate social performance. "It became apparent during this time, particularly through social activism and regulatory activity, that social expectations of business had outstripped managers' comprehension and capabilities." (p.383) Thus, scholars and managers began to define what corporate social responsibilities were and were not. In the 1970s, studies about CSR were based on the idea that business is an "actor in the environment and should respond to social pressures and demands."

PAGE 22

10 (Wood, 1991, p.384) Another classic definition also pointed out the broad responsibilities of a modern corporation which defined CSR as the firms consideration of and respond to issues beyond narrow economic, technical, and legal requirements of the firm. (Davis, 1973) By the early 1980s, there is a slight shift of the research in CSR from the conceptual idea that companies should be responsible for a more practical approach to addressing how companies could respond to business-related social issues (Clark, 2000). CSR Definition Since the inception of CSR research, the concepts have been plagued by ambiguity. A long and diverse history has been associated with the evolution and the definition of CSR. According to Woods (1991) conceptual development of CSR theory, she included three main principles: (a) the principle of legitimacy at the institutional level, (b) the principle of public responsibility at the organizational level, and (c) the principle of managerial discretion at the individual level. In particular, Wood (1991) articulated that businesses are not responsible for solving all social problems. However, they are responsible for solving problems that they have caused and for helping to solve other social problems related to their business operations (p.697). For example, a lumber company is rightly held responsible for keeping a forest sustainable. However, it might be harder for such a company to support other social causes that are irrelevant to their practices, such as AIDS prevention or adult literacy. Social responsibilities should be relevant to the firms interests, operations, and actions (Wood, 1991, p.698). But this principle leaves room for managerial discretion in

PAGE 23

11 determining what social problems and issues are relevant and how those issues should be addressed To sharpen the definition of CSR, Carroll (1991) proposed a pyramid that constituted total CSR from four dimensions of social responsibilities including economic, legal, ethical, and philanthropic. It is worth noting that Carrolls (1991) pyramid of CSR begins with the economic performance. According to this principle, if a company is not making profits and is not providing high quality of goods and services to meet consumers needs, it can not be considered socially responsible even when the company has devoted many efforts in social causes. Carroll (1991) pointed out that CSR includes philanthropic contributions but is not limited to them. In other words, philanthropy is highly desired and prized but actually less important than the other three categories of social responsibility. Carroll argued that corporate managers and public relations practitioners should be aware of the distinctions between philanthropy and corporate responsibility. More specifically, philanthropic responsibilities do not dominate the definitional construct of CSR. While developing CSR initiatives, corporate managers should pay more attention to other dimensions of CSR rather than only focusing on philanthropy. Perceived CSR and Public Relations From the public relations perspective, perceived CSR may be considered a more appropriate term in discussing the strategy of communicating corporate social performance as a function of issues management. This is because the public perception of corporations social performance is not always predictable. In other words, the same message regarding a companys CSR activities can be perceived differently by different

PAGE 24

12 publics. As previous research on persuasion has pointed out, some moderators such as message recipients involvement or prior attitude toward the issue or the company can influence a recipients perceptions about the companys social performance. Therefore, more research regarding effective communication methods is needed. Despite the importance of communication management in CSR, effective communication methods are largely absent from the corporate social responsibility literature (Clark, 2000). A few studies about the effectiveness of CSR campaigns come from marketing and consumer research (Varadarajan & Menon, 1988; Menon & Kahn, 2003; Szykman et al., 2004). Moreover, previous discussions about CSR messages have merely considered perceived CSR a function of cause-related marketing (Varadarajan & Menon, 1988) or advocacy advertising of social issues (Haley, 1996; Menon & Kahn, 2003; Szykman et al., 2004). CSR and Issues Management Issues management is a function dedicated to helping organizations understand and strategically adapt to their environment through issue scanning, tracking, and monitoring (Heath, 1998). Heath (1998) further pointed out that a rhetorical approach to issues management assumes that organizations sometimes do not keep up with key publics standardstheir expectations of one or more aspects of the organizations performance (p.275). Along this research line, the function of issues management is considered parallel to the function public relations in organizations. CSR or corporate social performance (CSP) in literature has often been discussed as issues management (Wartick & Cochran, 1985; Wartick & Rude, 1986; Heath, 1998;

PAGE 25

13 Wood, 1991; Clark, 2002). Carroll (1991), who contributed to the definition of CSR, also took a perspective of stakeholder management by which CSR was considered as a managerial function that helped corporate management to achieve organizational goals. Wartick and Cochran (1985) proposed issues management as the third facet of the corporate social performance (CSP) model which stated that CSP is the integration of principles of social responsibility, processes of social responsiveness, and policies developed to address social issues. (p.758) CSR is considered as a process of corporate social responses to external stakeholders (Wood, 1991; Caroll, 1991). In this study, the CSR campaign was designed based on a lumber company. For the lumber industry, the long-term management of environmental responsibility has been a major issue. In particular, issues related to the practice of clearcutting have long been debated publicly (Berger, 2003). From the issues management perspective, Berger (2003) suggested that a lumber company deal with long-term stewardship and ecological considerations as major driving forces that are very important to the public. Berger (2003) further argued that a lumber company not only needs to understand public perceptions about clearcutting but also address that issue within a context of the structure and functioning of the ecosystem. Therefore, the clearcutting issue is considered an particularly appropriate topic for lumber companies to address and discuss when developing CSR campaigns. Corporate Credibility CSR can be adopted by companies as part of business strategies in order to enhance their corporate image and credibility. The reputation of a company that produces products has been identified as a type of source credibility in marketing research (Goldberg and

PAGE 26

14 Hatiwick, 1990). In addition to having a credible endorser represent their brand, companies are also concerned with their corporate credibility. This is evident because of the widespread use of public relations campaigns and prevalence of institutional and corporate advertising (Fombrun, 1996). Companies use this type of promotional effort primarily to associate themselves with positive environmental and social issues to enhance their reputation, and also hope of increasing sales (Kolter and Armstrong, 1996). A few empirical studies in the field of advertising effectiveness have examined the impact corporate credibility has on three principal outcome variables: attitude toward the ad, attitude toward the brand, and purchase intentions. Lafferty and Goldsmith (1999) found that both endorser and corporate credibility influence attitude-toward-the-ad and attitude-toward-the-brand, but corporate credibility alone appears to have a significant influence on purchase intentions. Moreover, whereas endorser credibility seems to have a greater influence on attitude-toward-the-ad, corporate credibility seems to have a greater influence on attitude-toward-the-brand. In other words, it is suggested that corporate credibility can play an influential role on forming peoples attitude toward the brand. However, little empirical research is available on corporate credibility. Although less attention has been given to this concept in the marketing literature, there is reason to believe that high corporate credibility is also important in producing positive attitude change toward the ad and toward the brand, as well as influencing purchase intentions (Fombrun, 1996). Goldberg and Hartwick (1990), Newell (1993), and Fombrun (1996) indicated that subjects are influenced by the credibility of the company when formulating their attitude-toward-the-ad and toward the brand as well as purchase intentions. The effect size show

PAGE 27

15 that the endorsers credibility appeared to have a stronger impact on the subjects when they evaluated the advertisement, and the corporate credibility appeared to have a stronger impact when the subjects assessed the brand. In the case of high corporate credibility, when the brand attributes are missing, the reputation of the firm may give the consumers more confidence that the product is good and make them significantly more willing to purchase the brand. Research findings are quite consistent with previous research. Newells (1993) results indicated that perceived corporate credibility is positively associated with purchase intentions, and Davis (1994) found that an overwhelming majority of consumers have stated that their product purchase decisions are at least in part influenced by their view of the parent companys good citizenship. Previous research in marketing and advertising has indicated that corporate credibility and corporate reputation can be considered a function of branding that influences consumers attitudes toward the brand and purchase intentions. Some studies in marketing research also showed positive associations between corporate credibility and corporate philanthropy. According to previous research, it is assumed that corporate credibility may be enhanced by the companys corporate social performance. Therefore, peoples attitude toward the company may also be positively influenced though effective CSR campaigns. The Elaboration Likelihood Model Persuasion and attitude change have long been the focal point of research with the Elaboration Likelihood Model (ELM) (Petty and Cacioppo, 1986a). The ELM provides

PAGE 28

16 a fairly general framework for organizing, categorizing, and understanding the basic processes underlying the effectiveness of persuasive communications (p.125). The standard paradigm of ELM research has been to test it with two to three variables, such as involvement, source credibility or argument strength. Typically, involvement is used as a motivational indicator of the route a person will use to process a message (central or peripheral route). According to the ELM, a highly involved person scrutinizes the arguments presented in the message and bases his or her attitudes toward the message and the message sponsors on the strengths of the arguments. In contrast, a person who is uninvolved when processing the message will use simple heuristics cues to form his or her attitudes toward the message (Petty and Cacioppo, 1986a). Past research has shown that the ELM is effective in explaining attitude change under various conditions and with certain accuracy. In this study, the researcher adopted two variables: involvement (high and low) and message sidedness (one-sided, refutational two-sided and nonrefutational two-sided). A review of previous studies based on the ELM will explain how each variable works and interacts in developing a persuasive message. The Theoretical Framework of the ELM The ELM is based on the conception that receivers will vary in the degrees to which they are likely to engage in the elaboration of information relevant to the persuasive issue. The term elaboration in the ELM refers to the extent to which a person thinks about issue-relevant arguments contained in a message (Petty & Cacioppo, 1986b, p.128). Such elaboration may be viewed as a continuum of commitment of cognitive resources which ranges from no thought to complete elaboration of every

PAGE 29

17 argument and complete integration of these elaborations into a persons attitude schema (p.129). Central Route and Peripheral Route According to the ELM (Petty & Cacioppo, 1981), the central route is theorized to occur under conditions where the message receiver is both motivated (e.g. due to the personal relevance of the issue) and has ability to process message content. When a message is processed through the central route, persuasion likely resulted from the information presented in support of an advocacy (Petty & Cacioppo, 1986b, p.125), and a high level of elaboration and cognitive responses will be fostered and ultimately lead to a significant impact on the message receivers attitudes to the communicated issue. In contrast, unmotivated or cognitively unable individuals follow a peripheral route to persuasion where elaboration is relatively low. The message receiver is theorized not to focus on the primary arguments of the message presented, but instead to focus on various peripheral cues, such as the communicators credibility or attractiveness as guides to attitude and beliefs. While both routes are persuasive, the literature on the ELM suggests that the central route produces more enduring results (Petty and Cacioppo,1986a, 1986b). The ELM was applied in the current study, because it is comprehensive in outlining the multiple roles by which the variables selected (message sidedness and involvement) might impact persuasion. The central route to persuasion in this study was the focus of personal relevance to environmental issues the clearcutting issue. If message receivers consider environmental issues highly personally relevant, according to the ELM, they will pay more attention and scrutinize the messages presented. In other words, it is

PAGE 30

18 predicted that message receivers who consider the environmental issues highly relevant will engage in the central route in which they give thoughtful consideration to issue-relevant information presented in this study. On the other hand, it is predicted that message receivers will not pay attention to the issues discussed in the message if they have low involvement with environmental issues. In that case, message receivers are likely to engage in the peripheral route and pay little attention the message. The Variable of Personal Relevance/Involvement In the ELM, the most important variable affecting the motivation to process a persuasive message is the personal relevance of the advocacy (Petty & Cacioppo, 1986b, p81). Petty and Cacioppo regard personal relevance as the extent to which the advocacy has personal meaning (Sherif et al., 1973). Petty and Cacioppo (1979) have termed their use of involvement as issue involvement, which is the extent to which the attitudinal issue under consideration is of personal importance (p.1915). In the current study, the researcher applied Petty and Cacioppo s (1981) definition of involvement as the amount of cognitive processing and/or interest the person has in the stimulus or attitude object being attended to. The ELM argues that when involvement is high, people are likely to use more cognitive processing and induce more message related thoughts, thus engaging the central route. Many researchers found consistent expectations when a given issue becomes increasing personally relevant to the message receivers, their motivations of engaging in thoughtful considerations of that issue presumably increase (Sherif & Hovland, 1961; Petty & Cacioppo, 1979, 1981, 1984; Petty, Cacioppo, & Goldman, 1981; Petty, Cacioppo, & Schumann, 1983).

PAGE 31

19 In contrast, when the level of cognitive elaboration is low, the message receiver is said to be in the low involvement condition. Then, the peripheral route of persuasion is more likely to change the receivers attitudes. The peripheral route is often used when the material presented is not as personally relevant, and the individual is not motivated to analyze the facts presented (Petty, Gleicher, & Jarvis, 1993). Involvement and Environmental Advocacy Involvement has been widely studied in the ELM by marketing and advertising research. Advertisers strive to produce messages that will involve people, prompting audiences to devote more attention to and engage in more elaboration of the information contained in the ads (Gordon, McKeage, & Fox, 1998). Advertising literature has observed that the peripheral cues, such as source credibility or the number of arguments are more effective for low-involvement rather than high-involvement products (Rhine & Severence, 1970; Petty, Ostrom, and Brock, 1981; Heesacker et al., 1983; Gotlieb & Sarel, 1991; Mazursky & Schul, 1992). When the advertising messages are related to environmental issues, Schuhwerk and Letkoff-Hagius (1995) found that issues involvement played a key role in the effectiveness of green advertising appeals. Those who were highly involved with environment issues were not swayed by green appeals by advertisers. On the contrary, people who were less involved with environment issues found the appeal of green advertising significantly more persuasive than non-green appeal. Several consumer-based studies have investigated psychological aspects of environmental concerns and environmental behaviors. Previous study found a positive correlation between individuals environmental attitudes and their intentions of

PAGE 32

20 purchasing ecologically packaged products (Schwepker & Cornwell, 1991). Ellen, Wiener, and Cobb-Walgren (1991) found that consumer efficacy, or the degree to which individuals think they can make a difference in the quality of the environment, was positively related to intent to purchase environmentally safe products. Message Sidedness Effect Definitions Persuasion research has long questioned whether the persuasive communication is more effective when it only presents the persuaders arguments or when it includes some discussion (with or without refutation) of opposing arguments (Lumsdaine & Janis, 1953). Two approaches of a one-sided message and a two-sided message are often contrasted and measured by researchers to examine various effects on attitudes toward an issue or a brand (OKeefe, 1999). It is defined that a one-sided message presents only arguments favoring the position advocated by the source while a two-sided message, on the other hand, presents both arguments opposing the sources position and those favoring it (Hovland, Lumsdaine, & Sheffield, 1949). This basic contrast ignoring or discussing opposite arguments has commonly been discussed in persuasion research as the difference between one-sided and two-sided messages. Some researchers even break down the two-sided message into the refutational two-sided message and the nonrefutational two-sided message depending on whether a subsequent refutation is included in the message or not. It is indicated that the effects of message sidedness are moderated by the type of two-sided message employed (Allen, 1991; Allen, et al., 1990). Literature has suggested that the message sidedness effect may vary significantly depending upon whether the opposing arguments are refuted (Allen,

PAGE 33

21 1991, 1994; Crowley & Hoyer, 1994). A nonrefutational two-sided message acknowledges opposing arguments but does not include direct refutation of the opposing arguments (e.g. Bettinghaus & Basehart, 1969). A refutational two-sided message, on the other hand, attempts to directly refute the opposing arguments (e.g., McCroskey, Young, & Scott, 1972). Origins In persuasion research, the different persuasive effects that resulted from types of messages is called message sidedness effect, which has a long research history that can be traced back to earlier persuasive research by Hovland, Lumsdaine, and Sheffield (1949). In their groundbreaking study, an issue concerning the military propaganda effect was discussed. Two treatment groups were presented either with the one-sided message that supported the communicators position or the two-sided message which contained additional opposite arguments. Their results did not find any significant differences between the persuasive effect of the one-sided message and the two-sided message. However, when participants have initial opposition to the communicators view, it was found that the two-sided message was more effective in producing opinion change in the desired direction than was the one-sided presentation. The one-sided presentation proved to be more effective only with those audiences who were already favorably disposed toward the communicators position. Moreover, their study also showed that a two-sided message was more effective with better educated people, while a one-sided message was more persuasive for less educated people (Hovland et al., 1953; Hovland et al., 1949).

PAGE 34

22 Previous Studies Previous discussions of message sidedness effect have different conclusions. Some research argued that two-sided messages appear more likely than one-sided messages to produce persuasive effects in the desired direction. (e.g., Hovland et al., 1953; Etgar & Goodwin, 1982; Albert & Golden, 1982; Golden & Alpert, 1987; Kamins & Assael, 1987; Kamins, Brand, Hoeke, & Moe, 1989; Pechmann, 1992; Bettinghaus & Cody, 1994; Crowley & Hoyer, 1994; Lang, Lee & Zwick, 1999; Stiff & Mongeau, 2003), while other studies (e.g., Earl & Pride, 1980; Belch, 1981; Hastak & Park, 1990) do not support the same findings. Thus, some other researchers (e.g., OKeefe, 1999; Stiff & Mongeau, 2003) urged caution in accepting research findings related to the message sidedness effect. Advertising and Non-Advertising Messages According to Allens (1991) meta-analytic review of the effects of one-sided and two-sided persuasive messages, the accumulated research shows that the refutational two-sided message appears to be more persuasive than both the one-sided message and the nonrefutational two-sided message (Allen, 1991; Allen et al., 1990). However, types of two-sided messages may have different persuasive effects according to various topics of messages that are presented. For non-advertising messages such as those on social or political issues, research found that refutational two-sided messages have a persuasive advantage over one-sided messages but there was no such advantage for nonrefutational two-sided messages (Allen, 1991, 1994; OKeefe, 1999). The message sidedness effect also differs according to receivers initial attitudes. Consistent with Hovland et al.s (1949) findings, a marketing study found that a two

PAGE 35

23 sided message can be an extremely efficient persuasive technique when consumers already hold negative attitudes toward a brand and when consumers are to be exposed to negative counterclaims by competitors or by a neutral third party, including news media (Sawyer, 1973; Golden & Alpert, 1987). Therefore some researchers suggest that adding negative information to positive messages can increase credibility with an audience (Golden & Alpert, 1987). It is asserted that acknowledging opposing arguments may boost the communicators credibility by suggesting his or her honesty and lack of bias and thereby increase the effectiveness of messages (e.g., Hovland, Lumsdaine, & Sheffield, 1949, p. 204; Settle & Golden, 1974; Pechmann, 1990). However, nonrefutational two-sided messages on a non-advertising topic do not produce the same enhancement of credibility (OKeefe, 1999). It is predicted that when communicating a social issue, receivers may feel confused when reading arguments from both supporting and opposing positions of view. Therefore, the persuasive effect is greatly decreased. Prior Knowledge, Motivation, Personality Previous studies have indicated that prior knowledge about the issue influences the effectiveness of each type of messages. More specifically, one-sided messages are more effective when the audience is uninformed about the issue than when the audience is knowledgeable. On the contrary, two-sided messages tend to persuade well-informed recipients more than individuals who are unfamiliar with the issue (Chu, 1967; Hovland, Lumsdaine, & Sheffield, 1949). It is asserted that individuals motivation to engage in processing messages influences the effectiveness of the types of sided messages. One-sided messages are more

PAGE 36

24 persuasive for people with lower motivations, while two-sided messages are more effective when audiences have higher motivation. Sorrentino et al. (1988) explained that recipients of the one-sided message may adopt the recommendations made by the communicator without much cognitive work while those of a two-sided message engage in active cognitive integration of conflicting arguments. Thus, a two-sided message is expected to be more effective when recipients have the motivation to process the arguments thoughtfully (Sorrentino et al., 1988). Crowley and Hoyer (1994) also assumed that a two-sided message would increase the recipients motivation to process the message, which also leads to more cognitive responses and relatively enduring attitude change. Previous studies have discussed message sidedness in the framework of ELM. Sorrentino et al.s (1988) findings on personality and effects of message sidedness indicated that under high personal relevance, uncertainty-oriented persons are more likely centrally or systematically process the information. When under low personal relevance, they will use peripheral or heuristic processing of information. However, contrary to current theorizing, personal relevance does not necessary increase recipients cognitive analyses of persuasive messages. It was found that the way people process information depends not only on the effects of personal relevance but also on the personality of the recipients. Issue Involvement Most previous studies about message sidedness effect are mainly from marketing or advertising research. In particular, consumer research has examined other effects of two-sided message on increasing source credibility perception (Kamins & Assael, 1987;

PAGE 37

25 Bohner et al., 2003), generating attitudinal resistance (Bither et al., 1971; Kamins & Assael, 1987; Bohner et al., 2003), or reducing counter arguing (Belch, 1981; Kamins & Assael, 1987; Swinyard, 1981). Other moderating variables such as the order of argument, the type of two-sided message (refutational or nonrefutational), the availability of counterargument, the initial agreement of recipients, the familiarity with the topic, or educational background (Allen et al., 1990; OKeefe, 1999) have also been examined by researchers. A substantial body of literature has been devoted to message sidedness effect (for a review see Allen et. al, 1990). However, little previous studies (Chebat & Picard, 1985; Sorrention et al., 1988; Hastak & Park, 1990) have examined the moderating role of receivers issue involvement. While a compelling case can be made for the type of messages as a moderator of sidedness effects, research does not provide insight as to why the messages differ in persuasiveness. One approach that can be taken to address this issue is to study message recipients cognitive responses of processing messages (Hale, et al., 1991). Three Theoretical Accounts for the Message Sidedness Crowley and Hoyer (1994) contend that inconsistency in results on the effects of message sidedness stems from the lack of a theoretical framework that specifically addresses two-sided message effects. In this section, the researcher reviewed three theories that addressed the reasons why two-sided messages have an advantage of persuasiveness over one-sided messages.

PAGE 38

26 Inoculation Theory According to the inoculation theory, the two-sided refutational message appears to work best to minimize counter argumentation (McGuire and Papageorgis, 1961). The authors suggested that refutation enabled participants to be pre-exposed to weakened counterarguments that heightened their defenses and alerted them to possible attacks on established beliefs. Inoculation theory also predicts that the refutational appeal should be the highest among the three appeals in the degree of support argumentation incurred (McGuire and Papageorgis, 1961). It is suggested by the inoculation theory that, refutational two-sided appeal leads to greater acceptance of the communicators position than the one-sided or two-sided nonrefutational appeal, even when the receiver is in agreement with the communicators position (McGuire, 1961; Etgar and Goodwin 1982). Inoculation should bolster the individuals cognitive defense by triggering the search for supporting arguments in light of potential counterarguments. In an advertising study, an inoculation of product disclaimers for relatively nonsalient product attributes comprised in a nonrefutational appeal could trigger a moderate to high degree of support argumentation. Because of the absence of inoculation and lack of perceived advertiser truthfulness, the one-sided appeal should produce the lowest degree of support argumentation (Kamins & Assael, 1987). Correspondence Theory From the perspective of communicators credibility, Tannenbaum (1967) argued that refutational two-sided appeals are effective communication devices because their credibility is enhanced by the opposing arguments presented to consumers. Applying

PAGE 39

27 correspondence theory to advertising study, researchers (Smith & Hunt, 1978) suggest that people perceive a lower degree of credibility and show higher levels of source derogation for one-sided appeals than for two-sided communications. In other words, consumers consider one-sided advertisements are not as credible as two-sided appeals. Kamins and Assaels (1987) empirical study on the sided appeals in advertising research found that two-sided refutational appeal resulted in significantly more support argumentation and significantly less counterargumententation than use of the one-sided appeal. Consistent with correspondence theory, the most novel appeal should lead to correspondent attribution and the lowest degree of source derogation upon exposure. The two-sided nonrefutational appeal should be perceived as most novel as no attempt is made to refute product disclaimers. In particular, research on two-sided nonrefutational appeals has been abundant in the marketing literature. Generally, two-sided nonrefutational appeals were more effective in increasing perceived advertiser truthfulness (Smith and Hunt, 1978; Swinyard, 1981) and believability (Etgar & Goodwin, 1982; Anderson & Golden, 1984; Golden & Alpert, 1987). Inoculation and correspondent theories contribute to predicting cognitive responses. The inoculation theory is predicting participants responses counter or supportive based on the communications stimuli. On the other hand, correspondence theory is concerned primarily with the attribution to the source enhancing or derogatory based on the same communication stimuli.

PAGE 40

28 Cognitive Response Theory Previous research has shown clearly that the message sidedness effect is weak, because the effect is indirect (Hale et al., 1991). Hale et al. (1991) asserted that cognitive responses and message evaluation should be considered as two mediating variables between the transmission of the message and the attitude measure (p.387). Numerous studies that have reasoned that message sidedness can be explained by the cognitive response theory (Allen et al., 1990; Hale et al., 1991; Crowley & Hoyer, 1994). Ford and Smith (1991) assume that such refutational messages may require more commitment and thus result in high elaboration. Sorrentino et al. (1988) explain that recipients of a one-sided message may adopt the recommendations made by the communicator without investing much cognitive work while those of two-sided messages engage in active cognitive integration of conflicting arguments. Thus, two-sided messages are expected to be more persuasive when receivers are more motivated to process the arguments thoughtfully (Sorrentino et al., 1988). Crowly and Hoyer (1994) also assumed that a two-sided message will increase receivers motivation to process the message, which also lead to more cognitive response and relatively enduring attitude change. According to Hale et al. (1991), cognitive responses and message evaluation play an order effect in processing persuasive messages. Their data appears that through exposure to a message prompts the generation of cognitions, which produces an evaluation of the message, and that evaluation influences a recipients attitude (p.387).

PAGE 41

29 It was also found that the number of positive cognitions generated is greater for refutational two-sided messages than for either nonrefutational two-sided message or one-sided messages. Hale et al. (1991) suggested that message sidedness can be considered part of the ELM. Their findings indicated that messages sidedness influenced message elaboration, which is the generation of message relevant cognition. One intuitive possibility is that direct refutation of an opposing argument increases the perceived strengths of the argument. However, in the nonrefutational two-sided message, receivers may have trouble comparing arguments from conflicting points of view they perceive them to be weaker. The one-sided message may simply be perceived as being weaker than the two-sided message because any mention of opposing arguments is missing. There is considerable evidence demonstrating that perceived argument strength is positively related to the quantity of positive cognitions (Petty & Cacioppo, 1986b). In the current research, it is predicted that the refutational two-sided message will be perceived to have the strongest argument strength, therefore, induce the highest number of positive cognitions. The ELM suggests that attitude change results from cognitive elaborations of audiences after receiving persuasion messages (Petty & Cacioppo, 1981, 1986a, 1986b). However, it is important to note that the ELM does not directly predict an attitudinal advantage for two-sided messages (Hastak & Park, 1990). All that this model asserts is that attitudinal advantage for a two-sided message will be mediated by message-related cognitive responses (p.330). However, message sidedness is connected to the ELM when

PAGE 42

30 motivation and ability to think about the issue will determine the route to persuasion (Allen, 1991, p.391). The current study of message sidedness is a response to the research call of Hale et al. (1991) that explores the possibility of message sidedness functions as a central persuasive cue. Hypotheses The Table 2-1 shows the factorial design of the experiment. As noted before, involvement is defined as the message recipients motivation to process information about an issue from a message (Petty and Caciopppo,1979; Johnson & Eagly, 1989) Participants in the high involvement condition will be more motivated to process the message than those in the low involvement condition. Hypothesis 1 predicted that participants in a high involvement condition (X1) would generate more message-related thoughts than those in a low involvement condition. (Predicted main effect) Hypothesis 2 predicted that a two-sided refutational message (Y1) would induce more message-related thoughts than a one-sided (Y3) or a nonrefutational two-sided message (Y2). (Predicted main effect) Few results on the interactive effects of message sidedness and perceived CSR are available. No theory-articulated reasoning is provided to fill the gap and a hypothesis is provided. Since it is hard to find any research or any theoretical accounts that explain the effect of message sidedness on perceived CSR, a research question is proposed. RQ1: How does message sidedness influence peoples perception about the corporate social responsibility of a company? Will a two-sided message or a one-sided message lead to more positive perceived CSR?

PAGE 43

31 According to Petty and Cacioppo (1979) under high involvement condition, participants engaged message processing, which enhances persuasion for favorable and strong argument. The ELM argues that when involvement is high, people are likely to use more cognitive processing and more message related thoughts, which engage the central route. Further, Hale, et al.s (1991) study has shown that when participants are exposed to refutational two-sided messages, they are likely to generate more positive cognitive cognitions than those exposed to one-sided messages. Following this line of research: Hypothesis 3 predicted that participants exposed to a refutational two-sided message would show more positive attitudes than those who were exposed to a one-sided or a non-refutational two-sided message. Hypothesis 4 predicted that participants in a low involvement condition would show more positive attitudes with a one-sided message while those in a high involvement condition would express more positive attitudes with a refutational two-sided message. (Predicted interaction effect) Hypothesis 5 predicted that higher perceived CSR would be positively associated with a positive attitude toward the company. Table 2-1. Factorial design of the experiment Message Sidedness (Y1, Y2 & Y3) Personal Relevance/Involvement (X1 & X 2) Refutational Two-Sided Nonrefutational Two-Sided One-Sided Main Effects for Involvement High Involvement A B C X1 > Low Involvement D E F X2 Main effects for Sidedness Y1 > Y2> Y3

PAGE 44

CHAPTER 3 METHODOLOGY The present study investigates empirically the persuasiveness of a CSR campaign in influencing message recipients attitudes and perceptions. Based on the level of issue involvement and the effects of message sidedness presented in CSR messages, this study also examines the main effect and the interaction effect between these two variables on message recipients attitude toward the company and their perceptions of the companys social achievements. In addressing the socially responsible achievements, the present study asked whether a company should discuss a relevant issue to communicate with key stakeholders or active publics. In other words, should a company present simply a CSR message or utilize the CSR campaign as an opportunity to address a relevant environmental or a social issue? To answer the above questions, a factitious lumber company named CALCO was created in the experiment. The reason for choosing a lumber company in experiment design is that timber industry has long been questioned for the practice of lumbering such as clearcutting that result in critical environmental damages. In addition, the continuing controversy over clearcutting has become essentially a social issue. Therefore, the present study is able to test whether peoples attitude toward the company can be positively enhanced after receiving a CSR campaign which advocates the social performance of that company. Moreover, this study further tests whether the enhancements of the positive attitude toward the company can be increased when the company particularly addresses 32

PAGE 45

33 controversial issues that are critical to the business. In brief, the purpose of this study is to test whether people will have more positive attitude if the company is considered as socially responsible. By applying a sophisticated experimental study on attitude that examines the cognitive processing model of a CSR message, the present study adds a body of knowledge to the research of CSR and the ELM. In particular, the present study advances the existing research and literature on the effect of message sidedness on CSR messages and environmental issues. Moreover, this study is the first to link between CSR communication and issues management and empirically examine the persuasive effects of the CSR messages on attitude toward the company and the perceptions of the companys social performance. This innovative approach is based on the growing trend of todays corporate world to bring the economics together with the environmental and the social needs as part of a business strategy. From an issues management perspective, the present study also presents a good opportunity for public relations practitioners to demonstrate the importance of PR function, pursuing an excellence in CSR communication campaigning. An Experimental Design This study is grounded on a cognitive response perspective in examining the underlying processes that mediates the persuasive effects of a CSR campaign. An experimental method was considered the most appropriate approach in testing a cognitive response model proposed in this study. A mixed 2x3 factorial experiment was designed with two between-subject variables being personal involvement (high or low) and message sidedness (one-sided, refutational two-sided and nonrefutational two-sided). Participants attitudes toward the company and

PAGE 46

34 their perceptions of the companys social performance were evaluated after reading the CSR campaigns presented in the stimuli. Involvement level and message sidedness have been chosen as two independent variables of testing the persuasiveness of a CSR campaign. According to the ELM (Petty and Cacioppo, 1986a, 1986b), issue involvement plays an important role in influencing which route people choose to process the message. The ELM has indicated that for the desired central route of attitude change to occur, participants must be highly involved with the issue and able to cognitively process the persuasive messages. Therefore, participants level of involvement with the issue discussed in the CSR messages was controlled in the experiment. Previous research has indicated that message sidedness could result in different persuasive effects on attitude change. Therefore, message sidedness was experimentally varied so that participants in this study received either a one-sided CSR campaign, a refutational two-sided CSR campaign, or a nonrefutational two-sided CSR campaign. All three CSR campaigns addressed a lumber companys social performance and the clearcutting issue which was associated with the companys forestry management. In this study, the one-sided CSR campaign only presented arguments that support the companys environmental commitment and corporate citizenship. The refutational two-sided CSR campaign comprised pairs of arguments including both supportive and refutational information. For example, arguments against the companys clearcutting practices were followed by refutational explanations that defended the companys goodwill. The nonrefutational two-sided CSR campaign acknowledged the concerns from the opposite positions such as environmental activists groups, but did not attempt to

PAGE 47

35 refute those claims directly. These procedures are similar to the two-sided operationalisation used in previous message sidedness studies (Allen, 1991). Consistent with previous research of the ELM, it is assumed in this study that the more a person is involved with the environmental issue, the more he or she thinks carefully about (or cognitively elaborates on) the CSR message, eventually the more likely positive attitudes are to occur. The results of the current study will contribute to the research of cognitive response theory and demonstrate the value of the effects of CSR campaigns on attitude toward the company and the function of CSR communication as issues management. Pilot Study Procedure In order to make sure the manipulation checks of message sidedness and involvement work successfully, a total of fifty-four college students participated in a pilot test. All students were recruited on the campus of the University of Florida and assigned to one of six treatment conditions randomly. The message stimulus was a CSR campaign presented on the Web site of a lumber company named CALCO. The purpose of the CSR campaign is to advocate the social performance and the corporate citizenship of that company. Participants were presented with a booklet including a background news story, a CSR campaign message, and a questionnaire. The fist part of the booklet was a one-page newspaper article about an environmental activist named Julie Butterfly Hills who lived in a tree for two years to protest against the lumber companys clearcutting practices. CALCOs CSR campaigns on the company Web site were the second part of the booklet, and finally a five-page questionnaire was included as the third part of the booklet.

PAGE 48

36 Participants were instructed with the following statement to complete the questionnaire once they have finished reading all the messages presented in the bookelt. Based on the background story and the CALCOs statement you just read from the printed Web pages, please answer the following questions. Please circle the number that best describes your thoughts or feelings. The questionnaire included Liker-type items designed to measure how interested respondents were in environmental and clearcutting related issues, their feelings toward the company presented, and their level of agreement with various attitudes statements. Participants level of involvement was manipulated by the location of the lumber company presented in the message stimulus. For high involvement groups, the lumber company presented in the stimuli was located in Florida. On the other hand, the lumber company was located in Taiwan for low involvement groups. Involvement level was measured using four questions (see Table3-1): (1) How important are environmental issues in Taiwan (or Florida) to you personally?; (2) How important is the deforestation issue in Taiwan (or Florida) to you personally?; (3) How much are you concerned about environmental issues in Taiwan (or Florida)?; (4) How much are you concerned about deforestation issue in Taiwan (or Florida)? Table 3-1. Liker-type items of the manipulation check of involvement level for high involvement groups How important are environmental issues in Florida to you personally? How important is the deforestation issue in Florida to you personally? How much are you concerned about environmental issues in Florida? How much are you concerned about the deforestation issue in Florida?

PAGE 49

37 For the manipulation check of message sidedness, we used two items from Trent and Greers (2001) items: (1) CALCOs statement shows two-sided arguments; (2) CALCOs statement refutes the opponents arguments (see Table 3-2). Table 3-2. Liker-type items for the manipulation check of message sidedness CALCOs statement shows two-sided arguments. CALCO refutes opponents arguments. Pilot Test Results From the pilot test, a successful manipulation of involvement was obtained (M low =3.94, SD=1.69; M high = 4.95 SD=1.19; t=2.54, df=50, p<.05). The mean score of involvement index from the high involvement group is significantly higher than those from the low involvement group. However, the results of the manipulation check of message sidedness did not yield significant differences in the pilot test. In other words, participants can not recognize the differences between three groups of CSR campaigns (one-sided CSR message, refutational two-sided CSR message, and nonrefutational two-sided CSR message). The manipulation of message sidedness failed in the pilot test. The major problem was attributed to the design of background information page about the clearcutting issue associated with CALCO's practices. In the pilot test, a one-page newspaper article about Julie Hills' tree sitting protest against CALCO's clearcutting practice for two years was provided. It was assumed that this new article might have delivered a too strong message that overrode the message sidedness effect presented in CSR campaigns. The results from the pilot test shown overall negative attitudes toward CALCO. Thus, to reveal the effect of message sidedness in CSR campaigns, CALCOs

PAGE 50

38 "About Us" page (see Appendix A) including a brief introduction of the company's history and visions was substituted for the news article in the main experiment. Main Study Sample A convenient sample was adopted in this study. It was considered appropriate to use a convenient sample in this study due to the theoretical testing nature of the experiment (Calder, Phillips, & Tybout, 1981). Calder, Phillips, & Tybout (1981) stated that it is important to use a homogeneous sample for an experiment because of its theoretical testing in nature. They argued that a homogeneous sample permits more exact theoretical predictions than from a heterogeneous group. In addition, a homogenous sample also decreases the chance of making a false conclusion about whether there is a covariation between the variables under study (Cook & Campbell, 1979). The participants for the main study were recruited from two colleges of the University of Florida. The experiment was conducted in two undergraduate classes from the College of Agricultural and Life Sciences and the College of Journalism and Communications. A total of 205 students (68 male, 113 females, 24 missing) participated in the study. Therefore, educational level and the age of participants appeared to be indicative of traditional college students. Participants completed the study at the beginning of the class session and received extra credit points as compensation for their participation. One hundred and four participants were in the 2 (involvement: high vs. low) by 3 (message sidedness: one-sided vs. refutational two-sided vs. nonrefutational two-sided) groups, and 21 were in the control group. Thus, all participants were randomly assigned to one of the following seven conditions: 1)high vs. one-sided message, 2)low vs. one-sided message, 3)high vs.

PAGE 51

39 nonrefutational two-sided message, 4)low vs. nonrefutational two-sided message, 5)high vs. refutational two-sided message, 6)low vs. refutational two-sided message or 7)a control group (see Table 3-3). Table 3-3. The conditions of the 2x3 experimental design ONE-SIDED NR TWO-SIDED R TWO-SIDED HIGH INVOLVEMENT Group (1) Group (3) Group (5) LOW INVOLVEMENT Group (2) Group (4) Group (6) Note. 1. NR stands for nonrefutational. R stands for refutational. Stimuli The CSR campaign stimuli used in the main study were slightly different from that of the pilot study. The newspaper article about Julie Hills' protest against CALCO's clearcutting practices was replaced by a brief introduction of the company's history and visions titled as "About Us" in the main experiment. Participants were asked to read three printed Web pages including a brief introduction about the company, the companys environmental commitment, and corporate social activities. The Web site of one of the largest lumber companies in the United States was mirrored and slightly changed for simulating Web pages of a fictitious company named CALCO. To manipulate different levels of involvement, two versions of CALCOs About Us page were created. For high involvement conditions, CALCO was presented as The Caribbean Lumber Company in Southwest Florida in the About Us page (See Appendix A). On the other hand, for low involvement conditions, CALCO was presented as The Chiayi Lumber Company in Taiwan on the same page (See Appendix B). To manipulate message sidedness (one-sided vs. refutational two-sided vs. nonrefutational two-sided), three versions of CALCOs Environmental Commitment

PAGE 52

40 Web pages were created. For the one-sided version, the CSR campaign only contained one page of supportive messages that advocated CALCOs corporate social activities in three areas: environment, community, and employees. Each domain of CSR activities was subtitled as Responsibility to Our Environment, Responsibility to Our Community, and Responsibility to Our Employees. Moreover, this page also has two versions for high involvement groups (See Appendix C) and for low involvement groups (See Appendix D) depending on the location of CALCO presented in the campaign. For the refutational two-sided version, one page of Myth and Fact (See Appendix E) which contained activist groups concerns of clearcutting issues and CALCOs refutational arguments was included in addition to the CSR messages the same as the one-sided group. Therefore, the refutational two-sided version has two pages in total. The refutational message dealt with three controversies related to the timber industrys clearcutting practice by displaying activist groups claims about the impact of clearcutting on aesthetic and recreational value, deforestation, and ecosystem. Each argument was refuted with CALCOs positions. For the nonrefutational two-sided version, in addition to the same CSR messages as the one-sided version, one page of activist groups claims about clearcutting issues was presented without any refutations or arguments from CALCOs perspectives (See Appendix F). Thus, the nonrefutational two-sided version also has two pages. To ensure the manipulation of both involvement level and message sidedness successfully instituted in different treatment conditions, a pilot test with a separate pool of 54 participants in total was conducted.

PAGE 53

41 Procedure Students were informed during that days class session that there would be an extra credit opportunity if they complete the study. Students were also informed that this opportunity would require them to read a couple of corporate Web pages, and answer several questions based on what they read. Participants were told that the purpose of the experiment was to develop creative headlines and body copies for a lumber companys Web site. Seven versions of booklets including six for the treatment groups and one for the control group were given to participants based on randomly assigned experimental conditions. The stimuli booklets included one printed Web page of CALCOs About Us, two printed Web pages of CALCOs CSR advocacy titled Environmental Commitment, and finally a five-page questionnaire. Two versions of questionnaires were provided including a U.S. version for high involvement groups (See Appendix G) and a Taiwanese version for low involvement groups (See Appendix H). After reading the stimuli, participants were asked to complete a questionnaire that contained a series of measures on attitudes, perceived CSR, and message-related thoughts. After participants responded anonymously to questions regarding the persuasive effects of CSR messages, they completed a measure for the relationship study and answered demographic questions. The results of the relationship questions will not be discussed in this thesis as they are part of another study that will be explained in a different report. Finally, participants were debriefed and thanked for their participation. Students were informed that their completion of the measures was voluntary and anonymous. In order to assure the anonymity of participants responses, each participant turned in a separate consent form with his or her name written on it.

PAGE 54

42 Operational Definitions: Independent Variables Involvement Involvement is a motivational variable that moderates how individuals process information. The manipulation of involvement differentiates the motivation people have while processing the information which in this study is about the companys CSR initiatives. In this study, involvement is defined in terms of issue involvement which can be explained by the extent to which the attitudinal issue under consideration is of personal importance (Petty & Cacioppo, 1979, p.1915). The manipulation of issue involvement in this study is also closely paralleled by the study of Petty, Cacioppo, and Schumann (1983). In Petty et al.s (1983) research, the high involvement group was led to believe that a comprehensive exam would be held at their university while the low involvement group was told that the exam would be conducted in a different university. Personal relevance occurs when people expect the issue to have significant consequences for their own lives (Apsler & Sears, 1968). Therefore, it is clear that in high involvement conditions, participants would be more motivated to engage in thoughtful consideration of the issue because they would be affected personally whereas in low involvement conditions, they would not. According to these findings, participants levels of involvement were manipulated by receiving different background information about CALCO which was presented either as a lumber company located in Florida or located in Taiwan. Take the high-involvement condition for example, participants read the following messages which were included on the first page of their stimuli package: CALCO is The Caribbean Lumber Company in Southwest Florida. We have been growing trees, making wood products, and supporting generations of families in Southwest Florida since 1863. On the other hand, for the low

PAGE 55

43 involvement condition, CALCO was presented as The Chiayi Lumber Company in Taiwan on the same page. Message sidedness Message sidedness was experimentally varied and participants received one of the three message packages: one-sided message, refutational two-sided or nonrefutational two-sided messages to discuss CALCOs CSR activities. All three packages include one page of background information about the company, and another Environmental Commitment describing CALCOs CSR activities which is to be found on the last page. For the refutational two-sided CSR campaign and the nonrefutational two-sided CSR campaign, an additional page on Environmental Commitment discussing the clearcutting issue was included as well. The About Us Web page contains the brief introduction about CALCOs name, products, history and visions. The following paragraphs are some examples of sentences: Established more than 140 years ago, CALCO still embodies the vision and spirit of its founders. Today, their vision is a reality. The modern CALCO is a dynamic and innovative company whose employees are dedicated to providing superior lumber products to customers and value to shareholders. For the refutational two-sided message, an additional page of Myth and Fact contained three controversies about the impact of clearcutting on (1) deforestation, (2) aesthetic and recreational value, and (3) ecosystem. The following paragraphs are examples of arguments related to the impact of clearcutting on deforestation: Myth: Environmental activist groups have argued that clearcutting leads to deforestation. They have also alleged that nothing can be regenerated once a forest is cut down and that it will never be the same again. Fact: Deforestation is the removal of a forest with no intention of establishing a future stand of trees. For instance, deforestation is occurring today in Brazil and

PAGE 56

44 other tropical areas in the form of agricultural conversion. Our silvicultural clearcutting is both a harvest and a regeneration of the forest, and is done to improve future stand quality, growth, genetics, and species composition. While regarding the nonrefutational two-sided message, environmental activists groups claims were included without refutational arguments from CALCOs perspective. The following sentences are provided as examples: Environmental activist groups have argued that clearcutting leads to deforestation. They have also alleged that nothing can be regenerated once a forest is cut down and it will never be the same again. Another argument against clearcutting is that the resulting bare patches of land look unsightly. Activist groups argue that clearcutting is the destruction of aesthetic values and recreational opportunity. In all the three message packages, one page of Environmental Commitment was included on the last page of the CSR campaign. Three domains of social activities are presented: (1) Responsibility to the Environment, (2) Responsibility to Our Community, and (3) Responsibility to Our Employees. The following sentences about Responsibility to the Environment are provided as examples: CALCO now offers office paper with 100 percent post-consumer recycled content. Tree-free papers made from kenaf, hemp, or agricultural waste such as wheat straw, rice straw and bagasse (waste from sugar cane processing) are also available. There is also an increasing range of forest-friendly products for use in construction and furniture making. The following sentences about Responsibility to Our Community are provided as examples: Sustaining the communities in which we operate is a responsibility and a commitment we take very seriously. We believe that we have an obligation to contribute to the quality of life in our communities. Our major contribution to the people in Sarasota County is in our ability to provide hundreds of family-wage jobs. As part of that commitment, we also provide a medical and dental facility for CALCO employees and people living in our communities. Each year, the CALCO

PAGE 57

45 College Scholarship program awards over $ 350,000 to students from Sarasota County. The following sentences related to Responsibility to Our Employees are provided as examples: CALCO fosters personal and professional growth and learning for all employees. Employees in CALCO are empowered to make decisions and to share both the risks and rewards of making decisions. In addition, CALCO employees are eligible to receive $ 12,000 for attendance at any four-year accredited university or college and $4,600 for any two-year college or trade or religious school. Operational Definitions: Dependent Variables Cognitive elaboration According to Petty and Cacioppo (1986a, 1986b), cognitive elaboration was defined as message-related thoughts. In order to evaluate participants level of cognitive elaboration when they were processing the presented CSR messages, participants were requested to list all the thoughts that came to their mind while reading the printed Web pages of the company. Participants read the following instructions on writing their message-related thoughts, Please list the thoughts that came to your mind while reading CALCOs statement generate as many as you can. Please write your thoughts in the following lines, one thought per line. Eight lines were provided and they were followed by the above instructions. Perceived CSR In this study, the construct of CSR as the audiences subjective perceptions is more likely to be formed by exposure to corporate CSR advocacy messages. Thus, the term of perceived CSR is used in this study instead of CSR. To access participants perceptions about the companys CSR activities and commitment, Menon and Kahns (2003)

PAGE 58

46 measurements of perceived CSR was applied in this study with a slight change in wording. Participants responses of CALCOs environmental policies were asked by the following five questions: (1) CALCO believes in environmental commitment; (2) CALCO is likely to follow environmental friendly rules and policies; (3) CALCO is highly involved in community activities; (4) CALCO is highly concerned about environmental issues; (5) CALCO is genuinely concerned about public welfare (see Table 3-4). Participants were told to show their agreement with above questions by 7-point rating scales anchoring as strongly disagree, and as strongly agree. The statistical results show that these items on the perceived CSR scale is very reliable (alpha = .92). Table 3-4. Liker-type items for perceived CSR CALCO believes in environmental commitment. CALCO is likely to follow environment-friendly rules and policies. CALCO is highly concerned about environmental issues. CALCO is highly involved in community activities. CALCO is genuinely concerned about public welfare. Attitudes A 7-point semantic differential scale with six items was used to evaluate participants attitudes toward CALCO. Priester and Pettys (2003) attitude scale is applied with a slight change in wording for evaluating participants attitudes toward the company in this study. The items of attitude scale included in this study were negative-positive, harmful-beneficial, foolish-wise, unfriendly-friendly, bad-good, and

PAGE 59

47 unfavorable-favorable (see Table 3-5). The following instructions were given for participants: Please evaluate how you feel about CALCO by circling a number on each of the scales below. If you feel that you have no reaction, please circle the number 4 to indicate your neutrality. The statistic results of the items applied in this study turned out to be very reliable (alpha = .95). Table 3-5. Semantic differential type items for attitude toward the company Negative 1 2 3 4 5 6 7 Positive Harmful 1 2 3 4 5 6 7 Beneficial Foolish 1 2 3 4 5 6 7 Wise Unfriendly 1 2 3 4 5 6 7 Friendly Bad 1 2 3 4 5 6 7 Good Unfavorable 1 2 3 4 5 6 7 Favorable Manipulation Check Involvement The manipulation check for involvement remained the same as in the pilot study. In this study, participants levels of involvement were controlled according to the different locations of the company. For the low involvement condition, CALCO was presented as a lumber company in Taiwan, whereas for the high involvement group CALCO was presented as a lumber company located in Florida. It was assumed that participants would feel more involved with the issue presented in the messages if the company was located in Florida. To evaluate the manipulation of involvement, four questions were asked to examine participants involvement level: (1) How important are environmental issues in Taiwan (or Florida) to you personally?; (2) How important is the deforestation issue in Taiwan (or Florida) to you personally?; (3) How much are you concerned with

PAGE 60

48 environmental issues in Taiwan (or Florida)?; (4) How much are you concerned with deforestation issue in Taiwan (or Florida)? (see Table 2). Message sidedness The manipulation check for message sidedness also remained the same as in the pilot study (see Table 3). For the manipulation check of message sidedness, we used two items from Trent and Greers (2001) with a slight change in wording: (1) CALCOs statement shows two-sided arguments (2) CALCO refutes opponents arguments. Participants show their agreement with 7-point rating scales anchoring as follows: as strongly disagree, and as strongly agree.

PAGE 61

CHAPTER 4 FINDINGS The analyses of this study contained four parts. First section is the descriptive statistics of the present participants in this experiment. Second part contains the manipulations checks of the main experiment and reliability of dependent measures. Third, the hypotheses and the research question are addressed, using t-test, One-Way ANOVA, post-hoc analysis, Pearson correlation and regression analyses (hierarchical regression). Finally, the path analysis was performed to measure the causal relationships between (1) the effects of message sidedness and involvement level on perceived CSR and (2) the impact of perceived CSR on attitude toward the company. Profile of Participants All respondents used for analysis in this experiment were college students at the University of Florida. A total number of 205 students participated in this study. Of the total 205 participants, 68 (33.2%) were males, 113 (55.1%) were females and 24 (11.7%) did not report. Of those who responded to the question of ethnicity, 114 (55.6 %) were White-Non Hispanic, 30 (14.6%) were Hispanic American, 16 ( 7.8%) were Asian American, 14 (6.8%) were African American, and 7 (3.1%) reported other. Participants age ranged from 18 to 40. However, over 90% of participants were in the 18-24 age groups. This study was conducted in two undergraduate classes. Therefore, educational level and the age of participants appeared to be of traditional college students. 49

PAGE 62

50 205 participants were randomly assigned to one of the seven groups. 21 participants were assigned in the control group, and the remaining 184 participants were assigned in the following treatment groups: 1) 31 in the high and one-sided message group, 2) 30 in the low and one-sided message group, 3) 31 in the high and nonrefutational two-sided message group, 4) 31 in the low and nonrefutational two-sided message group, 5) 30 in the high and refutational two-sided message group, 6) 31 in the low and refutational two-sided message group (see Table 4-1). Table 4-1. The number of participants in each cell Involvement vs. Message One-Sided NR Two-Sided R Two-Sided Control Total Low 30 (32.6%) 31 (33.7%) 31 (33.7%) 92 (100%) High 31 (33.7%) 31 (33.7%) 30 (32.6%) 92 (100%) Control 21 (100%) 21 (100%) Total 61 (29.8%) 62 (30.2%) 61 (29.8%) 21 (10.2%) 205 (100%) Note. 1. NR stands for nonrefutational. R stands for refutational. Results of Manipulation Checks Involvement Consistent with the results from the pilot test, the manipulation check of involvement worked successfully. The four indicators used to check the effectiveness of the involvement manipulation yielded converging and supportive results. Table 4-2 shows the mean scores for two groups that pertained to either a highor a low-involvement condition were significantly different (M low = 3.89, SD = 1.62; M high = 5.08, SD = 1.32; t = 5.39, df = 179, p < .001). In other words, the manipulation and the measurements of involvement level applied in this study which closely paralleled by the study of Petty, Cacioppo, and Schumann (1983) work successful.

PAGE 63

51 Table 4-2. The manipulation check of involvement Measure Treatment N Mean SD t df Low 90 3.89 1.62 Involvement High 91 5.08 1.32 5.39 *** 179 Note. Items in the involvement scale were measured on a 7-point scale ranging from 1 (not at all) to 7 (very much). *** p < .001 Message Sidedness For the manipulation check of message sidedness, the revised background story turned out to be a success. After the news article about clearcutting issue was replaced by a CALCOs About Us Web page, the manipulation check of message sidedness worked successfully. The original background story about Julie Butterfly Hills' tree-sitting protests might have delivered a strong message to the experiment group that possibly had overridden the message sidedness effect. Therefore, it was difficult to find significant differences between participants' attitude toward the company even after exposure to CSR campaigns advocating the company's good citizenship and its corporate social performance. Therefore, a brief introduction to the company's history was substituted for the news article. Message sidedness resulted in a significant mean difference between three treatment groups. As researchers expectation, the mean score of the message sidedness index for the one-sided message (M=3.44, SD=1.06) turned out to be the lowest among three sided message groups. Table 4-3 indicates that the mean score for a refutational (R) two-sided message was significantly higher than that of a nonrefutational (NR) two-sided message and that of a one-sided message.

PAGE 64

52 Table 4-3. The manipulation check of message sidedness Measure Treatment N Mean SD F df One-Sided 61 3.44 1.06 NR Two-Sided 62 3.90 1.03 Message Sidedness R Two-Sided 61 4.85 1.33 23.67 2/181 Note. 1. NR stands for nonrefutational. R stands for refutational. 2. Items in the message sidedness scale were measured on a 7-point scale ranging from 1 (strongly disagree) to 7 (strongly agree). Question Items used in the message sidedness were CALCOs statement refutes the opponents arguments, and CALCOs statement shows two-sided arguments. *** p < .001 Reliability Checks for Dependent Measures In order to establish the reliability of dependent measures used in this study, reliability analyses were conducted for each of construct used in this study including perceived corporate social responsibility (CSR) and attitude toward the company. Cronbachs alpha was computed to evaluate if the items within each index have high internal reliability. Alpha represents a coefficient that demonstrates how well the items measuring the same characteristic correlate with one another (Hon & J. E. Grunig, 1999). In general, reliability coefficients over .90 are considered excellent, over .80 are very good, and over .70 are considered adequate (Kline, 1998). The index of perceived CSR includes five questions relating to CALCOs environmental policies. Participants attitudes toward CALCO were evaluated by a 7-point semantic differential scale with six items. Reliability tests confirmed that the indexes of overall attitude toward the company (Cronbachs =.93) and perceived CSR ( = .90) applied in this research was appropriate. Therefore, the statistical analysis indicated that the dependent measures used in this experiment had a high internal reliability.

PAGE 65

53 Test of Hypotheses Cognitive Elaboration Hypothesis 1: Participants in a high involvement condition will generate more message-related thoughts than those in a low involvement condition. (Predicted main effect) Based on the literature review, it was assumed that message sidedness and involvement have main effects on cognitive elaboration. Contrary to the expectations, the main effects of involvement and message sidedness were not significant. Hypothesis 1, participants in a high-involvement condition would generate more message-related thoughts than those in a low-involvement condition, was rejected. A t-test was conducted to test the main effect of issue involvement level on cognitive elaboration. As Table 4-4 demonstrates, no statistical differences of message-related thoughts between a high and a low involvement group were found (p= .398). Thus, Hypothesis 1 was rejected. Contrary to previous research, this study did not find more cognitive elaboration among participants of the high-involvement condition. These results showed that regardless of whether the participant is highly involved or less involved with the issue, it is likely that he or she would generate the same amount of message-related thoughts about to the issue discussed in the messages. Table 4-4. A t-test of mean difference of message-related thoughts generated between the high and the low involvement group Measure Treatment N Mean SD t Df Low 91 3.42 2.21 Involvement High 92 3.36 1.99 .190 181 Note. 1. Participants were requested to generate message-related thoughts after reading experimental stimuli. 2. p= .398

PAGE 66

54 Hypothesis 2: A two-sided refutational message will induce more message-related thoughts than a one-sided or a nonrefutational two-sided message. (Predicted main effect) To test Hypothesis 2, a One-Way Analysis of Variance (ANOVA) test was run (see Table 4-5). In this experiment, no statistically significant differences were found between the numbers of message-related thoughts from different message sidedness groups (p=.134). Thus, Hypothesis 2 was also rejected. For cognitive elaboration, there were no significant effects attributed to involvement and message sidedness. Table 4-5. A t-test of mean difference of message-related thoughts generated between three groups of sided messages Measure Treatment N Mean SD F df One-Sided 61 3.70 2.22 NR Two-Sided 62 2.97 2.10 Attitudes R Two-Sided 61 3.49 1.91 2.03 2/180 Note. 1. NR stands for nonrefutational. R stands for refutational. 2. Items in the attitude scale were measured in a seven-point semantic differential scale. 3. p= .134 Main Effects of Message Sidedness on Perceived CSR RQ1: How does message sidedness influence peoples perception about corporate social responsibility of a company? What will lead to more positive perceived CSR between a two-sided message and a one-sided message? None of the previous studies from the literature of CSR and persuasion research had empirically tested the persuasiveness of a CSR campaign based on the effect of message sidedness. Therefore, a research question was proposed. To answer this research question, a one-way ANOVA test with perceived CSR as a dependent variable was undertaken. The test yielded a significant message sidedness effect, F (3, 201) = 15.74, p < .001.

PAGE 67

55 The results in Table 4-6 and Table4-7 show that there were significant mean differences between the treatment group and the control group. The mean score of overall perceived CSR for the control group (M=4.05) was significantly lower than those of treatment groups (M One-sided = 5.39; M NR Two-sided = 4.61; M R Two-sided = 4.86, Scheffe post-hoc test, p <. 05). In other words, participants perceptions about the companys corporate social performance were significantly different from that in the control group who did not receive CSR campaigns. Table 4-6. Perceived CSR by message sidedness Message Sidedness Control (N=21) One-Sided (N=61) NR Two-Sided (N=62) R Two-Sided (N=61) Overall Perceived CSR 4.00 (.62) 5.51 (.97) 4.61 (.88) 4.86 (1.15) Note. 1. NR stands for nonrefutational. R stands for refutational. 2. Standard deviation in parentheses. 3. The perceived CSR index shows the average of five items on a 7-point scale, with higher value indicating higher perceived CSR. Table 4-7. F-test of perceived CSR by message sidedness Df MS F Overall Perceived CSR Between groups 3 15.01 Within groups 201 .95 15.74*** Note. 1. NR stands for nonrefutational. R stands for refutational. 2. *** p < .001 In particular, significant mean differences were also found between the scores of perceived CSR among the three message sidedness groups. Table 4-8 shows the mean score of perceived CSR for the one-sided message group is significantly higher than that of two-sided message groups. It is noteworthy that participants of all the three groups received the same CSR message(see Appendix C & D) while only that in two-sided message groups received additional arguments on the clearcutting issue (see Appendix E

PAGE 68

56 & F). Moreover, participants in nonrefutational two-sided message groups rated the lowest score on perceived CSR among all the three groups. Participants in the refutational two-sided message groups, on the other hand had a lower score on perceived CSR than the one-sided message group. However, the finding is contrary to previous research. These results indicated that the one-sided CSR message is the most persuasive in influencing participants perceptions about the companys corporate social performance. Table 4-8. Perceived CSR by message sidedness Note. 1. NR stands for nonrefutational. R stands for refutational. Subset for Alpha .05 Sidedness N 1 2 3 Control 21 4.00 NR Two-Sided 62 4.61 R Two-Sided 61 4.86 One-Sided 5.51 2. Displayed are means for groups in homogeneous subset. 3. Scheffe post-hoc test was performed; p < .05. Interaction Effect on Perceived CSR As results of Table 4-9 and Table 4-10 show, a significant interaction effect was found only when the participant is in the low involvement condition. The results indicated that perceived CSR was more positive in a one-sided condition (M=5.85, SD=.82) than in the remaining conditions (M for NR TWO = 4.57, M for R TWO=4.70) under a low involvement condition, however the mean difference between the refutational two-sided group (M = 5.02) and the one-sided group (M = 5.18) decreased under a high involvement condition. The mean score of the nonrefutational two-sided group remained relatively unchanged by the two involvement levels. In short, the

PAGE 69

57 interaction effect of involvement and message sidedness on perceived CSR was statistically significant, F (2,178) = 4.17, p < .05. Table 4-9. Interaction of involvement and message sidedness on perceived CSR Low Involvement High Involvement Message Sidedness N M (SD) N M (SD) ONE-SIDED 30 5.85 (0.82) 31 5.18 (1.01) NR TWO 31 4.57 (0.89) 31 4.65 (0.88) R TWO 31 4.70 (1.14) 30 5.02 (1.09) Note. 1. NR stands for nonrefutational. R stands for refutational. 2. p < .05 Table 4-10 Tests of between-subject effects (dependent variable: perceived CSR) MS df F Intercept 4591.72 1 4656.13 *** Involvement .38 1 .38 Message Sidedness 13.40 2 13.59 *** Involvement* Message Sidedness 4.11 2 4.17 Error .99 178 Note. 1. p < .05; *** p < .001 2. R2 = .17 The Figure 4-1 clearly shows the differences in the interaction effects of message sidedness on perceived CSR when the participant is in the low involvement and the high involvement conditions. These findings indicated that a one-sided CSR message appeared to be most persuasive in influencing peoples perceptions about a companys corporate social performance. Particularly, the persuasion effect of a one-sided CSR message is more obvious when the message recipient is in a low-involvement situation. However, the advantage of persuasiveness for one-sided message decreased when involvement was high.

PAGE 70

58 An interaction of message sidedness and involvement on perceived CSR44.24.44.64.855.25.45.65.866.26.4LowHighinvolvement One-sided R Two-sided NR two-sided Figure 4-1. Interaction of involvement and message sidedness on perceived CSR. Attitudes toward a Company Hypothesis 3: Participants exposed to a refutation two-sided message would show more positive attitudes than those who were exposed to a one-sided or a non-refutational two-sided message. (Predicted main effect) Another important dependent variable that was widely studied by researchers was participants attitude toward a brand or a company. To test this Hypothesis 3, a one-way ANOVA test was applied to the overall attitudes toward CALCO, followed by the Scheffe homogeneous subset comparison procedure upon significance level at = .05. Table 4-11 and Table 14-2 demonstrate that results from the ANOVA test turn out to be contrary to the prediction of Hypothesis 3.

PAGE 71

59 Contrary to Hypothesis 3, it was found that the one-sided message group showed the most positive attitude toward CALCO with M=5.82, SD=1.03. Results from the one-way ANOVA test revealed a significant main effect of message sidedness on attitudes toward CALCO, F (3, 200) = 20.56, p < .001. Table 4-11. Attitudes toward CALCO by message sidedness Message Sidedness Control (N=21) One-sided (N=60) NR TWO (N=62) R TWO (N=61) Attitudes Toward CALCO 4.18 (.63) 5.82 (1.03) 4.73 (.96) 4.75 (1.11) Note. 1. NR stands for nonrefutational. R stands for refutational. 2. Two indicates two-sided. 3. Standard deviation in parentheses. 4. The attitude index shows the average of six items with a 7-point semantic differential scale, with higher value indicating more positive attitudes toward CALCO. This finding again suggested that one-sided CSR messages have an advantage in persuasiveness over refuational two-sided messages and non-refutational two-sided messages. According to the findings of the current study, one-sided CSR messages appear to be more persuasive in communicating CSR activities, in terms of generating positive attitudes toward the company and inducing higher perceptions of the companys corporate social performance. In other words, a one-sided CSR campaign presenting simply the positive information related to the social activities of the company generated more positive attitudes toward the company than a two-sided CSR campaign. Table 4-12. F-test for attitudes toward CALCO by message sidedness Df MS F Attitudes toward CALCO Between groups 3 20.77 Within groups 200 1.01 20.56*** Note. 1. NR stands for nonrefutational. R stands for refutational. 2. *** p < .001

PAGE 72

60 Since the main effect for control of message sidedness on attitudes toward the company was significant, a subsequent post-hoc analysis with Scheffes homogeneous subset test showed that the effects of message sidedness on the attitudes toward CALCO could be attributed to the differences between the one-sided message group and the two-sided message groups (see Table 4-13). The mean attitude score of the one-sided message group (M = 5.82) was significantly higher than that of both the nonrefutational two-sided message group (M= 4.73) and the refutational two-sided message group (M = 4.75) under p < .05. Table 4-13. Attitudes toward CALCO by message sidedness Note. 1. NR stands for nonrefutational. R stands for refutational. Subset for Alpha .05 Sidedness N 1 2 Control 21 4.18 NR TWO 62 4.73 R TWO 61 4.75 One-side d 5.82 2. Displayed are means for groups in homogeneous subset. 3. Scheffe post-hoc test was performed; p < .05. These findings indicated that significant differences were found between the one-sided message and the two-sided message (refutational two-sided and nonrefutational two-sided) in terms of the effects on perceived CSR and attitudes toward the company. However, no significant differences were found between the refutational two-sided message and the nonrefutational two-sided message. The results of the current study might explain why the mainstream of message sidedness research mainly focused on one-sided and two-sided messages, instead of comparing two types of two-sided messages. Interaction Effect on Attitude toward a Company Hypothesis 4: Participants in a low involvement condition will show more positive attitudes with a one-sided message while those in a high involvement condition

PAGE 73

61 would express more positive attitudes with a refutational two-sided message. (Predicted interaction effect on attitude) In Hypothesis 4, an interaction effect of message sidedness and involvement on attitude toward CALCO was predicted. It was more specifically hypothesized that participants in a low involvement condition would show more positive attitudes to a one-sided message while those in a high involvement condition would express more positive attitudes to a refutational two-sided message. Table 4-14 and Table 4-15 revealed that there was no significant interaction effect of message sidedness and involvement on attitudes toward the company. Therefore, Hypothesis 4 was rejected. There was only a main effect of message sidedness on attitude toward the company as confirmed in the earlier analysis. The one-sided message resulted in more positive attitude toward CALCO both in a high involvement and in a low involvement condition than other remaining conditions. Table 4-14. Interaction of involvement and message sidedness on attitudes Low Involvement High Involvement Message Sidedness N M (SD) N M (SD) One-Sided 30 5.93 (1.16) 30 5.71 (0.90) NR Two-Sided 31 4.45 (0.98) 31 5.01 (0.97) R Two-Sided 31 4.70 (1.27) 30 4.75 (1.11) Note. 1. NR stands for nonrefutational. R stands for refutational. 2. Two indicates two-sided. 3. Standard deviation in parentheses. 4. The attitude index shows the average of six items with a 7-point semantic differential scale, with higher value indicating more positive attitudes toward CALCO.

PAGE 74

62 Table 4-15. Tests of between-subject effects (dependent variable: attitudes) MS df F Intercept 4758.61 1 4474.25 *** Involvement 1.00 1 .94 Message Sidedness 23.31 2 21.92 *** Involvement* Message Sidedness 2.39 2 2.25 Error 188.25 177 Note. 1. *** p < .001 2. R2 = .22 A sub Analysis between Two Sample Groups The participants of the present experiment were recruited from two undergraduate class from two different colleges: the College of Agricultural and Life Sciences and the College of Journalism and Communications. It is appropriate to assume that there might be some differences between the two groups of participants because of the majors of students in the College of Agricultural and Life Sciences are related to environmental issues and therefore assumed to be more environmental concerned. A sub analysis was run between two groups of participants to test the above four hypotheses. However, the results did not indicate any significant statistical differences. A Correlation between Perceived CSR and Attitudes toward a Company Hypothesis 5: Higher perceived CSR would be positively associated with more positive attitude toward CALCO. As a test of this hypothesis, a correlation analysis was conducted. In order to determine any significant relationships between perceived CSR and attitudes toward the company, the Pearson correlation was undertaken. Pearson correlations coefficient is a measure of association which caries from -1 to +1, while 0 indicating no linear relationship, -1 indicating a perfect negative linear relationship and +1 indicating a perfect linear relationship (Garson, 2005).

PAGE 75

63 In the Table 4-16, the Pearson correlation coefficient shows that there is a positive correlation between perceived CSR and attitude toward a company. Thus, Hypothesis 5 was supported. This finding indicated that perceived CSR was highly associated with attitudes toward the company. In other words, when a person perceives a company as socially responsible, it was likely that he or she would have more positive attitude toward the company. Table 4-16. A Pearson Bivariate Correlation between perceived CSR and attitudes Perceived CSR Attitudes Perceived CSR 1.00 Attitudes .697 ** 1.00 Note. 1. ** p < .01 (2-tailed). A Hierarchical Regression Analysis between Perceived CSR and Attitudes toward a Company To look into the relationship between perceived CSR and attitude toward the company, a hierarchical regression analysis was performed (see Table 4-17). Attitudes toward CALCO was the dependent variable for this hierarchical regression model. In this hierarchical regression analysis, three blocks of variables were entered into the hierarchical regression analysis The three blocks of variables consisted of participants demographics, experimental variables (i.e., message sidedness and involvement), and mediators. Participants demographics consisted of age, gender, and ethnicity: gender and ethnicity were dummy-coded (M1). The experimental variables of involvement and messages sidedness were also entered as dummy variables (M2). Finally, in order to examine mediating impacts of cognitive elaboration and perceived CSR on the dependent variable, the total number of message-related thoughts and perceive CSR were entered into the third step (M3). The regression model helped verify the hypothesis by comparing the

PAGE 76

64 standardized coefficient () and R2. Standardized regression coefficients () and R-square changes (R2) in explained variance were examined. Table 4-17. Hierarchical regression analysis for proposed model Attitude toward the company as a DVa M1 M2 M3 Std. Error Std. Error Std. Error Demographics Age -.01 .08 .08 .07 .00 .06 Gender (female)b -.01 .08 .01 .07 -.02 .06 Ethnicity (White)b -.00 .08 -.01 .07 .01 .05 Experimental Variables One-sidedb .52*** .08 .21** .07 Refutational Twosidedb .13 .08 -.02 .07 Involvement: Highb .05 .07 .09 .05 Mediators Cognitive Elaboration .02 .05 Perceived CSR .63*** .06 Constant 5.11*** 3.80*** 1.50** R2 .00 .21 .54 Adjusted R2 -.02 .18 .52 R2 .00 .21 .33*** F for Regression .02 7.68*** 25.13*** Notes: Standardized regression coefficients () are shown; Std. Error means standard errors. a. DV = dependent variable b. dummy coded Mean for the items means the average score for the 7 items in the Fortune reputation index. ** p < .01; *** p < .01 In the regression analysis, three demographic variables were entered first. No significant influence was found from entering this block (i.e., M1) of variables. Since these demographics were considered as control variables in this experiment, it could be

PAGE 77

65 interpreted that individual differences that may hurt the results of this experiment were strongly controlled. The second bloc of the analysis (M2) observed a significant impact of message sidedness (i.e., one-sided message) on the dependent variables of attitudes toward CALCO ( = .52, p <. 001). The refutational two-sidedness entered as a dummy variable did not contribute to the dependent variable, nor did involvement manipulation. The experimental variables entered in step 2 accounted for 21 percent of the variance (p < .001). In the final step of analysis (M3), the entire variables were entered to look into a relative contribution of each variable to the dependent variable. The beta coefficient of message sidedness ( = .21, p < 01) was reduced but it was still significantly associated with attitudes toward CALCO, the dependent variable. Therefore, the one-sided message could be interpreted as a reliable contributor to positive attitudes toward CALCO. The results also showed that perceived CSR could be an effective predictor for achieving positive attitudes toward CALCO ( = .63, p < 001). Therefore, Hypothesis 5 that predicted a positive relationship between perceived CSR and attitudes toward the company was corroborated. In order to be perceived as socially responsible, the findings of this study suggest that companies should emphasize not only practicing CSR but also communicating its corporate social performance to their stakeholders. In brief, the findings of this study demonstrate the effectiveness of CSR campaigns on enhancing peoples perceptions of the corporate social performance and attitude toward the company. More specifically, in terms of CSR communication strategy, one-sided messages presenting supportive

PAGE 78

66 arguments for the companys social performances are more persuasive in influencing peoples perceptions in a desired direction regardless of involvement level. Path Analysis The researcher proposed a cognitive processing model of a CSR message in this study (see Figure 4-2). However, no significant difference of the numbers of message related thoughts was found between treatment groups. The assumption related to the main effects and interaction effects of message sidedness and issue involvement on participants levels of cognitive elaboration was rejected. Therefore, in order to revise the cognitive processing model of a CSR message, an additional path analysis was undertaken to explore the causal relationships between message sidedness, involvement, perceived CSR and attitude toward the company. Moreover, in response to the research call for effective communication methods of CSR (Clark, 2000), it is essential to demonstrate the impact of communicating CSR activities or commitment on achieving organizational goals. Based on the notion that effective CSR communication should have positive impact on organizational goals, the present study attempts to explore the causal relationships between the perceptions of a companys corporate social performance and the attitudes toward the company by conducting path analysis. Path analysis is a statistical procedure consisting of a series of regression analyses. The key advantage of using the path analysis approach in this research study is that the researcher can demonstrate the causal relationships between variables so that in turn the theories and relationships among variables can be clarified and strengthen (Agresti & Finlay, 1997).

PAGE 79

67 The equations for a five-variable path model were depicted in Figure 4-3. Based on the equations, the researcher tested direct and indirect effects of exogenous variables and different patterns of CSR messages and involvement level on both attitude toward the company and perceived CSR. It is assumed in this study that perceived CSR has a mediating effect on attitude toward the company. The results of the path analysis suggest that the one-sided message has both significantly direct effects on perceived CSR (P=.38, p< .001) and on attitude toward the company (P=.43, p< .001). Refutational two-sided message also has a direct effect on perceived CSR (P=.18, p< .05), but only marginal impact on attitude toward the company (P=.10, p< .10).The path analysis also indicated that a significant direct effect between perceived CSR and attitude toward the company (P=.65, p< .001). That is, the higher the person perceived a companys corporate social activities, the more likely to the person will induce positive attitudes toward the company. However, Figure 4-3 shows that involvement level did not link to either perceived CSR or attitude toward the company. In other words, according to the results of the path analysis, involvement level does not affect peoples perceptions of a companys corporate social performance and their attitudes toward the company. The most important findings in the causal model were related to the power of perceived CSR in predicting attitude toward the company. Consistent with the hypothesis of this study, it was found that perceived CSR was highly associated with attitude toward the company. The higher participants perceived the corporate social performance of a company, the more they had positive attitudes toward the company. In this regard, communicating a companys social performance is as important as conducting corporate

PAGE 80

68 social activities. Therefore it is needed to address a communication management perspective of corporate social responsibility in order to include CSR as part of a business strategy. Another interesting finding from this path analysis is the effects of message sidedness on perceived CSR and attitude toward the company. The path coefficients of message sidedness in the path model show that a one-sided message has a direct impact on perceived CSR and attitude toward the company. In other words, a one-sided CSR campaign which presents only positive information has a significant advantage over a two-sided CSR campaign in influencing message recipients perceptions of the companys social performance and their attitudes toward the company. The involvement variable however, has no direct effect on perceived CSR nor has it on attitudes toward the company. Regardless of involvement level, a CSR campaign has the same effect on message recipients perceived CSR and their attitudes toward the company. Therefore, according to the results of the path analysis, a revised cognitive processing model of a CSR message is presented in Figure 4-4. The effect of message sidedness is highly associated with both perceived CSR, and attitude toward the company. In brief, the results of the path analysis show that five paths, all direct, are statistically significant (see Figure 4-3): from a one-sided message to perceived CSR (P=.38, p< .001), from a one-sided message to attitude toward the company (P=.43, p< .001), from perceived CSR to attitude toward the company (P=.65, p< .001), from refutational two-sided message to perceived CSR (P=.18, p< .05), from refutational two-sided message to attitude toward the company (P=.10, p< .10).

PAGE 81

69 Perceived CSR Message Sidedness Cognitive Elaboration Issue Involvement Attitude to the Company Figure 4-2. The original cognitive processing model of a CSR message proposed in this research. One-Sided Message .43*** .38*** .65*** Perceived CSR Attitude toward the Company Refutational Two-Sided Message .18* .10 n.s. n.s. Involvement Figure 4-3. The path diagram based on the path analysis. Note. 1. Message sidedness and involvement were dummy coded. 2. p < .10, p < .05, *** p< .001 3. n.s. represents nonsignificant

PAGE 82

70 Message Attitude toward the Company Perceived CSR Sidedness Figure 4-4. The revised cognitive processing model of a CSR message proposed in this research.

PAGE 83

CHAPTER 5 DISCUSSION Over View of the Present Study The present study represents an innovative experiment of a cognitive processing model of a CSR message. The purpose of this study is to test empirically the persuasive effect of a CSR campaign in influencing participants attitudes toward a company and perceived CSR regarding involvement level and the effect of message sidedness. In particular, the relationships of a CSR campaign on attitudes toward the company via message recipients perceptions of the companys social performance were verified in this study. By testing the persuasiveness of a CSR campaign and proposing a cognitive processing model, this study sheds light on the research of CSR communication and persuasion. Even though previous research has discussed CSR from the issues management perspective (Carroll, 1991; Clark, 2000; Esrock & Leichty, 1998; Heath, 1998; Heath & Ryan, 1989; LEtant, 1994; Wartick & Cochran, 1985; Wartick & Rude, 1986), the discussions about the effects of a CSR campaign on public perceptions of a companys social performance were missing. To fill this research lacuna, this study explored the persuasiveness of a CSR message in influencing public perceptions of a companys social performance and attitudes toward the company. The present study asked whether presenting a one-sided CSR campaign or a two-sided CSR campaign is more persuasive in advocating a companys corporate citizenship and addressing the controversial issues related to the 71

PAGE 84

72 fundamental practice of the business. The two-sided CSR campaigns were further broken down into two different forms: 1) a nonrefutational two-sided campaign in which the opposing messages were shown without additional refutations, and 2) a refutational two-side message in which the opposing arguments were shown with refutational explanations. From the public relations perspectives, this study contributed to explore the effectiveness of CSR campaigns as proactive tools of issues management. The main objective of the present study was to test whether attitude toward the company can be influenced by peoples perceptions of the companys corporate social performance. Moreover, this study also explored the functions of a CSR campaign not only in advocating the corporate citizenship of the company but also preventing crisis situations by actively addressing controversial issues with external stakeholders. Little research had empirically examined the effectiveness of a CSR campaign. The present study attempted to explore this uncharted research area with relevant theories in persuasion, such as the cognitive response research and the ELM. In addition, this study tested the causal relations between perceived CSR and attitude toward the company. The experimental design of this study can provide some practical implications to public relations practitioners who have to communicate with active publics concerning a critical issue. According to Grunigs (1997) situational theory, the active public is identified as a high involvement group. It is assumed in this study that if participants are highly involved with the issue, they would generate more message-related thoughts particularly when they were exposed to a refutational two-sided message.

PAGE 85

73 In addition, another important assumption related to involvement level was message recipients cognitive elaboration of the CSR messages. Based on the ELM, it was reasoned in this study that an individuals perceptions about a companys corporate social performance and attitudes toward the company would be mediated by his or her motivation to process the message and the issue presented in CSR campaigns. However, findings in this study did not support the assumption that message sidedness and issue involvement have impact on participants cognitive elaboration level of processing a CSR message. Even though there were no significant differences between the numbers of recipients message-related thoughts, it was found in this study that a one-sided CSR campaign has an advantage of persuasiveness over both refutational two-sided and nonrefutational two-sided CSR campaigns. The findings in this study indicated that presenting only supportive messages about a companys corporate social activities will induce more positive attitudes toward the company. Regardless of the involvement level, people are likely to form a positive attitude toward a company when the company is considered as socially responsible. Contrary to previous findings on the effect of message sidedness, it was found in this study that a one-sided CSR message had generated more positive attitudes toward the company and more positive perceptions of the companys corporate social achievements than two-sided one. In addition, this study also found that the perceived CSR was positively associated with attitudes toward the company (r = .697), implying that the perceived CSR might play a mediating role in the process of attitude formation when a person was exposed to a CSR advocacy message.

PAGE 86

74 Overview of the Hypotheses and Research Question In the following section, each of the five hypotheses and one research question is discussed in detail based on the results of this study. Hypothesis 1: Participants in a high involvement condition will generate more message-related thoughts than those in a low involvement condition The first hypothesis investigated the main effect of involvement level on cognitive elaboration of the CSR messages presented in the experiment. To test this hypothesis, a t-test was conducted. However, the results showed no statistical differences of message-related thoughts between a high and a low involvement group (p= .398). Thus, Hypothesis 1 was rejected. Even though a successful manipulation check of involvement was obtained in this study, participants in high involvement groups did not demonstrate higher level of cognitive elaboration as previous research has indicated. The results showed that regardless of participants involvement level, the numbers of message-related thoughts generated by participants in high involvement conditions were not significantly different from that in low involvement conditions. One explanation of these results is that the manipulation of involvement used in this experiment did not affect respondents motivation to process the messages presented in the stimuli booklets. In this study, participants level of involvement was manipulated by the location of the lumber company presented in the message stimulus. For high involvement groups, the lumber company presented in the stimuli was located in Florida, whereas the lumber company was located in Taiwan for low involvement groups. Involvement level was

PAGE 87

75 measured using four questions regarding participants concerns about the environmental issues in Florida or in Taiwan. Even thought the manipulation check of involvement worked successfully in this experiment, it did not show that participants motivation to cognitively processing the CSR messages was also manipulated by the stimuli. Therefore, it is appropriate to assume that the amounts of effort participants invested in reading the CSR messages and answering the questionnaire were not influenced by the involvement manipulation presented in the stimuli materials. Participants in the high involvement group were not necessary more motivated to process the stimuli even though their responses to the manipulation check were significantly different from that in the low involvement group. Therefore, to test the cognitive elaboration, future research should consider other approaches to enhance participants motivation in responding experimental materials. Hypothesis 2: A two-sided refutational message will induce more message-related thoughts than a one-sided or a nonrefutational two-sided message. To test Hypothesis 2, a One-Way ANOVA test was run. In this experiment, no statistically significant differences were found between the numbers of message-related thoughts from different message sidedness groups. Thus, Hypothesis 2 was also rejected. For cognitive elaboration, there were no significant effects attributed to involvement and message sidedness. However, a previous research by Hale et al. (1991) found that more positive cognitions were generated by refutational two-sided messages than nonrefutational two-sided messages and one-sided messages. Contrary to previous study, this experiment found that one-sided CSR message had induced more message related thoughts than two

PAGE 88

76 sided CSR messages. One explanation may be that the total number of message-related thoughts was considered for measuring cognitive elaboration in this study, instead of the number of positive and negative thoughts. In conclusion, these results failed to support the main effects of message sidedness and involvement on cognitive elaboration. Further analysis is needed to revise the cognitive elaboration model of a CSR message proposed in the beginning of this study. Therefore, path analysis, regression analysis and Pearson correlation analysis were conducted in this study and discussed in the following section to explain the causal relationships between the three variables in the revised cognitive elaboration model of a CSR message. RQ1: How does message sidedness influence peoples perception about corporate social responsibility of a company? What will lead to more positive perceived CSR between a two-sided message and a one-sided message? To answer this research question, a one-way ANOVA test with perceived CSR as a dependent variable was undertaken. The test yielded a significant message sidedness effect and the results showed that participants perceptions about the companys corporate social performance were significantly influenced by CSR campaigns. In particular, these results indicated that the one-sided CSR message is the most persuasive in influencing participants perceptions about the companys corporate social performance. A significant interaction effect was also found in this study. However, the interaction effect was limited to the low involvement condition. The results indicated that

PAGE 89

77 one-sided CSR message had induced more positive perceived CSR in the low involvement condition. In conclusion, the present study confirmed previous studies that demonstrated a relationship between CSR communication and perceptions about a companys corporate social performance. The CSR campaign designed for the present experiment appears to represent the majority of CSR initiatives and activities, because the three areas of CSR activities chosen for research, environment, employees and community, are the focus of CSR initiatives and activities for the majority of todays corporations. Hypothesis 3: Participants exposed to a refutation two-sided message would show more positive attitudes than those who were exposed to a one-sided or a non-refutational two-sided message. (Predicted main effect) To test this Hypothesis 3, a one-way ANOVA test was applied to the overall attitudes toward the company. Results revealed a significant main effect of message sidedness on attitudes toward the company (p < .001). However, contrary to the prediction, this finding again suggested that one-sided CSR messages have an advantage in persuasiveness over refuational two-sided messages and non-refutational two-sided messages. Hypothesis 4: Participants in a low involvement condition will show more positive attitudes with a one-sided message while those in a high involvement condition will express more positive attitudes with a refutational two-sided message. (Predicted interaction effect on attitude) The results indicated that no interaction effect of message sidedness and involvement on attitudes toward the company was found. Therefore, Hypothesis 4 was

PAGE 90

78 rejected. There was only a main effect of message sidedness on attitude toward the company as confirmed in the earlier analysis. Two main effects of message sidedness were found on perceived CSR and attitude toward the company, while only one interaction effect between message sidedness and perceived CSR was found when participants involvement levels were low. In conclusion, the results to above hypotheses showed that (1) generally the effect of message sidedness on perceived CSR and on attitude toward the company were significant, (2) contrary to previous research, one-sided CSR messages appear to have an overall advantage of persuasiveness in influencing peoples perceptions about a companys social performance and their attitude toward the company, and (3) the relationships between message sidedness and involvement on cognitive elaboration were do found. Based on these findings, it can be suggested that the proposed cognitive elaboration model of processing a CSR message can be revised by taking out the cognitive elaboration. Hypothesis 5: Higher perceived CSR will be positively associated with more positive attitude toward CALCO. As expected, the Pearson correlation coefficient shows that there is a positive correlation between perceived CSR and attitude toward a company. This finding indicated that perceived CSR was highly associated with attitudes toward the company. A hierarchical regression analysis was performed to look into the relationship between CSR messages, perceived CSR and attitude toward the company. The regression analyses tested whether CSR messages and perceived CSR had significant effect on predicting

PAGE 91

79 attitudes toward the company. According to the results, the one-sided message could be interpreted as a reliable contributor to positive attitudes toward CALCO. Using the path analysis, the researcher examined the effects of message sidedness and involvement had on perceived CSR and attitude toward the company. As expected, the analysis showed that both message sidedness had both direct effects on perceived CSR and attitude toward the company, while no significant effects were found between involvement and other variables. The results of path analysis also indicated the direct relationship between perceived CSR and attitude toward the company is statistically significant. Path analysis confirmed five paths, all direct, are statistically significant: one-sided message had two direct effects on perceived CSR and attitude toward the company, while refutational two-sided message also had two direct effects on perceived CSR and attitude toward the company, and refutational two-sided message had a direct effect on attitude toward the company. Consequently, the present study proposed a model of relationship between message sidedness, perceived CSR and attitude toward the company as the final model. In conclusion, the present study demonstrated that (1) CSR messages (one-sided message and refutational two-sided message) were significantly associated with perceived CSR, (2) perceived CSR was significantly associated with attitude toward the company. Although most of hypotheses were rejected except for the last one hypothesized on the relationship between perceived CSR and attitude toward the company, some of interesting and information were found. First, regardless of peoples involvement level

PAGE 92

80 with issues, the one-sided CSR message has an overall advantage of persuasiveness in influencing peoples perceptions about a companys corporate social performance and their attitudes toward the company. Contrary to previous research, the results of the present study reflected the conflicting findings of the mainstream research in message sidedness. Second, perceived CSR was found to have a significant effect on predicting attitude toward the company. The present study had verified the relationship between a CSR message, perceived CSR and attitude toward the company. The results of this study suggest that CSR communication is highly associated with the attitude toward the company. In particular, a one-sided CSR message presenting only positive information about the companys social commitment and activities are most persuasive in positively influencing the attitude toward the company. Despite these unexpected results related to the message sidedness effect, the present study supports the proposition that perceived CSR can function as an important predictor of attitude toward the company. The present study concluded that CSR campaigns had a directly effect on people perceptions about the companys social performance, while these perceived CSR also directly influence attitudes toward the company.

PAGE 93

CHAPTER 6 CONCLUSION It has been a crucial topic in public relations research to explore the ways in which public relations can contribute to organizational goals. Clark (2000) indicated a set of similarities between CSR and public relations and suggested that both disciplines are seeking to enhance the quality of the relationship of an organization among key stakeholder groups. However, for both fields to solidify a permanent place as a management function, more research into the value of their efforts is needed. Moreover, recent developments in the fields of CSR and public relations suggest that reputation is an important area for future research to measure the value of both fields (Clark, 2000). According to Fombrun (1996), organizations can build a sound reputation through effective communication. Previous research in communication has looked specifically into how best to communicate the social responsibility of an organization (J. E. Grunig, 1986; Ledingham & Bruning, 1999). However, more specifically, a question about the chosen messages presented in CSR campaigns and how they affect the reputation or perception of an organization as responsible remain (Clark, 2000, p.376). To overcome the research gap, the present study explored empirically the persuasive effects of a CSR message on attitude toward the company and perceived CSR. The effect of message sidedness and involvement level were also tested as a way of designing an effective CSR communication strategy. 81

PAGE 94

82 The results of this study indicated that presenting a one-sided CSR campaign which presented only positive information about a companys social activities was more persuasive in influencing public perceptions about the companys social performance. Moreover, this study verified the relationships between the CSR message, perceived CSR and attitude toward the company. The results of this study demonstrated that there were direct effects between the CSR message, perceived CSR and attitude toward the company. The results of the present study not only responded to Clarks (2000) research call on the effects of CSR messages but also verified the communication management functions on CSR. It was showed in this study that an effective CSR campaign can influence public perceptions about a companys social performance and induce positive attitude toward the company. The present study had build on the knowledge of CSR communication, cognitive response, and persuasive effect of message sidedness on CSR activities. Discussions about the Effect of Message Sidedness A couple of previous studies found that a one-sided message had more persuasive effect than a refutational two-sided message (e.g., Halverson, 1975; Smith et al, 1994). OKeefe (1998) had argued that the conclusion of the message sidedness effect should be interpreted carefully by considering the message topics. He argued that: given the different effects of nonrefutational forms in advertising and nonadvertising contexts, one should not too easily assume that refutational forms will function identically in the two circumstances (p. 238). Even thought an abundant number of studies have shown superior persuasiveness of refutational two-sided messages, other studies have found that one-sided messages can

PAGE 95

83 be more persuasive under certain conditions, implying the existence of moderators in the message sidedness effect. For future research on message sidedness, the possible moderators of the effect of message sidedness were discussed. From the very beginning of sidedness research, a number of possible moderators have been proposed. Hovland et al. (1949, p.225), for example, suggested that the audiences educational level is an important determinant of the consequences of sidedness variations. Several variables such as the message recipients initial attitude toward the sources position, topic familiarity (Allen, 1991), and intelligence (Hovland et al., 1949; Allen et al., 1990) are commonly mentioned as moderators of the persuasive effects of sidedness variations. Other proposed moderators have included perceived source motivation (Pechmann, 1990), and exposure to subsequent opposing communications (Lumsdaine & Janis, 1953). Specifically, one-sided messages were posited to be more effective than two-sided messages when the recipient initially agreed with the sources position, while being unfamiliar with the topic, or unintelligent (Allen et al., 1990). The following sections are discussions about the moderators of the effect of message sidedness. Even though these variables were controlled or in this experiment, previous studies had indicated that they may have an impact on the persuasiveness of the message sidedness which required some attention for future research. Credibility Perceptions Credibility perceptions are very important to the interpretation of the findings from this study. According to Allen et al.s (1990) meta-analytic review, the credibility perception about the communicator was reported to play an important role in a persuasive process. In another comprehensive meta-analysis of the message sidedness effect by

PAGE 96

84 OKeefes (2000), the credibility perceptions were discussed in terms of a mediator in a persuasive process. The variable of source credibility was controlled in this experiment. All the participants received CSR messages from the same source: the CALCOs company Web site. Previous research indicates that refutational two-sided messages might evoke closer message scrutiny, and the persuasive effect was discounted because the message source could have been perceived to be less credible. In other words, participants may discount the message if it is delivered from a less credible source or because the message is considered as the one with a self-serving bias. Regarding the impact of message-sidedness on persuasion, Chebat, Filiatrault, Laroche, and Watson (1988, p.619) had an important finding. They found that the attention span of negatively oriented recipients was more limited when the source was less credible and the message was two-sided. Therefore, an unexpected finding regarding the message sidedness in this study might be attributed to the impact of source credibility. Pre-Attitudes toward an Issue Previous research found that a one-sided message could be more persuasive when the audience was positively predisposed (Etgar & Goodwin, 1982). As discussed previously, a one-sided message is more persuasive to message recipients with positive initial attitudes toward an issue while a two-sided message is more persuasive to those who initially have negative attitudes. In this experiment, some participants in the experimental conditions reported that the background information about CALOC which is delivered on CALCOs About Us page generated positive initial attitudes toward the company. Even though we tried to deliver a brief company introduction, we had not anticipated that a well-designed

PAGE 97

85 corporate Web page would have such an effect in forming a positive pre-attitude toward the company. Fear Arousal Smith et al.s (1994) study on the persuasive message regarding organ donation found that a one-sided message was more persuasive than a refutational one. Their study showed the effects of prior thoughts and intent on outcomes associated with persuasion. They found that refutational messages induced significantly higher fear and anxiety in the segment of audience who are lacking in prior thought and intent to donate an organ. In other words, the decreased persuasive effects of refutational two-sided messages in the present research might also be associated with unintended consequences such as the recipients avoidance of negative discussion. It was possible that participants would have perceived the refutational messages to be self-serving, which in turn generates negative attitudes toward the message itself and consequently toward the company. Application The current study responded to the research call of developing an effective method in communicating CSR. The findings in this study suggest that a one-sided CSR message may have an advantage in communicating the companys corporate social performance as opposed to a two-sided one. More importantly, findings in this study also confirmed that communicating corporate social responsibility enhanced peoples perceptions about the companys social performance and also generated more positive attitudes toward the company advocated. From the positive role of a CSR campaign in generating positive image and attitudes toward company, the CSR could be recommended as an effective tool of building a long-term relationship between the public and an organization. Further research examining the effect of CSR on building relationships is needed.

PAGE 98

86 However, the conclusion of this study should be overstated because of several limitations mentioned previously. Regarding the level of issue involvement, the participants recruited in this study may mirror the general public who do not particularly concern about environmental issues. The results of this study did not suggest that all CSR campaigns should be presented as one-sided. On the contrary, for public relations practitioner to design an effective CSR communication strategy, it is essential to consider the characteristic of targeted audiences. For example, if the CSR campaign is targeted to communicate with environmental active audiences, then the conclusion of this study should be applied with caution. Future research is needed to explore the impact of CSR campaign on different audiences Limitations This study examines only a small part of the process and the effects of CSR communication with a limited group of people and limited stimulus message. The results of this experimental study offer interesting and useful findings for public relations practitioners and researchers. However, any generalization from the findings of this study should be made with caution. Despite a couple of limitations, this study could be improved with some considerations. In this research, the total number of message-related thoughts was considered for measuring cognitive elaboration, instead of the number of positive and negative thoughts. Several studies in this cognitive response paradigm suggest that valence of cognitive elaboration be considered. In the current research, the researcher did not test whether a refutaitonal two-sided message generates more positive message-related thoughts or it leads to negative cognitive elaboration. Future research can test the

PAGE 99

87 cognitive process model of a CSR message with the consideration of valence of cognitive elaboration. Suggestions for Future Research Knowledge Level Previous research in persuasion has indicated that participants educational levels have an influence on their responses of a one-sided or a two-sided message (Chu, 1967; Hovland, Lumsdanie, & Sheffield, 1949). It has been found that lower education people would be more likely to respond to one-sided messages. However, controlling for the educational level, the preset study still found that one-sided CSR messages had an advantage of influencing attitude toward the company and perceived CSR. The clearcutting issue discussed in the stimuli materials was not a common topic in Florida areas. Therefore, it was assumed that most participants in this study were not familiar with the issue and had a low knowledge level. That is, because of the lower knowledge level about the clearcutting issue, participants in this study were more likely to respond to a one-sided message, regardless of their educational level. Message Length Apparently, another consideration is the length of message in this study. The lengths of two-sided CSR messages are much longer than the one-sided CSR message. Previous research suggested that the message length has an impact on the persuasive effect of message sidedness. It was found that low involvement people are more likely to respond to short message length. Therefore, for future research, it is suggested that the length of messages should be controlled.

PAGE 100

88 Source Credibility As mentioned earlier, the source perception of the CSR message may have influenced the message effect. Tit is possible that the audiences will discount the corporate source, presumably because they thought that the arguments presented from the corporate source at best served for self-interests. Source credibility is a key role in persuasion and has been studied widely by researchers in different fields, such as marketing, advertising and social psychology. Future research could advance the current research with a consideration of source credibility manipulation. For instance, a future research design might use an experimental study that presents the same message used in the current study from either a high credible source (e.g., from a third party or a researcher in forestry) or the company. Attitudes toward Issues From an issues management perspective, future research may add some considerations on the interaction between peoples attitudes toward an issue and attitudes toward the company that advocated the issue. An additional dependent variable such as the attitudes toward the issue may be included in an experimental study via a pre-and posttest, so that an attitudinal change can be measured. The present research has shed light on developing an effective communication strategy on CSR based on the message sidedness effect. However, in order to fully understand the persuasive effects of a CSR campaign on peoples perceived CSR, attitudes toward the company, and even the long-term relationship building between the company advocated and stakeholders, additional primary research is needed concerning the sidedness effect on CSR messages.

PAGE 101

89 Contrary to previous studies, findings in this study show that a one-sided CSR campaign turned out to be more persuasive than a two-sided one. As discussed previously, unintended consequences regarding the message receivers skepticism of the credibility of acknowledged counterarguments might play a mediating effect on the persuasiveness of message sidedness. Differences in initial mistrust might be responsible for the observed differences in the perceived credibility of presenting counterarguments in the CSR messages. Exploration of the role of initial receiver skepticism seems warranted. This possibility might be explored through systematic examination of the relationship between sidedness effects and receivers background expectations about the sidedness of messages that are potentially characterized by such mistrust. Finally, future research may also consider other topics of a CSR campaign considering the companys unique characteristics of the industry: for instance, an experiment to test the persuasive effects of a CSR campaign targeting the AIDS issue in Africa a global pharmaceutical company.

PAGE 102

APPENDIX A CALCO ABOUT USUNITED STATES 90

PAGE 103

APPENDIX B CALCO ABOUT USTAIWAN VERSION 91

PAGE 104

APPENDIX C CALCO ENVIRONMENTAL COMMITMENT-UNITED STATES 92

PAGE 105

APPENDIX D CALCO ENVIRONMENTAL COMMITMENT-TAIWAN 93

PAGE 106

APPENDIX E CALCO MYTH & FACT 94

PAGE 107

APPENDIX F CALCO ENVIRONMENTAL COMMITMENT-NONREFUTATIONL 95

PAGE 108

APPENDIX G QUESTIONNAIRE FOR HIGH INVOLVEMENT GROUPS Q1. Please evaluate how you feel about CALCO by circling a number on each of the scales below. If y Based on CALCOs statement you just read from the printed Web pages, please answer the following questions. Please circle the number that best describes your thoughts or feelings. ou feel that you have no reaction, please circle the number 4 to indicate yoneutra ur lity. 1 Negative 1 2 3 4 5 6 7 Positive 2 HBeneficial armful 1 2 3 4 5 6 7 3 Foolish 1 2 3 4 5 6 7 Wise 4 Unly Friendly friend 1 2 3 4 5 6 7 5 Bad 1 2 3 4 5 6 7 Good 6 Unfavorable 1 2 3 4 5 6 7 Favorable Q2. Please evaluate how you feel about CALCOs environmental policies Please circle the number that best indicates your agreement with each item. If you strongly disagree with the provided statement, please circle . If you strongly agree with tprovided statement, pleas he e circle 7. If you have no reaction, please circle to dicate your neutrality. nt do you agree with the following in To what exte statements? Strongly Strongly Disagree Agree (2) CALCO is likel y to follow environment-friendly 1 2 3 4 5 6 7 rules and policies. (3) CAL CO is highly concerned about environmental 1 2 3 4 5 6 7 issues. (1) CALCO believes in environmental commitment. 1 2 3 4 5 6 7 (4) CALCO is highly involved in community activities. 1 2 3 4 5 6 7 (5) CALCO is genuinely concerned about public welfare. 1 2 3 4 5 6 7 [Please go to the next page] 96

PAGE 109

97 Q3. Please evaluate how you feel about the statement provided by CALCO. Please circle the number that best indicates your agreement with each item. If you stronglydisagree with the provide d statement, please circle . If you strongly agree with the rovided statement, please circle . If you have no reaction, please circle to statements? p indicate your neutrality. To what extent do you agree with the following Strongly Strongly Disagree Agree (1) CALCOs statement made th eir points effectively. 1 2 3 4 5 6 7 (2) I like CALCOs statement. 1 2 3 4 5 6 7 (3) The arguments presented in CALCOs statement were convincing. 1 2 3 4 5 6 7 (4) Considering both content and style, I think CALCOstatement is well written. s 1 2 3 4 5 6 7 (5) CALCOs statement was effective in increasing the commitment. awareness about the companys environmental 1 2 3 4 5 6 7 (6) CALCOs statement shows two-sided arg uments. 1 2 3 4 5 6 7 (7) CALCO refutes opponents arguments. 1 2 3 4 5 6 7 (8) CALCOs statement helps me understand the issu e. 1 2 3 4 5 6 7 (9) CALCOs statement corrects misunde rstanding. 1 2 3 4 5 6 7 (1 0) CALCOs statement provides facts. 1 2 3 4 5 6 7 Q4. Please give your impressions or feelings about CALCO on each of the scales below. If you have no reaction to CALCO on either s cale, please circle the number 4 to dicate your neutrality. For example, how trustworthy or untrustworthy do you think CACOabo theirmmitnts? U T in L is to inform you ut co me 1 ntrustworthy 1 2 3 4 5 6 7 rustworthy 2 Selfish 1 2 3 4 5 6 7 Unselfish 3 U Believable nbelievable 1 2 3 4 5 6 7 4 Not reliable 1 2 3 4 5 6 7 Reliable 5 Not honest 1 2 3 4 5 6 7 Honest [Please go to the next p age]

PAGE 110

98 Q 5. Please list the thoughts that came to your mind while reading CALCOs statement ought per line. Please spend your time to list your thoughts. Thank you. hought1________________________________________________________________ _______________________________________________________________________ hought2________________________________________________________________ _______________________________________________________________________ hought3________________________________________________________________ _______________________________________________________________________ hought4________________________________________________________________ _______________________________________________________________________ hought5________________________________________________________________ _______________________________________________________________________ hought6________________________________________________________________ _______________________________________________________________________ hought7________________________________________________________________ _______________________________________________________________________ hought8________________________________________________________________ ________________________________________________________________________ [Please go to the next page] as many as you can generate. Please write your thoughts in the following lines, one th T T T T T T T T

PAGE 111

99 Q 6. Please evaluate how you respond to the following questions. indicates not at all tion, pleasthe following N and indicates very much. If you have no reacneutrality. e circle to indicate your To what extent do you respond to questions? ot at All Very Much (1) How much did you p ay attention to read the 1 2 3 4 5 6 7 statement presented by CALCO? (2) How much were you motivated to read the statement presented by CALCO? 1 2 3 4 5 6 7 (3) How much di d you make an effort to think of the 1 2 3 4 5 6 7 issue addressed in the statement presented by CALCO? (4) How importan t are environmental issues in Florida to 1 2 3 4 5 6 7 you personally? (5) How important is the deforestation issue in Florida to 1 2 3 4 5 6 7 you personally? (6) How much are you concerned about environmental 1 2 3 4 5 6 7 issues in Florida? (7) How mu ch are you concerned about the deforestation 1 2 3 4 5 6 7 issue in Florida? (8) How often do you think about environmental issues 1 2 3 4 5 6 7 in general? (9) How often do you think about the deforestation issue 1 2 3 4 5 6 7 in particular? (10) How confident are you that you can do something to lve environment issues in general? 1 2 3 4 5 6 7 so (1 1) How confident are you that you can do something to lve the deforestation issue in particular? 1 2 3 4 5 6 7 so [Please go to the next page]

PAGE 112

100 Q7. Please evaluate how you feel about the relationship between you and CALCO. Please circle the number tha t best indicates your agreement with each item. If you rongly disagree with the provided statement, please circle . If you strongly agree no reactionour neutrality. SDisagree Agree st with the provided, please circle . If you have to indicate y please circle in the box To what extent do you agree with the following statements trongly Strongly know it will be concerned about people like me. (2) CALCO can be relied on to keep its promis es. 1 2 3 4 5 6 7 (3) I beli eve that CALCO takes the opinions of people 1 2 3 4 5 6 7 like me into account when making decisions. will do. (5) Sound principles seem to guid e CALCOs behavior. 1 2 3 4 5 6 7 (6) I can see that CALCO wants to maintain a 1 2 3 4 5 6 7 relationship with people like me. (7) There is a long-lasting bond between CALCO and 1 2 3 4 5 6 7 people like me. relationship. (9) I feel people like me are important to CALCO. 1 2 3 4 5 6 7 (10) CALCO see (1) Whenever CALCO makes an important decision, I 1 2 3 4 5 6 7 (4)CALCO has the ability to accomplish what it says it 1 2 3 4 5 6 7 (8) CALCO and people like me benefit from their 1 2 3 4 5 6 7 ms to be the kind of company that the community. 1 2 3 4 5 6 7 invests in (11) CAs the adevelo and keep le1 2 3 4 5 6 7 LCO hability to attract, p, talented peop. G ender [ ] What year are you in the college? ] Ethnic Native American [ ] Asian/Pacific IsOthers (Please specify): [ ] This is the End. I appreciate your participation. Male [ ], Female [ ] Age [ icity White-Non Hispan[ ] Hispanic American [ ] African American [ ] lander [ ]

PAGE 113

APPENDIX H QUESTIONNAIRE FOR LOW INVOLVEMENT GROUPS 101 Q1. Please evaluate how you feel about CALCO by circling a number on each of the scales below. If y Based on CALCOs statement you just read from the printed Web pages, please answer the following questions. Please circle the number that best describes your thoughts or feelings. ou feel that you have no reaction, please circle the number 4 to indicate yoneutra ur lity. 1 Negative 1 2 3 4 5 6 7 Positive 2 HBeneficial armful 1 2 3 4 5 6 7 3 Foolish 1 2 3 4 5 6 7 Wise 4 Unly Friendly friend 1 2 3 4 5 6 7 5 Bad 1 2 3 4 5 6 7 Good 6 Unfavorable 1 2 3 4 5 6 7 Favorable Q2. Please evaluate how you feel about CALCOs environmental policies Please circle the number that best indicates your agreement with each item. If you strongly disagree with the provided statement, please circle . If you strongly agree with tprovided statement, pleas he e circle 7. If you have no reaction, please circle to dicate your neutrality. nt do you agree with the following in To what exte statements? Strongly Strongly Disagree Agree (2) CALCO is likel y to follow environment-friendly 1 2 3 4 5 6 7 rules and policies. (3) CAL CO is highly concerned about environmental 1 2 3 4 5 6 7 issues. (1) CALCO believes in environmental commitment. 1 2 3 4 5 6 7 (4) CALCO is highly involved in community activities. 1 2 3 4 5 6 7 (5) CALCO is genuinely concerned about public welfare. 1 2 3 4 5 6 7 [Please go to the next page]

PAGE 114

102 Q3. Please evaluate how you feel about the statement provided by CALCO. Please circle the number that best indicates your agreement with each item. If you strongly disagree with the provided statement, please circle . If you strongly agree with the provided statement, please circle . If you have no reaction, please circle to indicate your neutrality. To what extent do you agree with the following statements? Strongly Strongly Disagree Agree (1) CALCOs statement made their points effectively. 1 2 3 4 5 6 7 (2) I like CALCOs statement. 1 2 3 4 5 6 7 (3) The arguments presented in CALCOs statement were convincing. 1 2 3 4 5 6 7 (4) Considering both content and style, I think CALCOs statement is well written. 1 2 3 4 5 6 7 (5) CALCOs statement was effective in increasing the awareness about the companys environmental commitment. 1 2 3 4 5 6 7 (6) CALCOs statement shows two-sided arguments. 1 2 3 4 5 6 7 (7) CALCO refutes opponents arguments. 1 2 3 4 5 6 7 (8) CALCOs statement helps me understand the issue. 1 2 3 4 5 6 7 (9) CALCOs statement corrects misunderstanding. 1 2 3 4 5 6 7 (10) CALCOs statement provides facts. 1 2 3 4 5 6 7 Q4. Please give your impressions or feelings about CALCO on each of the scales below. If you have no reaction to CALCO on either scale, please circle the number 4 to indicate your neutrality. For example, how trustworthy or untrustworthy do you think CALCO is to inform you about their commitments? 1 Untrustworthy 1 2 3 4 5 6 7 Trustworthy 2 Selfish 1 2 3 4 5 6 7 Unselfish 3 Unbelievable 1 2 3 4 5 6 7 Believable 4 Not reliable 1 2 3 4 5 6 7 Reliable 5 Not honest 1 2 3 4 5 6 7 Honest [Please go to the next page]

PAGE 115

103 Q5. Please list the thoughts that came to your mind while reading CALCOs statement as many as you can generate. Please write your thoughts in the following lines, one thought per line. Please spend your time to list your thoughts. Thank you. Thought1________________________________________________________________ ________________________________________________________________________ Thought2________________________________________________________________ ________________________________________________________________________ Thought3________________________________________________________________ ________________________________________________________________________ Thought4________________________________________________________________ ________________________________________________________________________ Thought5________________________________________________________________ ________________________________________________________________________ Thought6________________________________________________________________ ________________________________________________________________________ Thought7________________________________________________________________ ________________________________________________________________________ Thought8________________________________________________________________ ________________________________________________________________________ [Please go to the next page]

PAGE 116

104 Q6. Please evaluate how you respond to the following questions. indicates not at all and indicates very much. If you have no reaction, please circle to indicate your neutrality. To what extent do you respond to the following questions? Not at All Very Much (1) How much did you pay attention to read the statement presented by CALCO? 1 2 3 4 5 6 7 (2) How much were you motivated to read the statement presented by CALCO? 1 2 3 4 5 6 7 (3) How much did you make an effort to think of the issue addressed in the statement presented by CALCO? 1 2 3 4 5 6 7 (4) How important are environmental issues in Taiwan to you personally? 1 2 3 4 5 6 7 (5) How important is the deforestation issue in Taiwan to you personally? 1 2 3 4 5 6 7 (6) How much are you concerned about environmental issues in Taiwan? 1 2 3 4 5 6 7 (7) How much are you concerned about the deforestation issue in Taiwan? 1 2 3 4 5 6 7 (8) How often do you think about environmental issues in general? 1 2 3 4 5 6 7 (9) How often do you think about the deforestation issue in particular? 1 2 3 4 5 6 7 (10) How confident are you that you can do something to solve environment issues in general? 1 2 3 4 5 6 7 (11) How confident are you that you can do something to solve the deforestation issue in particular? 1 2 3 4 5 6 7 [Please go to the next page]

PAGE 117

105 Q7. Please evaluate how you feel about the relationship between you and CALCO. Please circle the number that best indicates your agreement with each item. If you strongly disagree with the provided statement, please circle . If you strongly agree with the provided, please circle . If you have no reaction, please circle in the box to indicate your neutrality. To what extent do you agree with the following statements Strongly Strongly Disagree Agree (1) Whenever CALCO makes an important decision, I know it will be concerned about people like me. 1 2 3 4 5 6 7 (2) CALCO can be relied on to keep its promises. 1 2 3 4 5 6 7 (3) I believe that CALCO takes the opinions of people like me into account when making decisions. 1 2 3 4 5 6 7 (4)CALCO has the ability to accomplish what it says it will do. 1 2 3 4 5 6 7 (5) Sound principles seem to guide CALCOs behavior. 1 2 3 4 5 6 7 (6) I can see that CALCO wants to maintain a relationship with people like me. 1 2 3 4 5 6 7 (7) There is a long-lasting bond between CALCO and people like me. 1 2 3 4 5 6 7 (8) CALCO and people like me benefit from their relationship. 1 2 3 4 5 6 7 (9) I feel people like me are important to CALCO. 1 2 3 4 5 6 7 (10) CALCO seems to be the kind of company that invests in the community. 1 2 3 4 5 6 7 (11) CALCO has the ability to attract, develop, and keep talented people. 1 2 3 4 5 6 7 Gender Male [ ], Female [ ] Age [ ] What year are you in the college? [ ] Ethnicity White-Non Hispanic [ ] Hispanic American [ ] African American [ ] Native American [ ] Asian/Pacific Islander [ ] Others (Please specify): [ ] This is the End. I appreciate your participation.

PAGE 118

LIST OF REFERENCES Agresti, A., & Finlay, B. (1997). Statistical methods for the social science (3rd ed.). Upper Saddle River, NJ: Prentice Hall. Albert, M. I., & Golden, L. L. (1982). The impact of education on the relative effectiveness of one-sided and two-sided communications. In B. J. Walker, W. O. Bearden, W. R. Darden, P. E. Murphy, J. R. Nevin, J. C. Olson & B. A. Weitz (Eds.), An assessment of marketing thought and practice (pp. 30-33). Chicago: American Marketing Association. Allen, M. (1991). Meta-analysis comparing the persuasiveness of one-sided and two-sided messages. Western Journal of Speech Communication, 55(4), 390-404. Allen, M. (1994). The persuasive effects of one and two sided messages. In M. Allen & R. W. Preiss (Eds.), Prospects and precautions in the use of meta-analysis (pp. 101-125). Dubuque: IA: Brown & Benchmark. Allen, M., Hale, J., Mongeau, P., Berkowitz-Stafford, S., Stafford, S., Shanahan, W., et al. (1990). Testing a model of message sidedness: Three replications. Communication Monographs, 57(December), 275-291. Anderson, W. T., & Golden, L. L. (1984). Life style and psychographics: A critical review and recommendation. In T. C. Kinnear (Ed.), Advances in consumer research (Vol. 11, pp. 405411). Provo, UT: Association for Consumer Research. Apsler, R., & Sears, D. O. (1968). Warning, personal involvement, and attitude change. Journal of Personality and Social Psychology, 9, 162-166. Belch, G. E. (1981). An examination of comparative and noncomparative television commercials: The effects of claim variation and repetition on cognitive response and message acceptance. Journal of Marketing Research, 18(August), 333-349. Berger, J. (2003). Incorporating ecology and ecological risk into-term stewardship on contaminated sites. Remediation Journal, 13, 107-119. Bettinghaus, E. P., & Basehart, J. R. (1969). Some specific factors affecting attitude change. Journal of Communication, 18, 227-239. Bettinghaus, E. P., & Cody, M. J. (1994). Persuasion communication (5th ed.): Fort Worth: Harcourt Brace College Publishers. 106

PAGE 119

107 Bither, S. W., Dolich, I. J., & Nell, E. B. (1971). The application of attitude immunization techniques in marketing. Journal of Marketing Research, 8, 56-61. Bliss, J. C. (2000). Public perceptions of clearcutting. Journal of Forestry, 12, 4-9. Bliss, J. C., Nepal, S. K., Brooks, R. T., & Larsen, M. D. (1994). Forestry community or granfalloom? Journal of Forestry, 92(6-10). Bohner, G., Einwiller, S., Erb, H.-P., & Siebler, F. (2003). When small means comfortable: relations between product attributes in two-sided advertising. Journal of Consumer Psychology, 13(4), 454-463. Calder, B. J., Phillips, L. W., & Tybout, A. M. (1981). Designing research for application. Journal of Consumer Research, 8(2), 197-201. Carroll, A. B. (1991). The pyramid of corporate social responsibility: Toward the moral management of organizational stakeholders. Business Horizons, July-August, 39-48. Chebat, J. C., Filiatrault, P., Laroche, M., & Watson, C. (1988). Compensatory effects of cognitive characteristic of the source, the message, and the receiver upon attitude change. Journal of Psychology, 122, 609-621. Chebat, J. C., & Picard, J. (1985). The effects of price and message sidedness on confidence in product and advertisement with personal involvement as a mediator variable. International Journal of Research in Market, 2(4), 129-141. Chu, G. C. (1967). Prior familiarity, perceived bias, and one-sided versus two-sided communication. Journal of Experimental Social Psychology, 3, 243-254. Clark, C. E. (2000). Differences between public relations and corporate social responsibility: An analysis. Public Relations Review, 26(3), 363-380. Clark, E. (2002). Management learning and knowledge transfer in transforming societies approaches, issues and future directions. Human Resource Development International, 5(3), 261-277. Cook, T. D., & Campbell, D. D. (1979). Quasi-experimentation: Design & analysis issues for field settings. Boston, MA: Houghton Mifflin Co. Crowley, A. E., & Hoyer, W. D. (1994). An integrative framework for understanding two-sided persuasion. Journal of Consumer Research, 20(4), 561-574. Davis, J. J. (1994). Consumer response to corporate environmental advertising. Journal of Consumer Marketing, 11, 25-47. Davis, K. (1973). The case for and against business assumption of social responsibilities. Academy of Management Review, 16, 312-322.

PAGE 120

108 Davis, K., & Blomstrom, R. L. (1975). Business and society environment and responsibility. New York: McGraw-Hill. Earl, R. L., & Pride, W. M. (1980). The effects of advertisement structure, message sidedness, and performance test results on print advertisement informativeness. Journal of Advertising, 9, 36-44. Ellen, P. S., Wiener, J. L., & Cobb-Walgren, C. (1991). The role of perceived consumer effectiveness in motivating environmentally conscious behaviors. Journal of Public Policy and Marketing, 10(2), 102-118. Esrock, S. L., & Leichty, G. B. (1998). Social responsibility and corporate Web pages: Self-presentation or agenda-setting? Public Relations Review, 24(3), 305-321. Etgar, M., & Goodwin, S. A. (1982). One-sided versus two-sided comparative message appeals for new brand introductions. Journal of Consumer Research, 8(March), 460-465. Fombrun, C. J. (1996). Reputation: Realizing value from the corporate image. Boston: Harvard Business School Press. Ford, L. A., & Smith, S. W. (1991). Memorability and persuasiveness of organ donation message strategies. American Behavioral Scientist, 34, 695-711. Fortune. (2003, June, 23). Features/ Corporate responsibility. Fortune, 98. Frederick, W. C. (1994). From CSR1 to CSR2. Businee and Society, 33(2), 150-164. Garson, G. D. (2003). PA 765 Statnotes: An online texbook, by David Garson. Retrieved March 3, 2005, from http://www2.chass.ncsu.edu/garson/pa765/statnote.htm Goldberg, M. E., & Hartwick, J. (1990). The effects of advertiser reputation and extremity of advertising claim on advertising effectiveness. Journal of Consumer Research, 17(September), 172-179. Golden, L., & Alpert, M. (1987). Comparative analysis of the relative effectiveness of one and two-sided communication for contrasting products. Journal of Advertising, 16(March), 18-25. Gordon, M. E., McKeage, K., & Fox, M. A. (1998). Relationship marketing effectiveness: the role of involvement. Psychology and Marketing, 15(5), 443-459. Gotlieb, J. B., & Sarel, D. (1991). Comparative advertising effectiveness: the role of involvement and source credibility. Journal of Advertising, 20(1), 38. Grunig, J. E. (1986). A new measure of public opinion on corporate social responsibility. Academy of Management Journal, 22(4), 738-764.

PAGE 121

109 Grunig, J. E. (1997). A situational theory of publics: Conceptual history, recent challenges and new research. In D. Moss, T. ManManus & D. Vercic (Eds.), Public relations research: An international perspective (Vol. 3-48). London: International Thomson Business Press. Grunig, J. E., & Hunt, T. T. (1984). Managing public relations. New York: Harcourt Brace Jovanovich College Publishers. Hale, J. L., Mongeau, P. A., & Thomas, R. M. (1991). Cognitive processing of one-and two-sided persuasive message. Western Journal of Speech Communication, 55(Fall), 380-389. Haley, E. (1996). Exploring the construct of organization as source: Consumer's understandings of organizational sponsorship of advocacy advertising. Journal of Advertising, 25, 19-36. Haley, E. (1996). Exploring the construct of organization as source: consumers' understandings of organizational sponsorship of advocacy advertising. Journal of Advertising, 25(2), 19. Halverson, R. R. (1975). Commitment, cognitive structure, and oneversus two-sided persuasive communications. In Dissertation abstracts international (Vol. 35, 6164B-6165B). Hastak, M., & Park, J.-W. (1990). Mediators of message sidedness effects on cognitive structure for involved and uninvolved audiences. Advances in Consumer Research, 17, 329-336. Heath, R. L. (1998). New communication technologies: An issues management point of view. Public Relations Review, 24(3), 273-288. Heath, R. L., & Ryan, M. R. (1989). Public relations role in defining corporate social responsibility. Journal of Mass Media Ethics, 4, 21-38. Heesacker, M., Petty, R., & Cacioppo, J. (1983). Field dependence and attitude change: source credibility can alter persuasion by affecting message-relevant thinking. Journal of Personality, 51(4), 653-666. Hon, L. C., & Grunig, J. E. (1999). Guidelines for measuring relationships in public relations. Gainesville, FL: The Institution for Public Relations. Hovland, C. I., Janis, I. L., & Kelley, H. H. (1953). Communications and persuasion: Psychological studies in opinion change. New Haven: CT: Yale University Press. Hovland, C. I., Lumsdaine, A. A., & Sheffield, R. D. (1949). The effects of presenting "one side" versus "both sides" in changing opinions on a controversial subject. In Experiments on mass communication: Studies in social psychology in World World II (Vol. 3, pp. 201-227). Princeton NJ: Princeton University Press.

PAGE 122

110 Johnson, B., & Eagly, A. H. (1989). Effects of involvement on persuasion: A metaanalysis. Psychological Bulletin, 106, 290-314. Kamins, M. A., & Assael, H. (1987). Two-sided versus one-sided appeals: A cognitive perspective on argumentation, source derogation, and the effect of disconfirming trial on belief change. Journal of Marketing Research, February, 1987, 29-39. Kamins, M. A., Brand, M. J., Hoeke, S. A., & Moe, J. C. (1989). Two-sided versus one-sided celebrity endorsements: The impact on advertising effectiveness and credibility. Journal of Advertising, 18, 4-11. Kline, R. B. (1998). Principles and practices of structural equation modeling. NY: The Guilford Press. Kotler, P., & Armstrong, G. (1999). Principles of Marketing (8th ed.). NJ: Prentice-Hall: Upper Saddle River. Lafferty, B. A., & Goldsmith, R. E. (1999). Corporate credibility's role in consumers' attitudes and purchase intentions when a high versus a low credibility endorser is used in the ad. Journal of Business Research, 44, 109-116. Lang, B., Lee, C. K.-C., & Zwick, R. (1999). Message sidedness at the brand and product form levels: Overcoming the shortcomings of two-sided messages. Advances in Consumer Research, 26, 485-490. Ledingham, J. A., & Brunig, S. D. (1999). Relationships between organizations and publics. Development of a multi-dimensional organization-public relations scale. Public Relations Review, 25, 158. L'Etang, J. (1994). Public relations and corporate social responsibility: Some issues arising. Journal of Business Ethics, 13(2), 111-123. L'Etang, J. (1996). Public relations and rhetoric. In J. L. Etang & M. Pieczka (Eds.), Critical perspective in public relations (pp. 106). London: International Thomson Business Press. Lim, J. S. (2001). Intangible outcomes of corporate strategic philanthropy: A public relations perspective. Paper presented at the annual conference of the Association for Education in Journalism and Mass Communication. Lim, J. S., & Lin, S. Y. (2005, August). One-sided or two-sided? Corporate social responsibility as issues management. Paper presented at the Annual convention of Association for Education in Journalism and Mass Communications, San Antonio, TX. Lumsdaine, A., & Janis, I. L. (1953). Resistance to "counterpropaganda" produced by onesided and two-sided "propaganda" presentations. Public Opinion Quarterly, 17, 311-318.

PAGE 123

111 Mazursky, D., & Schul, Y. (1992). Learning from the ad or relying on related attitudes: the moderating role of involvement. Journal of Business Research, 25(1), 81. McCroskey, J. C., Young, T. J., & Scott, M. D. (1972). The effects of message sidedness and evidence on inoculation against counterpersuasion in small group communication. Speech Monographs, 34, 205-212. McGuire, W. J. (1961). The effectiveness of supportive and reputational defenses in immunizing and restoring beliefs against persuasion. Sociometry, 24, 184-197. McGuire, W. J., & Papageorgis, D. (1961). Effectiveness of forewarning in developing resistance to persuasion. Public Opinion Quarterly, 26, 24-34. Menon, S., & Kahn, B. E. (2003). Corporate sponsorship of philanthropic activities: When do they impact perception of sponsor brand? Journal of Consumer Psychology, 13(3), 316-327. Newell, S. J. (1993). Developing a measurement scale and a theoretical model defining corporate credibility and determining its role as an antecedent of consumers' attitude toward the advertisement. Unpublished Doctoral Dissertation, Florida State University, Tallahassee. Nunn, J. (1998). Laboratory osychology: A beginner's guide: Psychology Press Ltd. O'Keefe, D. J. (1998). Justification explicitness and persuasive effect: A meta-analytic review of the effects of varying support articulation in persuasive messages. Argumentation and Advocacy, 35, 61-75. O'Keefe, D. J. (1999). How to handle opposing arguments in persuasive messages: A meta-analytic review of the effects of one-sided and two-sided messages. In M. E. Roloff (Ed.), Communication Yearbook 22 (pp. 209-249): Sage Publication Inc. Pechmann, C. (1990). How do consumer influences moderate the effectiveness of two-sided messages? In M. E. Goldberg, G. Gorn & R. W. Pollay (Eds.), Advances in consumer research (pp. 337-341). Provo, UT: Association for Consumer Research. Pechmann, C. (1992). Predicting when two-sided ads will be more effective than one-sided ads: The role of correlational and correspondent inferences. Journal of Marketing Research, November, 441-453. Peng, S. (2001, August). Investigating corporate social responsibility: A content analysis of top Chinese corporate Web pages. Paper presented at the annual conference of the Association for Education in Journalism and Mass Communication, Washington DV. Brand Strategy (2002, July 26). Petroleum and principles. Brand Strategy, 16.

PAGE 124

112 Petty, R. E., & Cacioppo, J. T. (1979). Issue-involvement can increase or decrease persuasion by enhancing message-relevant cognitive responses. Journal of Personality and Social Psychology, 37, 1915-1926. Petty, R. E., & Cacioppo, J. T. (1981). Attitudes and persuasion: Classic and contemporary approached. Dubuque, IA: Wm. C. Brown. Petty, R. E., & Cacioppo, J. T. (1984). The effects of involvement on responses to argument quantity and quality: Central and peripheral routes to persuasion. Journal of Personality and Social Psychology, 46(69-81). Petty, R. E., & Cacioppo, J. T. (1986a). Communication and persuasion: Central and peripheral routes to attitude change. New York: Springer-Verlagnc. Petty, R. E., & Cacioppo, J. T. (1986b). The elaboration likelihood model of persuasion. Advances in Experimental Social Psychology, 19, 123-205. Petty, R. E., Cacioppo, J. T., & Goldman, R. (1981). Personal involvement as a determinant of argument-based persuasion. Journal of Personality and Social Psychology, 41, 847-855. Petty, R. E., Cacioppo, J. T., & Schumann, D. (1983). Central and peripheral routes to advertising effectiveness: The moderating role of involvement. Journal of Consumer Research, 10, 135-146. Petty, R. E., Gleicher, F. H., & Jarvis, B. (1993). Persuasion theory and AIDS prevention. In J. B. Pryor & G. Reeder (Eds.), The social psychology of HIV infection (pp. (155-182). NJ: Hillsdale. Petty, R. E., Kasmer, J. A., Haugtvedt, C. P., & Cacioppo, J. T. (1987). Source and message factors in persuasion: A reply to Stiff's critique of the elaboration likelihood model. Communication Monographs, 54(September), 233-249. Petty, R. E., Ostrom, T. M., & T.C. Brock H. (1981). Cognitive responses in persuasion. Hillsdale, NJ: Erlbaum Associates. Post, J. E., Frederick, W. C., Lawrence, A. T., & Weber, J. (1996). Business and society: Corporate strategy, public policy, ethics (8th ed.). New York: McGraw Hill Inc. Priester, J. R., & Petty, R. E. (2003). The influence of spokesperson trustworthiness on message elaboration, attitude strength, and advertising effectiveness. Journal of Consumer Psychology, 13, 408-421. Rhine, R., & Laurence, S. (1970). Ego-involvement, discrepancy, source credibility and attitude change. Journal of Personality and Social Psychology, 16(2), 175-190. Sawyer, A. G. (1973). The effects of repetition and supportive advertising appeals. Journal of Marketing Research, 10(February), 23-33.

PAGE 125

113 Schuhwerk, M. E., & Lefkokk-Hagius., R. (1995). Green or non-green? Does type of appeal matter when advertising a green product? Journal of Advertising, 24(2), 45-55. Schwepker, C. H., & Cornwell, T. B. (1991). An examination of ecologically concerned consumers and their intention to purchase ecologically packaged products. Journal of Public Policy and Marketing, 10(2), 77-101. Settle, R. B., & Golden, L. L. (1974). Attribution theory and advertiser credibility. Journal of Marketing Research,, 9, 181-185. Sherif, C. W., Kelly, M., Rodgers, H. L., Sarup, G., & Tittler, B. (1973). Personal involvement, social judgment, and action. Journal of Personality and Social Psychology, 27, 311-327. Sherif, M., & Hovland, C. I. (1961). Social judgment: Assimilation and contrast effects in communication and attitude change. New Haven. Connecticut: Yale University Press. Shindler, B., List, P., & Steel, B. S. (2003). Managing federal forests: Public attitudes in oregon and nationwide. Journal of Forestry, 91, 36-42. Smith, R. E., & Hunt, S. (1978). Attributional processes and effects in promotional situations. Journal of Consumer Research, 5(December), 149-158. Smith, S. W., Morrison, K., Kopfman, J. E., & Ford, L. A. (1994). The influence of prior thought and intend on the memorability and persuasiveness of organ donation message strategies. Health Communication, 6, 1-20. Sorrentino, R. M., Bobocel, D. R., Z.Gitta, M., Olson, J. M., & Hewitt, E. C. (1988). Uncertainty orientation and persuasion: Individual differences in the effects of personal relevance on social judgments. Journal of Personality and Social Psychology, 55, 357371. Stiff, J. B., & Mongeau, P. A. (2003). Persuasive communication (2 ed.). New York: Guilford. Swinyard, W. R. (1981). The interaction between comparative advertising and copy claim variation. Journal of Marketing Research, 19(May), 75-186. Szykman, L. R., Bloom, P. N., & Blazing, J. (2004). Does corporate sponsorship of a socially-orientated message make a difference? An investigation of the effects of sponsorship identity on responses to an anti-drinking and driving message. Journal of Consumer Psychology, 14(1&2), 13-20. Tannenbaum, P. H. (1967). The congruity principle revisited: studies in the reduction, induction, and generalization of persuasion. Advances in Experimental Social Psychology, 3, 272.

PAGE 126

114 Trent, J., & Greer, J. (2001, August). Do new media messages mitigate the effect of corporate environmental Ads? A test of source credibility and message balance. Paper presented at the annual conference of the Association for Education in Journalism and Mass Communication, Washington D.C. Varadarajan, P. R., & Menon, A. (1988). Cause-related marketing: A coalignment of marketing strategy and corporate philanthropy. Journal of Management, 52, 58-74. Wartick, S. L., & Cochran, P. L. (1985). The evolution of the corporate social performance model. Academy of Management Review, 10, 758-769. Wartick, S. L., & Rude, R. E. (1986). Issues management: Corporate fad or corporate function. California Management Review, Fall(1), 124-140. Wood, D. J. (1991). Corporate social performance revisited. Academy of Management Review, 16(4), 691-718.

PAGE 127

BIOGRAPHICAL SKETCH Shu Yu Lin is a masters candidate for the Master of Arts in Mass Communication degree focusing on public relations. She was born in Taiwan and received her B.S. degree in Economics and a certification of marketing from National Chengchi University in 2002. Her research interests were particularly focused on issues management, corporate social responsibility campaigns, and integrated marketing communication. 115


Permanent Link: http://ufdc.ufl.edu/UFE0010812/00001

Material Information

Title: A Public Relations Campaign of Corporate Social Responsibility: A Test of a Cognitive Processing Model of a CSR Message
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0010812:00001

Permanent Link: http://ufdc.ufl.edu/UFE0010812/00001

Material Information

Title: A Public Relations Campaign of Corporate Social Responsibility: A Test of a Cognitive Processing Model of a CSR Message
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0010812:00001


This item has the following downloads:


Full Text












A PUBLIC RELATIONS CAMPAIGN OF CORPORATE SOCIAL
RESPONSIBILITY: A TEST OF A COGNITIVE PROCESSING MODEL OF A CSR
MESSAGE















By

SHU YU LIN


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ARTS IN MASS COMMUNICATION

UNIVERSITY OF FLORIDA


2005

































Copyright 2005

by

Shu Yu Lin

































This thesis is dedicated to my father and my mother. Without their love and support, I
could not have accomplished this study.















ACKNOWLEDGMENTS

The accomplishment of this thesis could not have realized without the help of many

people. First of all, I want to express deep appreciation for my parents, Ji-Shun Lin and

Pi-Hua Su, for their endless love and support. Their beliefs in me had given me strengths

throughout the difficulties in completing the degree and the thesis. They have given me

more than I could ever acknowledge.

I would like to thank my committee chair, Dr. Spiro Kiousis, and my committee

members, Dr. Margarete Hall and Dr. Mary Ann Ferguson, for their support of carrying

out this research study. I also want to express my gratitude to Dr. Lynda Kaid, Dr.

Amanda Ruth, and Dr. Tracy Irani for their generous help to accommodate the

experiment in their class periods.

I would like to attribute special thanks to Joon Soo Lim for his generous help and

insightful suggestions. I am especially grateful for the opportunity to work together with

him for a paper submitted to the conference of AEJMC. Without him, I could not have

hoped to accomplish such a challenging and creative experimental study like this.

I like to thank my friends at UF who have lifted my spirits during the stressful time

of writing the thesis: Chin-Hsin Liu, Wan Ping Chao, Yimin Wung, Yang Hsin Hsu,

Chen-Hsuan Chen (Judith), Ting Bing Wu, and Allan. In particular, I want to thank all

the friends who participated in the pilot study and provided useful suggestions. In

particular, I would like to show my appreciation to Camelia Baluta for her generous help

of editing my thesis.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES ................................ .......... .. .... .. .... .............. viii

LIST OF FIGURES ............................... ... ...... ... ................. .x

ABSTRACT .............. .................. .......... .............. xi

CHAPTER

1 IN TR OD U CTION ............................................... .. ......................... ..

Corporate Social Responsibility Communication ....................... ................ ........... 1
CSR Communication from the Issues Management Perspectives.............................2
State ent of the Purpose ......................................................... ................ ............ 4
B background ................................... ........................... ..... ...... ....... 4
R research Q uestions........... .................................................................... ........ .. .. ...
T heoretical Fram ew ork .......... ..... ........................................................ ................ .5

2 LITER A TU R E REV IEW ............................................................. ....................... 8

Corporate Social Responsibility ...........................................................................8
The Origin of CSR ........................................... ............. .... ......... 9
C SR D definition ............................................................. .......... .... 10
Perceived CSR and Public Relations ........................................ .................11
C SR and Issues M anagem ent ...................................................................... .... 12
Corporate Credibility ........................................... .... ..................... 13
The Elaboration Likelihood M odel ........................................ ........................ 15
The Theoretical Framework of the ELM ................ ..................................16
Central Route and Peripheral Route..... ....................... ..............17
The Variable of Personal Relevance/Involvement ........................ ............. .18
Involvement and Environmental Advocacy ................................... ............... 19
M message Sidedness Effect .............. ................ ......... .. ..... ............... 20
Definitions ..... ........... ......... .................. 20
Origins ......................... .............................21
Previous Studies ..........................................22
Advertising and Non-Advertising Messages ...............................................22
Prior Knowledge, Motivation, Personality ............................................. 23


v










Issue Involvem ent.............. .. ....... ............ .................... 24
Three Theoretical Accounts for the Message Sidedness................................25
Inoculation Theory ...................................... ................... ..... .... 26
C correspondence T heory ............................................................ .....................26
Cognitive R response Theory ........................................ .......................... 28
H ypotheses ................................................. 30

3 M E T H O D O L O G Y ............................................................................ ................... 32

A n Experim ental D esign.................................................. ............................... 33
P ilo t S tu d y ................................................................3 5
Procedure ................................. .......................... .. .... ......... 35
Pilot Test Results ................................. ........................... ... .......37
M ain Stu dy ........................................................................................... 3 8
Sam ple ................................. .......................... .... ..... ......... 38
S tim u li ........................................................................................................... 3 9
Procedure .....................................................41
Operational Definitions: Independent Variables ........................... ...............42
Involvement ................ .. ............. ..........42
Message sidedness..........................................43
Operational Definitions: Dependent Variables ...................................... 45
Cognitive elaboration ........................... ........ .. ....... 45
Perceived CSR................ ........ ........ .........45
A ttitu d e s ................................................................................................. 4 6
M manipulation Check ................................................ ............... 47
Involvement ................ .. ............. ..........47
M message sidedness.................................................. 48

4 FINDINGS ............... ............................... ...............49

Profile of Participants ......................... ......... .........49
Results of M manipulation Checks ........................................................................ .. ......50
Involvement .......... .. ............ ...................50
Message Sidedness ................... ....................51
Reliability Checks for Dependent Measures ........................................52
Test of Hypotheses .................. ............. .......... 53
Cognitive Elaboration............................... ........... .......... 53
Main Effects of Message Sidedness on Perceived CSR ..............................54
Interaction Effect on Perceived CSR ..................................... ............... 56
Attitudes toward a Company .................. ............. ..... ....... 58
Interaction Effect on Attitude toward a Company ...........................................60
A sub Analysis between Two Sample Groups ........................... ............... 62
A Correlation between Perceived CSR and Attitudes toward a Company .........62
A Hierarchical Regression Analysis between Perceived CSR and Attitudes
tow ard a C om p any ..................................................................................... 6 3
P ath A n aly sis .........................................................................................6 6









5 D ISC U S SIO N ............................................................................... 7 1

Over V iew of the Present Study ...................................................... ................71
Overview of the Hypotheses and Research Question...............................................74

6 C O N C L U SIO N ......... ......................................................................... ......... .. .... 1

Discussions about the Effect of Message Sidedness .............. ..... ............... 82
Credibility Perceptions .............................. ..... ... ...... .. .... ............. 83
Pre-A attitudes tow ard an Issue......................................... ......................... 84
F ear A rou sal ................................................................8 5
A p p lic a tio n ........................................................................................................... 8 5
L im station s ...................................................................................... ..................... 86
Suggestions for Future Research ........................................................ 87
K n ow led g e L ev el ........................................................................................... 8 7
Message Length .............. .................... ......... 87
Source Credibility .............. ......... .......... ......... 88
A ttitu des tow ard Issu es.................................................................................. 88

APPENDIX

A CALCO "ABOUT US"- UNITED STATES ........................................ ....90

B CALCO "ABOUT US"- TAIWAN VERSION ........................................ ...91

C CALCO "ENVIRONMENTAL COMMITMENT"-UNITED STATES ................... 92

D CALCO "ENVIRONMENTAL COMMITMENT"-TAIWAN ............................. 93

E CALCO "M YTH & FACT" ................................................................................ 94

F CALCO "ENVIRONMENTAL COMMITMENT"-NONREFUTATIONL ............95

G QUESTIONNAIRE FOR HIGH INVOLVEMENT GROUPS ..............................96

H QUESTIONNAIRE FOR LOW INVOLVEMENT GROUPS ..............................101

L IST O F R E F E R E N C E S ........................................................................................... 106

BIOGRAPHICAL SKETCH .............................................................. ...............115
















LIST OF TABLES


Table page

2-1 Factorial design of the experim ent................................................ ........ ....... 31

3-1 Liker-type items of the manipulation check of involvement level for high
involved ent groups .................. ........................................ .. ...... .... 36

3-2 Liker-type items for the manipulation check of message sidedness ........................37

3-3 The conditions of the 2x3 experimental design ....................................... .......... 39

3-4 Liker-type item s for perceived CSR ..................................... ........................ 46

3-5 Semantic differential type items for attitude toward the company ........................47

4-1 The number of participants in each cell ............ ............ ...................... ............... 50

4-2 The manipulation check of involvement..... ...................... ............51

4-3 The manipulation check of message sidedness................................ ... ................ 52

4-4 A t-test of mean difference of message-related thoughts generated between the
high and the low involvement group........ .......................................... 53

4-5 A t-test of mean difference of message-related thoughts generated between three
group s of sided m messages ........................................ ......................................... 54

4-6 Perceived CSR by message sidedness .......................................... ...............55

4-7 F-test of perceived CSR by message sidedness ............. ....................................... 55

4-8 Perceived CSR by message sidedness .......................................... ...............56

4-9 Interaction of involvement and message sidedness on perceived CSR ................57

4-10 Tests of between-subject effects (dependent variable: perceived CSR) ..................57

4-11 Attitudes toward CALCO by message sidedness ......... ......................................59

4-12 F-test for attitudes toward CALCO by message sidedness................................. 59









4-13 Attitudes toward CALCO by message sidedness......... ..............................60

4-14 Interaction of involvement and message sidedness on attitudes............. ..............61

4-15 Tests of between-subj ect effects (dependent variable: attitudes)..........................62

4-16 A Pearson Bivariate Correlation between perceived CSR and attitudes..................63

4-17 Hierarchical regression analysis for proposed model ...........................................64
















LIST OF FIGURES


Figure pge

1-1 The cognitive processing model of a CSR message proposed in this study .............7

4-1 Interaction of involvement and message sidedness on perceived CSR ................58

4-2 The original cognitive processing model of a CSR message proposed in this
re se a rc h ....................................................................... 6 9

4-3 The path diagram based on the path analysis. ................. ............................... 69

4-4 The revised cognitive processing model of a CSR message proposed in this
re se a rc h ....................................................................... 7 0















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Arts in Communication

A PUBLIC RELATIONS CAMPAIGN OF CORPORATE SOCIAL
RESPONSIBILITY: A TEST OF A COGNITIVE PROCESSING MODEL OF A CSR
MESSAGE

By

Shu Yu Lin

August, 2005

Chair: Spiro Kiousis
Major Department: Journalism and Communications

Little scholarship has examined the effectiveness of CSR campaigns in influencing

attitude toward the company. The purpose of this present study is to measure the impact

of CSR communication by proposing a cognitive processing model of a CSR message.

According to the Elaboration Likelihood Model (ELM), it is assumed in this study that

recipients' perceived CSR and attitudes toward a company can be influenced by CSR

campaigns in terms of the effect of message sidedness (one-sided, refutational two-sided

and nonrefutational two-sided) and message recipients' involvement level (high or low).

A mixed 2x3 factorial design is applied for this experiment in which involvement

and message sidedness are independent variables, while attitude toward the company and

perceived CSR are the dependent variables. Two hundred and five students from the

University of Florida participate in this study and randomly assigned to one of the seven

conditions. A CSR campaign advocating the corporate social performance of a factitious









lumber company named CALCO is created and presented on the printed Web pages of

CALCO. A controversy over CALCO's practice of clearcutting, which is fundamental to

the survival of the timber industry, is discussed in the stimuli.

Results of the present study show that the one-sided CSR message is significantly

more persuasive than the refutational two-sided and the nonrefutational two-sided one.

Two main effects are found between message sidedness on both perceived CSR and

attitude toward the company. These results indicate that the one-sided CSR campaign has

a significant advantage over the refutational two-sided and the nonrefutational two-sided

one in influencing participants' perceptions about the company's social performance and

their attitudes toward the company.

In addition, the path analysis indicates a significant direct effect between perceived

CSR and attitude toward the company. The findings of this study show that the perceived

CSR is highly associated with the attitude toward the company. In other words, when a

company's social performance is highly perceived, it is likely that people will generate

more positive attitudes toward the company. These results demonstrate the value of CSR

communication as part of a business strategy.

When a company attempts to communicate its corporate social performance, the

results of this study suggest that the one-sided CSR campaign simply presenting

supportive information related to the company's corporate social activities has more

persuasive effects than the two-sided CSR campaigns in influencing public perceptions

about the company's social commitment and generating more positive attitude toward the

company.














CHAPTER 1
INTRODUCTION

Corporate Social Responsibility Communication

Understanding the effects of communicating corporate social responsibility (CSR)

on public perceptions and attitudes toward the company can create a good opportunity for

public relations practitioners to demonstrate the importance of PR function in the

corporate world. Communication research has looked specifically into how best to

communicate the social responsibility of an organization (Wartick & Cochran, 1985;

Wartick & Rude, 1986; Heath & Ryan, 1989; Carroll, 1991; Wood, 1991; Frederick,

1994; L'Etant, 1994; Esrock & Leichy, 1998; Heath, 1998; Clark, 2000). However, little

of previous research on public relations has empirically tested the effects of CSR

campaigns on people's cognitive responses and attitudes.

The concept of CSR is not new. It has been widely discussed both in the business

and marketing field. More and more CEOs of Corporate America started to talk about a

business obligation to society that goes beyond making profits. In this emerging

corporate philosophy, contemporary corporations are expected to be both doing good in

terms of social performance and doing well in terms of business performance. With an

increasing emphasis on corporate social performance, thus, today's business

organizations tend to advocate CSR not only to gain measurable outcomes but also to

gain intangible outcomes (Lim, 2001). Corporate executives now believe that socially

responsible companies will outperform their peers by enhancing reputations, reducing

risks, and seizing new opportunities. An emerging agenda for today's CEOs is to bring









the economics together with the environmental and the social needs as part of a business

strategy.

There haven been a lot of discussions from marketing literature regarding the

effects of CSR initiatives on achieving organizational goals such as increasing the

organizational bottom-line or boosting shareholder value. However, any attempt to

account for CSR as part of a business strategy needs to address a communication

management perspective of corporate social responsibility. In this regard, an important

research topic for public relations researchers is how to communicate CSR effectively.

To achieve excellence in public relations campaign, a good communication strategy

should engage a good defense as well as offense. Then, presenting only a CSR message

should be considered to be a defense whereas managing an issue with the CSR message

could be an offense. In other words, should a company present simply CSR messages or

utilize the CSR campaign as an opportunity to address a relevant environmental or a

social issue?

This thesis attempted to investigate the best model for the CSR campaign in terms

of issues management.

CSR Communication from the Issues Management Perspectives

The power of information technology allows today's public to scrutinize corporate

performances in detail. Through the Internet, individuals are able to access all kinds of

information from a variety of sources and perspectives. The convenient accessibilities of

mass information of corporate performances allow people to actively discuss social issues

or environmental concerns related to corporate operations around the world. Irresponsible

corporate actions are likely to induce criticisms from external stakeholders such as the

media, governmental authorities, communities or even activists groups that organize









boycotts against the company. For example, Shell placed its 100 years of brand building

and reputation in jeopardy when the company was accused of the environmental damages

and the neglect of human rights in Nigeria in 1994 (Brand Strategy, 2002).

Due to this changing business environment, more corporations have woken up to

the environmental and social expectations of today's society. For instances, to reduce

emissions created by daily business operations, FedEx converted all its trucks to hybrid

electric-diesel engines while UPS also included 1, 800 alternative-fuel vehicles. To

ensure biodiversity, Starbucks is buying more organic and shade-grown coffee that

minimizes damages of rain forests (Fortune, 2003).

In academia, a sustainable body of literature has discussed CSR from the issues

management perspective (Wartick & Cochran, 1985; Wartick & Rude, 1986; Heath &

Ryan, 1989; Carroll, 1991; L'Etant, 1994; Esrock & Leichty, 1998; Heath, 1998; Clark,

2000;). According to Lim and Lin (2005), some researchers have specifically looked into

how CSR could play an active role in dealing with issues that are critical to

organizational survival (Heath & Ryan, 1989; Frederick, 1994; L'Etant, 1994; Esrock &

Leichty, 1998; Heath, 1998).

An abundant number of studies have conceptualized CSR in terms of issues

management. However, little research has empirically delved into the communication

management perspective of a CSR campaign. Moreover, a few studies that could be

identified as research on communication management have used the concept of CSR very

narrowly as either cause-related marketing or advocacy advertising (Lim & Lin, 2005, for

a full list of literature).









Statement of the Purpose

The purpose of the present study is to fill this research gap by proposing an

effective CSR communication strategy from an issues management perspective. This

study intends to contribute to a better understanding of theoretical and practical

implications of CSR communication. In this study, two communication goals are

included in the CSR campaigns. First of all, a CSR message aims to inform targeted

audiences about the corporate good citizenship. Secondly, in response to external

stakeholders' claims, a CSR message also addresses corporate positions or statements of

controversial issues that are important to the company's fundamental practices. It is

proposed in this study that communicating corporate social performance will eventually

influence people's attitude toward the company. Moreover, it is assumed in this study

that the persuasiveness of CSR campaigns depends on the effect of message sidedness.

Some companies may choose to simply present positive messages that support their

social and environmental initiatives, while others may mention additional

counterarguments from environmental groups.

Background

In this study, the topic of the experimental stimuli is the clearcutting issue that is

critical to the survival of the timber industry. Clearcutting is a management technique in

which all of the trees in an area are cut at the same time. Clearcutting is commonly

applied by many lumber companies because it is cost effective and efficient.

Bliss (2000) has addressed a problem of clearcutting issue. He argued that there

had been a perceptual gap between the public view and the industry's view. He

emphasized that corporate communication should play an important role in restoring

public trust on the clrearcutting issue. Especially, it was emphasized the social and









environmental performance of the timber industry so that it influences the dominant

beliefs and values among the public. It is proposed in this present study that CSR

campaigns might help to reduce the perceptual gap of clearcutting issues between the

lumber company and the general public.

Research Questions

In this study, a dilemma in communication strategy is discussed. In dealing with

controversial issues, should companies refute the counterarguments or not? This study

examines the effectiveness of communication strategies in influencing individuals'

perceptions of the company's social performance and their attitude toward the company.

From the persuasion perspectives, should the companies present one-sided CSR

campaigns that are simply describing the social performance of the company? Or, should

corporations design refutational two-sided CSR campaigns that include opposite claims

from activists groups but follow with refutational arguments emphasizing the advocated

positions? Moreover, companies could also consider the nonrefutational two-sided CSR

campaigns presenting both supportive and opposite arguments without refutational

arguments.

In order to address these questions, an experimental study of the CSR campaigns

was conducted to test empirically which strategy is more effective in influencing people's

attitudes and perceptions

Theoretical Framework

Richard Petty and John Cacioppo's (1981, 1986a, 1986b) Elaboration Likelihood

Model (ELM) is an attitude formation and change model that can be used to explain how

publics evaluate persuasive communication. Based on the ELM, two previously used









variables were employed in this study: involvement (high and low) and message

sidedness (one-sided, refutational two-sided and nonrefutational two-sided).

A cognitive processing model (see Figure 1-1) of CSR messages is proposed in this

study. In this model, the researcher assumes that CSR messages will influence the

audience's perception of the company's social performance and attitude toward the

company depending on the message sidedness effect (one-sided, refutational two-sided

and nonrefutational two-sided) and the audience's involvement level (high and low)

The level of subjects' involvement is manipulated in the booklet, directing them to

either browse through the presented message or to carefully read and think about the

message. Consistent with previous research, it is assumed in this study that those who are

highly involved will induce more cognitive elaboration.

The hypotheses of the current research were posited to address the main effects and

the interaction effects of message sidedness and involvement on perceived CSR and

attitudes toward the company advocated. The main effects of both involvement and

message sidedness are that high involvement and two-sided messages will induce more

cognitive elaboration than low involvement and one-sided messages.

The interaction effects predict that when highly involved people are exposed to

refutational two-sided messages, they will be more influenced by the message and create

more positive attitudes toward the company. On the other hand, when involvement is

low, people will be more influenced by a one-sided message in their evaluation of

perceived CSR and attitude toward a company. Nonrefutational two-sided messages are

said to have the least persuasiveness in terms of influencing attitudes toward the company

and perceived CSR in both high and low involvement conditions.









Based on the ELM, this study empirically tests the persuasive effects of CSR

messages on people's perceptions of corporate social performance and attitude toward the

company. This study intends to contribute to a better understanding of the theory and the

implications of CSR communication. In essence, the studies of ELM and CSR can be

strengthened and supported by connecting the persuasive effects of communicating CSR

on publics' attitude.


Figure 1-1. The cognitive processing model of a CSR message proposed in this study.














CHAPTER 2
LITERATURE REVIEW

Corporate Social Responsibility


Corporate social responsibility (CSR) has been widely discussed in terms of a

managerial function in business organizations. CSR has long been referred to the

managerial obligation to improve both the societal welfare and a company's enlightened

self-interest (Davis & Blomsstrom, 1975). A study of how U.S. corporate entities make

use of the Web to present themselves as socially responsible citizens showed that more

than 80% of a random sample from Fortune 500 companies had Web pages that

addressed at least one corporate social responsibility issue (Esrock & Leichty 1998). This

trend has reflected the pervasive belief among business leaders that CSR is an economic

imperative in today's global marketplace.

CSR is important for the public relations function in organizations. James Grunig

and Todd Hunt (1984), two well-known communication scholars, devoted a whole

chapter in their book to public relations and public responsibility. They argued that

"Public, or social responsibility has become a major reason for an organization to have a

public relations function and that the two-way symmetrical communication is the best

means by which to evaluate social responsibility." (p. 48) Through CSR programs, public

relations not only gets involved in the advisory management role by building

relationships with key stakeholders but also engages in communicational management by

delivering messages to target publics (L'Etang, 1994).









The Origin of CSR


The idea of CSR appeared around the turn of the 20th century (Post, Frederick,

Lawrence and Weber, 1996).Clark's (2000) comprehensive review of CSR and corporate

public relations found similarities of these two areas in the origins, theories, processes,

and primary responsibilities. Both public relations and CSR struggled to define their

disciplines beginning primarily in the 1920s (Clark, 2000, p.366).

Although a large number of today's business still think of CSR as paternalistic in

terms of charitable contributions, researchers argue that there is another form of CSR

from the stewardship perspective (Post, Frederick, Lawrence and Weber, 1996).

According to the stewardship principle, corporations become stewards or public trustees

by using their resources to affect all people in society in fundamental ways. This

principle led to modern stakeholder theory that argued that managers should recognize

the need to interact meaningfully with all groups who have a stake in the organization's

activities.

In the 1960s and 1970s, CSR underwent some of its most important iterations.

Wood (1991) pointed out the emerging importance of CSR in her conceptual

development of corporate social performance.

"It became apparent during this time, particularly through social activism and

regulatory activity, that social expectations of business had outstripped managers'

comprehension and capabilities." (p.383)

Thus, scholars and managers began to define what corporate social responsibilities

were and were not. In the 1970s, studies about CSR were based on the idea that business

is an "actor in the environment and should respond to social pressures and demands."









(Wood, 1991, p.384) Another classic definition also pointed out the broad responsibilities

of a modem corporation which defined CSR as "the firm's consideration of and respond

to issues beyond narrow economic, technical, and legal requirements of the firm." (Davis,

1973)

By the early 1980s, there is a slight shift of the research in CSR from the

conceptual idea that companies should be responsible for a more practical approach to

addressing how companies could respond to business-related social issues (Clark, 2000).

CSR Definition


Since the inception of CSR research, the concepts have been plagued by ambiguity.

A long and diverse history has been associated with the evolution and the definition of

CSR. According to Wood's (1991) conceptual development of CSR theory, she included

three main principles: (a) the principle of legitimacy at the institutional level, (b) the

principle of public responsibility at the organizational level, and (c) the principle of

managerial discretion at the individual level.

In particular, Wood (1991) articulated that businesses are not responsible for

solving all social problems. However, they are responsible for solving problems that they

have caused and for helping to solve other social problems related to their business

operations (p.697). For example, a lumber company is rightly held responsible for

keeping a forest sustainable. However, it might be harder for such a company to support

other social causes that are irrelevant to their practices, such as AIDS prevention or adult

literacy. Social responsibilities should be relevant to the firm's interests, operations, and

actions (Wood, 1991, p.698). But this principle leaves room for managerial discretion in









determining what social problems and issues are relevant and how those issues should be

addressed

To sharpen the definition of CSR, Carroll (1991) proposed a pyramid that

constituted total CSR from four dimensions of social responsibilities including economic,

legal, ethical, and philanthropic. It is worth noting that Carroll's (1991) pyramid of CSR

begins with the economic performance. According to this principle, if a company is not

making profits and is not providing high quality of goods and services to meet

consumers' needs, it can not be considered socially responsible even when the company

has devoted many efforts in social causes.

Carroll (1991) pointed out that CSR includes philanthropic contributions but is not

limited to them. In other words, philanthropy is highly desired and prized but actually

less important than the other three categories of social responsibility. Carroll argued that

corporate managers and public relations practitioners should be aware of the distinctions

between philanthropy and corporate responsibility. More specifically, philanthropic

responsibilities do not dominate the definitional construct of CSR. While developing

CSR initiatives, corporate managers should pay more attention to other dimensions of

CSR rather than only focusing on philanthropy.

Perceived CSR and Public Relations


From the public relations perspective, perceived CSR may be considered a more

appropriate term in discussing the strategy of communicating corporate social

performance as a function of issues management. This is because the public perception of

corporations' social performance is not always predictable. In other words, the same

message regarding a company's CSR activities can be perceived differently by different









publics. As previous research on persuasion has pointed out, some moderators such as

message recipient's involvement or prior attitude toward the issue or the company can

influence a recipient's perceptions about the company's social performance.

Therefore, more research regarding effective communication methods is needed.

Despite the importance of communication management in CSR, effective communication

methods are largely absent from the corporate social responsibility literature (Clark,

2000). A few studies about the effectiveness of CSR campaigns come from marketing

and consumer research (Varadarajan & Menon, 1988; Menon & Kahn, 2003; Szykman et

al., 2004). Moreover, previous discussions about CSR messages have merely considered

perceived CSR a function of cause-related marketing (Varadarajan & Menon, 1988) or

advocacy advertising of social issues (Haley, 1996; Menon & Kahn, 2003; Szykman et

al., 2004).

CSR and Issues Management


Issues management is a function dedicated to helping organizations understand and

strategically adapt to their environment through issue scanning, tracking, and monitoring

(Heath, 1998).

Heath (1998) further pointed out that a rhetorical approach to issues management

assumes that organizations sometimes do not keep up with key public's standards-their

expectations of one or more aspects of the organization's performance (p.275). Along this

research line, the function of issues management is considered parallel to the function

public relations in organizations.

CSR or corporate social performance (CSP) in literature has often been discussed

as issues management (Wartick & Cochran, 1985; Wartick & Rude, 1986; Heath, 1998;









Wood, 1991; Clark, 2002). Carroll (1991), who contributed to the definition of CSR, also

took a perspective of stakeholder management by which CSR was considered as a

managerial function that helped corporate management to achieve organizational goals.

Wartick and Cochran (1985) proposed issues management as the third facet of the

corporate social performance (CSP) model which stated that CSP is "the integration of

principles of social responsibility, processes of social responsiveness, and policies

developed to address social issues." (p.758) CSR is considered as a process of corporate

social responses to external stakeholders (Wood, 1991; Caroll, 1991).

In this study, the CSR campaign was designed based on a lumber company. For the

lumber industry, the long-term management of environmental responsibility has been a

major issue. In particular, issues related to the practice of clearcutting have long been

debated publicly (Berger, 2003). From the issues management perspective, Berger (2003)

suggested that a lumber company deal with long-term stewardship and ecological

considerations as major driving forces that are very important to the public. Berger

(2003) further argued that a lumber company not only needs to understand public

perceptions about clearcutting but also address that issue within a context of the structure

and functioning of the ecosystem. Therefore, the clearcutting issue is considered an

particularly appropriate topic for lumber companies to address and discuss when

developing CSR campaigns.

Corporate Credibility


CSR can be adopted by companies as part of business strategies in order to enhance

their corporate image and credibility. The reputation of a company that produces products

has been identified as a type of source credibility in marketing research (Goldberg and









Hatiwick, 1990). In addition to having a credible endorser represent their brand,

companies are also concerned with their corporate credibility. This is evident because of

the widespread use of public relations campaigns and prevalence of institutional and

corporate advertising (Fombrun, 1996). Companies use this type of promotional effort

primarily to associate themselves with positive environmental and social issues to

enhance their reputation, and also hope of increasing sales (Kolter and Armstrong, 1996).

A few empirical studies in the field of advertising effectiveness have examined the

impact corporate credibility has on three principal outcome variables: attitude toward the

ad, attitude toward the brand, and purchase intentions. Lafferty and Goldsmith (1999)

found that both endorser and corporate credibility influence attitude-toward-the-ad and

attitude-toward-the-brand, but corporate credibility alone appears to have a significant

influence on purchase intentions. Moreover, whereas endorser credibility seems to have a

greater influence on attitude-toward-the-ad, corporate credibility seems to have a greater

influence on attitude-toward-the-brand. In other words, it is suggested that corporate

credibility can play an influential role on forming people's attitude toward the brand.

However, little empirical research is available on corporate credibility. Although

less attention has been given to this concept in the marketing literature, there is reason to

believe that high corporate credibility is also important in producing positive attitude

change toward the ad and toward the brand, as well as influencing purchase intentions

(Fombrun, 1996).

Goldberg and Hartwick (1990), Newell (1993), and Fombrun (1996) indicated that

subjects are influenced by the credibility of the company when formulating their attitude-

toward-the-ad and toward the brand as well as purchase intentions. The effect size show









that the endorser's credibility appeared to have a stronger impact on the subjects when

they evaluated the advertisement, and the corporate credibility appeared to have a

stronger impact when the subjects assessed the brand.

In the case of high corporate credibility, when the brand attributes are missing, the

reputation of the firm may give the consumers more confidence that the product is good

and make them significantly more willing to purchase the brand. Research findings are

quite consistent with previous research. Newell's (1993) results indicated that perceived

corporate credibility is positively associated with purchase intentions, and Davis (1994)

found that an overwhelming majority of consumers have stated that their product

purchase decisions are at least in part influenced by their view of the parent company's

good citizenship.

Previous research in marketing and advertising has indicated that corporate

credibility and corporate reputation can be considered a function of branding that

influences consumers' attitudes toward the brand and purchase intentions. Some studies

in marketing research also showed positive associations between corporate credibility and

corporate philanthropy. According to previous research, it is assumed that corporate

credibility may be enhanced by the company's corporate social performance. Therefore,

people's attitude toward the company may also be positively influenced though effective

CSR campaigns.

The Elaboration Likelihood Model


Persuasion and attitude change have long been the focal point of research with the

Elaboration Likelihood Model (ELM) (Petty and Cacioppo, 1986a). The ELM provides









"a fairly general framework for organizing, categorizing, and understanding the basic

processes underlying the effectiveness of persuasive communications" (p. 125).

The standard paradigm of ELM research has been to test it with two to three

variables, such as involvement, source credibility or argument strength. Typically,

involvement is used as a motivational indicator of the route a person will use to process a

message (central or peripheral route). According to the ELM, a highly involved person

scrutinizes the arguments presented in the message and bases his or her attitudes toward

the message and the message sponsors on the strengths of the arguments. In contrast, a

person who is uninvolved when processing the message will use simple heuristics cues to

form his or her attitudes toward the message (Petty and Cacioppo, 1986a). Past research

has shown that the ELM is effective in explaining attitude change under various

conditions and with certain accuracy.

In this study, the researcher adopted two variables: involvement (high and low) and

message sidedness (one-sided, refutational two-sided and nonrefutational two-sided). A

review of previous studies based on the ELM will explain how each variable works and

interacts in developing a persuasive message.

The Theoretical Framework of the ELM


The ELM is based on the conception that receivers will vary in the degrees to

which they are likely to engage in the elaboration of information relevant to the

persuasive issue. The term elaboration in the ELM refers to "the extent to which a person

thinks about issue-relevant arguments contained in a message" (Petty & Cacioppo,

1986b, p. 128). Such elaboration may be viewed as a continuum of commitment of

cognitive resources which ranges from no thought to "complete elaboration of every









argument and complete integration of these elaborations into a person's attitude schema"

(p.129).

Central Route and Peripheral Route


According to the ELM (Petty & Cacioppo, 1981), the central route is theorized to

occur under conditions where the message receiver is both motivated (e.g. due to the

personal relevance of the issue) and has ability to process message content. When a

message is processed through the central route, persuasion "likely resulted from the

information presented in support of an advocacy" (Petty & Cacioppo, 1986b, p. 125), and

a high level of elaboration and cognitive responses will be fostered and ultimately lead to

a significant impact on the message receiver's attitudes to the communicated issue.

In contrast, unmotivated or cognitively unable individuals follow a peripheral route

to persuasion where elaboration is relatively low. The message receiver is theorized not

to focus on the primary arguments of the message presented, but instead to focus on

various peripheral cues, such as the communicator's credibility or attractiveness as guides

to attitude and beliefs. While both routes are persuasive, the literature on the ELM

suggests that the central route produces more enduring results (Petty and

Cacioppo,1986a, 1986b).

The ELM was applied in the current study, because it is comprehensive in outlining

the multiple roles by which the variables selected (message sidedness and involvement)

might impact persuasion. The central route to persuasion in this study was the focus of

personal relevance to environmental issues the clearcutting issue. If message receivers

consider environmental issues highly personally relevant, according to the ELM, they

will pay more attention and scrutinize the messages presented. In other words, it is









predicted that message receivers who consider the environmental issues highly relevant

will engage in the central route in which they give thoughtful consideration to issue-

relevant information presented in this study. On the other hand, it is predicted that

message receivers will not pay attention to the issues discussed in the message if they

have low involvement with environmental issues. In that case, message receivers are

likely to engage in the peripheral route and pay little attention the message.

The Variable of Personal Relevance/Involvement


In the ELM, the most important variable affecting the motivation to process a

persuasive message is the personal relevance of the advocacy (Petty & Cacioppo, 1986b,

p81). Petty and Cacioppo regard personal relevance as the extent to which the advocacy

has "personal meaning" (Sherif et al., 1973). Petty and Cacioppo (1979) have termed

their use of involvement as issue involvement, which is the "extent to which the

attitudinal issue under consideration is of personal importance" (p. 1915).

In the current study, the researcher applied Petty and Cacioppo' s (1981) definition

of involvement as "the amount of cognitive processing and/or interest the person has in

the stimulus or attitude object being attended to." The ELM argues that when

involvement is high, people are likely to use more cognitive processing and induce more

message related thoughts, thus engaging the central route. Many researchers found

consistent expectations when a given issue becomes increasing personally relevant to the

message receivers, their motivations of engaging in thoughtful considerations of that

issue presumably increase (Sherif & Hovland, 1961; Petty & Cacioppo, 1979, 1981,

1984; Petty, Cacioppo, & Goldman, 1981; Petty, Cacioppo, & Schumann, 1983).









In contrast, when the level of cognitive elaboration is low, the message receiver is

said to be in the low involvement condition. Then, the peripheral route of persuasion is

more likely to change the receiver's attitudes. The peripheral route is often used when the

material presented is not as personally relevant, and the individual is not motivated to

analyze the facts presented (Petty, Gleicher, & Jarvis, 1993).

Involvement and Environmental Advocacy


Involvement has been widely studied in the ELM by marketing and advertising

research. Advertisers strive to produce messages that will involve people, prompting

audiences to devote more attention to and engage in more elaboration of the information

contained in the ads (Gordon, McKeage, & Fox, 1998). Advertising literature has

observed that the peripheral cues, such as source credibility or the number of arguments

are more effective for low-involvement rather than high-involvement products (Rhine &

Severence, 1970; Petty, Ostrom, and Brock, 1981; Heesacker et al., 1983; Gotlieb &

Sarel, 1991; Mazursky & Schul, 1992).

When the advertising messages are related to environmental issues, Schuhwerk and

Letkoff-Hagius (1995) found that issues involvement played a key role in the

effectiveness of green advertising appeals. Those who were highly involved with

environment issues were not swayed by green appeals by advertisers. On the contrary,

people who were less involved with environment issues found the appeal of "green"

advertising significantly more persuasive than non-green appeal.

Several consumer-based studies have investigated psychological aspects of

environmental concerns and environmental behaviors. Previous study found a positive

correlation between individuals' environmental attitudes and their intentions of









purchasing ecologically packaged products (Schwepker & Cornwell, 1991). Ellen,

Wiener, and Cobb-Walgren (1991) found that consumer efficacy, or the degree to which

individuals think they can make a difference in the quality of the environment, was

positively related to intent to purchase environmentally safe products.

Message Sidedness Effect


Definitions


Persuasion research has long questioned whether the persuasive communication is

more effective when it only presents the persuader's arguments or when it includes some

discussion (with or without refutation) of opposing arguments (Lumsdaine & Janis,

1953). Two approaches of a one-sided message and a two-sided message are often

contrasted and measured by researchers to examine various effects on attitudes toward an

issue or a brand (O'Keefe, 1999). It is defined that a one-sided message presents only

arguments favoring the position advocated by the source while a two-sided message, on

the other hand, presents both arguments opposing the source's position and those

favoring it (Hovland, Lumsdaine, & Sheffield, 1949). This basic contrast -ignoring or

discussing opposite arguments- has commonly been discussed in persuasion research as

the difference between one-sided and two-sided messages.

Some researchers even break down the two-sided message into the refutational

two-sided message and the nonrefutational two-sided message depending on whether a

subsequent refutation is included in the message or not. It is indicated that the effects of

message sidedness are moderated by the type of two-sided message employed (Allen,

1991; Allen, et al., 1990). Literature has suggested that the message sidedness effect may

vary significantly depending upon whether the opposing arguments are refuted (Allen,









1991, 1994; Crowley & Hoyer, 1994). A nonrefutational two-sided message

acknowledges opposing arguments but does not include direct refutation of the opposing

arguments (e.g. Bettinghaus & Basehart, 1969). A refutational two-sided message, on the

other hand, attempts to directly refute the opposing arguments (e.g., McCroskey, Young,

& Scott, 1972).

Origins


In persuasion research, the different persuasive effects that resulted from types of

messages is called message sidedness effect, which has a long research history that can

be traced back to earlier persuasive research by Hovland, Lumsdaine, and Sheffield

(1949). In their groundbreaking study, an issue concerning the military propaganda effect

was discussed. Two treatment groups were presented either with the one-sided message

that supported the communicator's position or the two-sided message which contained

additional opposite arguments. Their results did not find any significant differences

between the persuasive effect of the one-sided message and the two-sided message.

However, when participants have initial opposition to the communicator's view, it

was found that the two-sided message was more effective in producing opinion change in

the desired direction than was the one-sided presentation. The one-sided presentation

proved to be more effective only with those audiences who were already favorably

disposed toward the communicator's position. Moreover, their study also showed that a

two-sided message was more effective with better educated people, while a one-sided

message was more persuasive for less educated people (Hovland et al., 1953; Hovland et

al., 1949).









Previous Studies


Previous discussions of message sidedness effect have different conclusions. Some

research argued that two-sided messages appear more likely than one-sided messages to

produce persuasive effects in the desired direction. (e.g., Hovland et al., 1953; Etgar &

Goodwin, 1982; Albert & Golden, 1982; Golden & Alpert, 1987; Kamins & Assael,

1987; Kamins, Brand, Hoeke, & Moe, 1989; Pechmann, 1992; Bettinghaus & Cody,

1994; Crowley & Hoyer, 1994; Lang, Lee & Zwick, 1999; Stiff & Mongeau, 2003),

while other studies (e.g., Earl & Pride, 1980; Belch, 1981; Hastak & Park, 1990) do not

support the same findings. Thus, some other researchers (e.g., O'Keefe, 1999; Stiff &

Mongeau, 2003) urged caution in accepting research findings related to the message

sidedness effect.

Advertising and Non-Advertising Messages


According to Allen's (1991) meta-analytic review of the effects of one-sided and

two-sided persuasive messages, the accumulated research shows that the refutational two-

sided message appears to be more persuasive than both the one-sided message and the

nonrefutational two-sided message (Allen, 1991; Allen et al., 1990).

However, types of two-sided messages may have different persuasive effects

according to various topics of messages that are presented. For non-advertising messages

such as those on social or political issues, research found that refutational two-sided

messages have a persuasive advantage over one-sided messages but there was no such

advantage for nonrefutational two-sided messages (Allen, 1991, 1994; O'Keefe, 1999).

The message sidedness effect also differs according to receivers' initial attitudes.

Consistent with Hovland et al.'s (1949) findings, a marketing study found that a two-









sided message can be an extremely efficient persuasive technique when consumers

already hold negative attitudes toward a brand and when consumers are to be exposed to

negative counterclaims by competitors or by a neutral third party, including news media

(Sawyer, 1973; Golden & Alpert, 1987). Therefore some researchers suggest that adding

negative information to positive messages can increase credibility with an audience

(Golden & Alpert, 1987).

It is asserted that acknowledging opposing arguments may boost the

communicator's credibility by suggesting his or her honesty and lack of bias and thereby

increase the effectiveness of messages (e.g., Hovland, Lumsdaine, & Sheffield, 1949, p.

204; Settle & Golden, 1974; Pechmann, 1990). However, nonrefutational two-sided

messages on a non-advertising topic do not produce the same enhancement of credibility

(O'Keefe, 1999). It is predicted that when communicating a social issue, receivers may

feel confused when reading arguments from both supporting and opposing positions of

view. Therefore, the persuasive effect is greatly decreased.

Prior Knowledge, Motivation, Personality


Previous studies have indicated that prior knowledge about the issue influences the

effectiveness of each type of messages. More specifically, one-sided messages are more

effective when the audience is uninformed about the issue than when the audience is

knowledgeable. On the contrary, two-sided messages tend to persuade well-informed

recipients more than individuals who are unfamiliar with the issue (Chu, 1967; Hovland,

Lumsdaine, & Sheffield, 1949).

It is asserted that individuals' motivation to engage in processing messages

influences the effectiveness of the types of sided messages. One-sided messages are more









persuasive for people with lower motivations, while two-sided messages are more

effective when audiences have higher motivation. Sorrentino et al. (1988) explained that

recipients of the one-sided message may adopt the recommendations made by the

communicator without much cognitive work while those of a two-sided message engage

in active cognitive integration of conflicting arguments. Thus, a two-sided message is

expected to be more effective when recipients have the motivation to process the

arguments thoughtfully (Sorrentino et al., 1988). Crowley and Hoyer (1994) also

assumed that a two-sided message would increase the recipient's motivation to process

the message, which also leads to more cognitive responses and relatively enduring

attitude change.

Previous studies have discussed message sidedness in the framework of ELM.

Sorrentino et al.'s (1988) findings on personality and effects of message sidedness

indicated that under high personal relevance, uncertainty-oriented persons are more likely

centrally or systematically process the information. When under low personal relevance,

they will use peripheral or heuristic processing of information.

However, contrary to current theorizing, personal relevance does not necessary

increase recipients' cognitive analyses of persuasive messages. It was found that the way

people process information depends not only on the effects of personal relevance but also

on the personality of the recipients.

Issue Involvement


Most previous studies about message sidedness effect are mainly from marketing or

advertising research. In particular, consumer research has examined other effects of two-

sided message on increasing source credibility perception (Kamins & Assael, 1987;









Bohner et al., 2003), generating attitudinal resistance (Bither et al., 1971; Kamins &

Assael, 1987; Bohner et al., 2003), or reducing counter arguing (Belch, 1981; Kamins &

Assael, 1987; Swinyard, 1981). Other moderating variables such as the order of

argument, the type of two-sided message (refutational or nonrefutational), the availability

of counterargument, the initial agreement of recipients, the familiarity with the topic, or

educational background (Allen et al., 1990; O'Keefe, 1999) have also been examined by

researchers.

A substantial body of literature has been devoted to message sidedness effect (for a

review see Allen et. al, 1990). However, little previous studies (Chebat & Picard, 1985;

Sorrention et al., 1988; Hastak & Park, 1990) have examined the moderating role of

receiver's issue involvement. While a compelling case can be made for the type of

messages as a moderator of sidedness effects, research does not provide insight as to why

the messages differ in persuasiveness. One approach that can be taken to address this

issue is to study message recipients' cognitive responses of processing messages (Hale, et

al., 1991).

Three Theoretical Accounts for the Message Sidedness


Crowley and Hoyer (1994) contend that inconsistency in results on the effects of

message sidedness stems from the lack of a theoretical framework that specifically

addresses two-sided message effects. In this section, the researcher reviewed three

theories that addressed the reasons why two-sided messages have an advantage of

persuasiveness over one-sided messages.









Inoculation Theory


According to the inoculation theory, the two-sided refutational message appears to

work best to minimize counter argumentation (McGuire and Papageorgis, 1961). The

authors suggested that refutation enabled participants to be pre-exposed to weakened

counterarguments that heightened their defenses and alerted them to possible attacks on

established beliefs. Inoculation theory also predicts that the refutational appeal should be

the highest among the three appeals in the degree of support argumentation incurred

(McGuire and Papageorgis, 1961).

It is suggested by the inoculation theory that, refutational two-sided appeal leads to

greater acceptance of the communicator's position than the one-sided or two-sided

nonrefutational appeal, even when the receiver is in agreement with the communicator's

position (McGuire, 1961; Etgar and Goodwin 1982).

Inoculation should bolster the individual's cognitive defense by triggering the

search for supporting arguments in light of potential counterarguments. In an advertising

study, an inoculation of product disclaimers for relatively nonsalient product attributes

comprised in a nonrefutational appeal could trigger a moderate to high degree of support

argumentation. Because of the absence of inoculation and lack of perceived advertiser

truthfulness, the one-sided appeal should produce the lowest degree of support

argumentation (Kamins & Assael, 1987).

Correspondence Theory


From the perspective of communicator's credibility, Tannenbaum (1967) argued

that refutational two-sided appeals are effective communication devices because their

credibility is enhanced by the opposing arguments presented to consumers. Applying









correspondence theory to advertising study, researchers (Smith & Hunt, 1978) suggest

that people perceive a lower degree of credibility and show higher levels of source

derogation for one-sided appeals than for two-sided communications. In other words,

consumers consider one-sided advertisements are not as credible as two-sided appeals.

Kamins and Assael's (1987) empirical study on the sided appeals in advertising

research found that two-sided refutational appeal resulted in significantly more support

argumentation and significantly less counterargumententation than use of the one-sided

appeal. Consistent with correspondence theory, the most novel appeal should lead to

correspondent attribution and the lowest degree of source derogation upon exposure. The

two-sided nonrefutational appeal should be perceived as most novel as no attempt is

made to refute product disclaimers.

In particular, research on two-sided nonrefutational appeals has been abundant in

the marketing literature. Generally, two-sided nonrefutational appeals were more

effective in increasing perceived advertiser truthfulness (Smith and Hunt, 1978;

Swinyard, 1981) and believability (Etgar & Goodwin, 1982; Anderson & Golden, 1984;

Golden & Alpert, 1987).

Inoculation and correspondent theories contribute to predicting cognitive responses.

The inoculation theory is predicting participants' responses- counter or supportive -

based on the communications stimuli. On the other hand, correspondence theory is

concerned primarily with the attribution to the source- enhancing or derogatory based

on the same communication stimuli.









Cognitive Response Theory


Previous research has shown clearly that the message sidedness effect is weak,

because the effect is indirect (Hale et al., 1991). Hale et al. (1991) asserted that cognitive

responses and message evaluation should be considered as two mediating variables

between the transmission of the message and the attitude measure (p.387).

Numerous studies that have reasoned that message sidedness can be explained by

the cognitive response theory (Allen et al., 1990; Hale et al., 1991; Crowley & Hoyer,

1994). Ford and Smith (1991) assume that such refutational messages may require more

commitment and thus result in high elaboration. Sorrentino et al. (1988) explain that

recipients of a one-sided message may adopt the recommendations made by the

communicator without investing much cognitive work while those of two-sided messages

engage in active cognitive integration of conflicting arguments. Thus, two-sided

messages are expected to be more persuasive when receivers are more motivated to

process the arguments thoughtfully (Sorrentino et al., 1988). Crowly and Hoyer (1994)

also assumed that a two-sided message will increase receivers' motivation to process the

message, which also lead to more cognitive response and relatively enduring attitude

change.

According to Hale et al. (1991), cognitive responses and message evaluation play

an order effect in processing persuasive messages. Their data appears that through

exposure to a message prompts the generation of cognitions, which produces an

evaluation of the message, and that evaluation influences a recipients' attitude (p.387).









It was also found that the number of positive cognitions generated is greater for

refutational two-sided messages than for either nonrefutational two-sided message or

one-sided messages.

Hale et al. (1991) suggested that message sidedness can be considered part of the

ELM. Their findings indicated that messages sidedness influenced message elaboration,

which is the generation of message relevant cognition.

One intuitive possibility is that direct refutation of an opposing argument increases

the perceived strengths of the argument. However, in the nonrefutational two-sided

message, receivers may have trouble comparing arguments from conflicting points of

view they perceive them to be weaker. The one-sided message may simply be perceived

as being weaker than the two-sided message because any mention of opposing arguments

is missing.

There is considerable evidence demonstrating that perceived argument strength is

positively related to the quantity of positive cognitions (Petty & Cacioppo, 1986b). In the

current research, it is predicted that the refutational two-sided message will be perceived

to have the strongest argument strength, therefore, induce the highest number of positive

cognitions.

The ELM suggests that attitude change results from cognitive elaborations of

audiences after receiving persuasion messages (Petty & Cacioppo, 1981, 1986a, 1986b).

However, it is important to note that the ELM does not directly predict an attitudinal

advantage for two-sided messages (Hastak & Park, 1990). All that this model asserts is

that attitudinal advantage for a two-sided message will be mediated by message-related

cognitive responses (p.330). However, message sidedness is connected to the ELM when









"motivation and ability to think about the issue will determine the route to persuasion"

(Allen, 1991, p.391). The current study of message sidedness is a response to the research

call of Hale et al. (1991) that explores the possibility of message sidedness functions as a

central persuasive cue.

Hypotheses


The Table 2-1 shows the factorial design of the experiment. As noted before,

involvement is defined as the message recipients' motivation to process information

about an issue from a message (Petty and Caciopppo,1979; Johnson & Eagly, 1989)

Participants in the high involvement condition will be more motivated to process the

message than those in the low involvement condition.

Hypothesis 1 predicted that participants in a high involvement condition (X1)

would generate more message-related thoughts than those in a low involvement

condition. (Predicted main effect)

Hypothesis 2 predicted that a two-sided refutational message (Y1) would

induce more message-related thoughts than a one-sided (Y3) or a nonrefutational

two-sided message (Y2). (Predicted main effect)

Few results on the interactive effects of message sidedness and perceived CSR are

available. No theory-articulated reasoning is provided to fill the gap and a hypothesis is

provided. Since it is hard to find any research or any theoretical accounts that explain the

effect of message sidedness on perceived CSR, a research question is proposed.

RQ1: How does message sidedness influence people's perception about the

corporate social responsibility of a company? Will a two-sided message or a one-

sided message lead to more positive perceived CSR?









According to Petty and Cacioppo (1979) under high involvement condition,

participants engaged message processing, which enhances persuasion for favorable and

strong argument. The ELM argues that when involvement is high, people are likely to use

more cognitive processing and more message related thoughts, which engage the central

route. Further, Hale, et al.'s (1991) study has shown that when participants are exposed to

refutational two-sided messages, they are likely to generate more positive cognitive

cognitions than those exposed to one-sided messages. Following this line of research:

Hypothesis 3 predicted that participants exposed to a refutational two-sided

message would show more positive attitudes than those who were exposed to a one-

sided or a non-refutational two-sided message.

Hypothesis 4 predicted that participants in a low involvement condition would

show more positive attitudes with a one-sided message while those in a high

involvement condition would express more positive attitudes with a refutational

two-sided message. (Predicted interaction effect)

Hypothesis 5 predicted that higher perceived CSR would be positively

associated with a positive attitude toward the company.

Table 2-1. Factorial design of the experiment
Personal Message Sidedness (Y1, Y2 & Y3)
Relevance/Involvement Refutational Nonrefutational One-Sided Main
(Xl & X 2) Two-Sided Two-Sided Effects for
Involvement
High Involvement A B C X1 >
Low Involvement D E F X2
Main effects for Sidedness Y1 > Y2> Y3














CHAPTER 3
METHODOLOGY

The present study investigates empirically the persuasiveness of a CSR campaign

in influencing message recipients' attitudes and perceptions. Based on the level of issue

involvement and the effects of message sidedness presented in CSR messages, this study

also examines the main effect and the interaction effect between these two variables on

message recipients' attitude toward the company and their perceptions of the company's

social achievements.

In addressing the socially responsible achievements, the present study asked

whether a company should discuss a relevant issue to communicate with key stakeholders

or active publics. In other words, should a company present simply a CSR message or

utilize the CSR campaign as an opportunity to address a relevant environmental or a

social issue?

To answer the above questions, a factitious lumber company named CALCO was

created in the experiment. The reason for choosing a lumber company in experiment

design is that timber industry has long been questioned for the practice of lumbering such

as clearcutting that result in critical environmental damages. In addition, the continuing

controversy over clearcutting has become essentially a social issue. Therefore, the present

study is able to test whether people's attitude toward the company can be positively

enhanced after receiving a CSR campaign which advocates the social performance of that

company. Moreover, this study further tests whether the enhancements of the positive

attitude toward the company can be increased when the company particularly addresses









controversial issues that are critical to the business. In brief, the purpose of this study is to

test whether people will have more positive attitude if the company is considered as

socially responsible.

By applying a sophisticated experimental study on attitude that examines the

cognitive processing model of a CSR message, the present study adds a body of

knowledge to the research of CSR and the ELM. In particular, the present study advances

the existing research and literature on the effect of message sidedness on CSR messages

and environmental issues. Moreover, this study is the first to link between CSR

communication and issues management and empirically examine the persuasive effects of

the CSR messages on attitude toward the company and the perceptions of the company's

social performance. This innovative approach is based on the growing trend of today's

corporate world to bring the economics together with the environmental and the social

needs as part of a business strategy. From an issues management perspective, the present

study also presents a good opportunity for public relations practitioners to demonstrate

the importance of PR function, pursuing an excellence in CSR communication

campaigning.

An Experimental Design

This study is grounded on a cognitive response perspective in examining the

underlying processes that mediates the persuasive effects of a CSR campaign. An

experimental method was considered the most appropriate approach in testing a cognitive

response model proposed in this study.

A mixed 2x3 factorial experiment was designed with two between-subject variables

being personal involvement (high or low) and message sidedness (one-sided, refutational

two-sided and nonrefutational two-sided). Participants' attitudes toward the company and









their perceptions of the company's social performance were evaluated after reading the

CSR campaigns presented in the stimuli.

Involvement level and message sidedness have been chosen as two independent

variables of testing the persuasiveness of a CSR campaign. According to the ELM (Petty

and Cacioppo, 1986a, 1986b), issue involvement plays an important role in influencing

which route people choose to process the message. The ELM has indicated that for the

desired central route of attitude change to occur, participants must be highly involved

with the issue and able to cognitively process the persuasive messages. Therefore,

participants' level of involvement with the issue discussed in the CSR messages was

controlled in the experiment.

Previous research has indicated that message sidedness could result in different

persuasive effects on attitude change. Therefore, message sidedness was experimentally

varied so that participants in this study received either a one-sided CSR campaign, a

refutational two-sided CSR campaign, or a nonrefutational two-sided CSR campaign. All

three CSR campaigns addressed a lumber company's social performance and the

clearcutting issue which was associated with the company's forestry management.

In this study, the one-sided CSR campaign only presented arguments that support

the company's environmental commitment and corporate citizenship. The refutational

two-sided CSR campaign comprised pairs of arguments including both supportive and

refutational information. For example, arguments against the company's clearcutting

practices were followed by refutational explanations that defended the company's

goodwill. The nonrefutational two-sided CSR campaign acknowledged the concerns from

the opposite positions such as environmental activists groups, but did not attempt to









refute those claims directly. These procedures are similar to the two-sided

operationalisation used in previous message sidedness studies (Allen, 1991).

Consistent with previous research of the ELM, it is assumed in this study that the

more a person is involved with the environmental issue, the more he or she thinks

carefully about (or cognitively elaborates on) the CSR message, eventually the more

likely positive attitudes are to occur. The results of the current study will contribute to the

research of cognitive response theory and demonstrate the value of the effects of CSR

campaigns on attitude toward the company and the function of CSR communication as

issues management.

Pilot Study

Procedure

In order to make sure the manipulation checks of message sidedness and

involvement work successfully, a total of fifty-four college students participated in a pilot

test. All students were recruited on the campus of the University of Florida and assigned

to one of six treatment conditions randomly.

The message stimulus was a CSR campaign presented on the Web site of a lumber

company named CALCO. The purpose of the CSR campaign is to advocate the social

performance and the corporate citizenship of that company. Participants were presented

with a booklet including a background news story, a CSR campaign message, and a

questionnaire. The fist part of the booklet was a one-page newspaper article about an

environmental activist named Julie "Butterfly" Hills who lived in a tree for two years to

protest against the lumber company's clearcutting practices. CALCO's CSR campaigns

on the company Web site were the second part of the booklet, and finally a five-page

questionnaire was included as the third part of the booklet.









Participants were instructed with the following statement to complete the

questionnaire once they have finished reading all the messages presented in the bookelt.

"Based on the background story and the CALCO's statement you just read from the

printed Web pages, please answer the following questions. Please circle the number that

best describes your thoughts or feelings." The questionnaire included Liker-type items

designed to measure how interested respondents were in environmental and clearcutting

related issues, their feelings toward the company presented, and their level of agreement

with various attitudes statements.

Participants' level of involvement was manipulated by the location of the lumber

company presented in the message stimulus. For high involvement groups, the lumber

company presented in the stimuli was located in Florida. On the other hand, the lumber

company was located in Taiwan for low involvement groups. Involvement level was

measured using four questions (see Table3-1): (1) How important are environmental

issues in Taiwan (or Florida) to you personally?; (2) How important is the deforestation

issue in Taiwan (or Florida) to you personally?; (3) How much are you concerned about

environmental issues in Taiwan (or Florida)?; (4) How much are you concerned about

deforestation issue in Taiwan (or Florida)?

Table 3-1. Liker-type items of the manipulation check of involvement level for high
involvement groups
How important are environmental issues in Florida to you personally?

How important is the deforestation issue in Florida to you personally?

How much are you concerned about environmental issues in Florida?

How much are you concerned about the deforestation issue in Florida?









For the manipulation check of message sidedness, we used two items from Trent

and Greer's (2001) items: (1) CALCO's statement shows two-sided arguments; (2)

CALCO's statement refutes the opponent's arguments (see Table 3-2).

Table 3-2. Liker-type items for the manipulation check of message sidedness
CALCO's statement shows two-sided arguments.

CALCO refutes opponents' arguments.



Pilot Test Results

From the pilot test, a successful manipulation of involvement was obtained (M low

=3.94, SD=1.69; M high = 4.95 SD=1.19; t=2.54, df=50, p<.05). The mean score of

involvement index from the high involvement group is significantly higher than those

from the low involvement group. However, the results of the manipulation check of

message sidedness did not yield significant differences in the pilot test. In other words,

participants can not recognize the differences between three groups of CSR campaigns

(one-sided CSR message, refutational two-sided CSR message, and nonrefutational two-

sided CSR message). The manipulation of message sidedness failed in the pilot test.

The major problem was attributed to the design of background information page

about the clearcutting issue associated with CALCO's practices. In the pilot test, a one-

page newspaper article about Julie Hills' tree sitting protest against CALCO's clearcutting

practice for two years was provided. It was assumed that this new article might have

delivered a too strong message that overrode the message sidedness effect presented in

CSR campaigns. The results from the pilot test shown overall negative attitudes toward

CALCO. Thus, to reveal the effect of message sidedness in CSR campaigns, CALCO's









"About Us" page (see Appendix A) including a brief introduction of the company's

history and visions was substituted for the news article in the main experiment.

Main Study

Sample

A convenient sample was adopted in this study. It was considered appropriate to

use a convenient sample in this study due to the theoretical testing nature of the

experiment (Calder, Phillips, & Tybout, 1981). Calder, Phillips, & Tybout (1981) stated

that it is important to use a homogeneous sample for an experiment because of its

theoretical testing in nature. They argued that a homogeneous sample permits more exact

theoretical predictions than from a heterogeneous group. In addition, a homogenous

sample also decreases the chance of making a false conclusion about whether there is a

covariation between the variables under study (Cook & Campbell, 1979).

The participants for the main study were recruited from two colleges of the

University of Florida. The experiment was conducted in two undergraduate classes from

the College of Agricultural and Life Sciences and the College of Journalism and

Communications. A total of 205 students (68 male, 113 females, 24 missing) participated

in the study. Therefore, educational level and the age of participants appeared to be

indicative of traditional college students.

Participants completed the study at the beginning of the class session and received

extra credit points as compensation for their participation. One hundred and four

participants were in the 2 (involvement: high vs. low) by 3 (message sidedness: one-sided

vs. refutational two-sided vs. nonrefutational two-sided) groups, and 21 were in the

control group. Thus, all participants were randomly assigned to one of the following

seven conditions: 1)high vs. one-sided message, 2)low vs. one-sided message, 3)high vs.









nonrefutational two-sided message, 4)low vs. nonrefutational two-sided message, 5)high

vs. refutational two-sided message, 6)low vs. refutational two-sided message or 7)a

control group (see Table 3-3).

Table 3-3. The conditions of the 2x3 experimental design
ONE-SIDED NR R
TWO-SIDED TWO-SIDED
HIGH Group (1) Group (3) Group (5)
INVOLVEMENT
LOW Group (2) Group (4) Group (6)
INVOLVEMENT
Note. 1. NR stands for "nonrefutational." R stands for "refutational."


Stimuli

The CSR campaign stimuli used in the main study were slightly different from that

of the pilot study. The newspaper article about Julie Hills' protest against CALCO's

clearcutting practices was replaced by a brief introduction of the company's history and

visions titled as "About Us" in the main experiment. Participants were asked to read three

printed Web pages including a brief introduction about the company, the company's

environmental commitment, and corporate social activities.

The Web site of one of the largest lumber companies in the United States was

mirrored and slightly changed for simulating Web pages of a fictitious company named

"CALCO." To manipulate different levels of involvement, two versions of CALCO's

"About Us" page were created. For high involvement conditions, CALCO was presented

as "The Caribbean Lumber Company" in Southwest Florida in the "About Us" page (See

Appendix A). On the other hand, for low involvement conditions, CALCO was presented

as "The Chiayi Lumber Company" in Taiwan on the same page (See Appendix B).

To manipulate message sidedness (one-sided vs. refutational two-sided vs.

nonrefutational two-sided), three versions of CALCO's "Environmental Commitment"









Web pages were created. For the one-sided version, the CSR campaign only contained

one page of supportive messages that advocated CALCO's corporate social activities in

three areas: environment, community, and employees. Each domain of CSR activities

was subtitled as "Responsibility to Our Environment," "Responsibility to Our

Community," and "Responsibility to Our Employees." Moreover, this page also has two

versions for high involvement groups (See Appendix C) and for low involvement groups

(See Appendix D) depending on the location of CALCO presented in the campaign.

For the refutational two-sided version, one page of "Myth and Fact" (See Appendix

E) which contained activist groups' concerns of clearcutting issues and CALCO's

refutational arguments was included in addition to the CSR messages the same as the

one-sided group. Therefore, the refutational two-sided version has two pages in total. The

refutational message dealt with three controversies related to the timber industry's

clearcutting practice by displaying activist groups' claims about the impact of

clearcutting on aesthetic and recreational value, deforestation, and ecosystem. Each

argument was refuted with CALCO's positions.

For the nonrefutational two-sided version, in addition to the same CSR messages as

the one-sided version, one page of activist groups' claims about clearcutting issues was

presented without any refutations or arguments from CALCO's perspectives (See

Appendix F). Thus, the nonrefutational two-sided version also has two pages.

To ensure the manipulation of both involvement level and message sidedness

successfully instituted in different treatment conditions, a pilot test with a separate pool

of 54 participants in total was conducted.









Procedure

Students were informed during that day's class session that there would be an extra

credit opportunity if they complete the study. Students were also informed that this

opportunity would require them to read a couple of corporate Web pages, and answer

several questions based on what they read. Participants were told that the purpose of the

experiment was to develop creative headlines and body copies for a lumber company's

Web site. Seven versions of booklets including six for the treatment groups and one for

the control group were given to participants based on randomly assigned experimental

conditions.

The stimuli booklets included one printed Web page of CALCO's "About Us," two

printed Web pages of CALCO's CSR advocacy titled "Environmental Commitment," and

finally a five-page questionnaire. Two versions of questionnaires were provided

including a U.S. version for high involvement groups (See Appendix G) and a Taiwanese

version for low involvement groups (See Appendix H). After reading the stimuli,

participants were asked to complete a questionnaire that contained a series of measures

on attitudes, perceived CSR, and message-related thoughts. After participants responded

anonymously to questions regarding the persuasive effects of CSR messages, they

completed a measure for the relationship study and answered demographic questions. The

results of the relationship questions will not be discussed in this thesis as they are part of

another study that will be explained in a different report.

Finally, participants were debriefed and thanked for their participation. Students

were informed that their completion of the measures was voluntary and anonymous. In

order to assure the anonymity of participants' responses, each participant turned in a

separate consent form with his or her name written on it.









Operational Definitions: Independent Variables

Involvement

Involvement is a motivational variable that moderates how individuals process

information. The manipulation of involvement differentiates the motivation people have

while processing the information which in this study is about the company's CSR

initiatives. In this study, involvement is defined in terms of issue involvement which can

be explained by the "extent to which the attitudinal issue under consideration is of

personal importance" (Petty & Cacioppo, 1979, p. 1915). The manipulation of issue

involvement in this study is also closely paralleled by the study of Petty, Cacioppo, and

Schumann (1983). In Petty et al.'s (1983) research, the high involvement group was led

to believe that a comprehensive exam would be held at their university while the low

involvement group was told that the exam would be conducted in a different university.

Personal relevance occurs when people expect the issue "to have significant

consequences for their own lives" (Apsler & Sears, 1968). Therefore, it is clear that in

high involvement conditions, participants would be more motivated to engage in

thoughtful consideration of the issue because they would be affected personally whereas

in low involvement conditions, they would not.

According to these findings, participants' levels of involvement were manipulated

by receiving different background information about CALCO which was presented either

as a lumber company located in Florida or located in Taiwan. Take the high-involvement

condition for example, participants read the following messages which were included on

the first page of their stimuli package: "CALCO is The Caribbean Lumber Company in

Southwest Florida. We have been growing trees, making wood products, and supporting

generations of families in Southwest Florida since 1863." On the other hand, for the low









involvement condition, CALCO was presented as "The Chiayi Lumber Company" in

Taiwan on the same page.

Message sidedness

Message sidedness was experimentally varied and participants received one of the

three message packages: one-sided message, refutational two-sided or nonrefutational

two-sided messages to discuss CALCO's CSR activities. All three packages include one

page of background information about the company, and another "Environmental

Commitment" describing CALCO's CSR activities which is to be found on the last page.

For the refutational two-sided CSR campaign and the nonrefutational two-sided CSR

campaign, an additional page on "Environmental Commitment" discussing the

clearcutting issue was included as well.

The "About Us" Web page contains the brief introduction about CALCO's name,

products, history and visions. The following paragraphs are some examples of sentences:

Established more than 140 years ago, CALCO still embodies the vision and spirit
of its founders.

Today, their vision is a reality. The modern CALCO is a dynamic and innovative
company whose employees are dedicated to providing superior lumber products to
customers and value to shareholders.

For the refutational two-sided message, an additional page of "Myth and Fact"

contained three controversies about the impact of clearcutting on (1) deforestation, (2)

aesthetic and recreational value, and (3) ecosystem. The following paragraphs are

examples of arguments related to the impact of clearcutting on deforestation:

Myth: Environmental activist groups have argued that clearcutting leads to
deforestation. They have also alleged that nothing can be regenerated once a forest
is cut down and that it will never be the same again.

Fact: Deforestation is the removal of a forest with no intention of establishing a
future stand of trees. For instance, deforestation is occurring today in Brazil and









other tropical areas in the form of agricultural conversion. Our silvicultural
clearcutting is both a harvest and a regeneration of the forest, and is done to
improve future stand quality, growth, genetics, and species composition.

While regarding the nonrefutational two-sided message, environmental activists

groups' claims were included without refutational arguments from CALCO's perspective.

The following sentences are provided as examples:

Environmental activist groups have argued that clearcutting leads to deforestation.
They have also alleged that nothing can be regenerated once a forest is cut down
and it will never be the same again.

Another argument against clearcutting is that the resulting bare patches of land look
unsightly. Activist groups argue that clearcutting is the destruction of aesthetic
values and recreational opportunity.

In all the three message packages, one page of "Environmental Commitment" was

included on the last page of the CSR campaign. Three domains of social activities are

presented: (1) "Responsibility to the Environment," (2) "Responsibility to Our

Community," and (3) "Responsibility to Our Employees." The following sentences about

"Responsibility to the Environment" are provided as examples:

CALCO now offers office paper with 100 percent post-consumer recycled content.
Tree-free papers made from kenaf, hemp, or agricultural waste such as wheat straw,
rice straw and bagasse (waste from sugar cane processing) are also available. There
is also an increasing range of forest-friendly products for use in construction and
furniture making.

The following sentences about "Responsibility to Our Community" are provided as

examples:

Sustaining the communities in which we operate is a responsibility and a
commitment we take very seriously. We believe that we have an obligation to
contribute to the quality of life in our communities. Our major contribution to the
people in Sarasota County is in our ability to provide hundreds of family-wage
jobs.

As part of that commitment, we also provide a medical and dental facility for
CALCO employees and people living in our communities. Each year, the CALCO









College Scholarship program awards over $ 350,000 to students from Sarasota
County.

The following sentences related to "Responsibility to Our Employees" are provided

as examples:

CALCO fosters personal and professional growth and learning for all employees.
Employees in CALCO are empowered to make decisions and to share both the
risks and rewards of making decisions. In addition, CALCO employees are eligible
to receive $ 12,000 for attendance at any four-year accredited university or college
and $4,600 for any two-year college or trade or religious school.

Operational Definitions: Dependent Variables

Cognitive elaboration

According to Petty and Cacioppo (1986a, 1986b), cognitive elaboration was

defined as message-related thoughts. In order to evaluate participants' level of cognitive

elaboration when they were processing the presented CSR messages, participants were

requested to list all the thoughts that came to their mind while reading the printed Web

pages of the company. Participants read the following instructions on writing their

message-related thoughts, "Please list the thoughts that came to your mind while reading

CALCO's statement generate as many as you can. Please write your thoughts in the

following lines, one thought per line." Eight lines were provided and they were followed

by the above instructions.

Perceived CSR

In this study, the construct of CSR as the audience's subjective perceptions is more

likely to be formed by exposure to corporate CSR advocacy messages. Thus, the term of

perceived CSR is used in this study instead of CSR. To access participant's perceptions

about the company's CSR activities and commitment, Menon and Kahn's (2003)









measurements of perceived CSR was applied in this study with a slight change in

wording.

Participants' responses of CALCO's environmental policies were asked by the

following five questions: (1) CALCO believes in environmental commitment; (2)

CALCO is likely to follow environmental friendly rules and policies; (3) CALCO is

highly involved in community activities; (4) CALCO is highly concerned about

environmental issues; (5) CALCO is genuinely concerned about public welfare (see

Table 3-4). Participants were told to show their agreement with above questions by 7-

point rating scales anchoring "1" as "strongly disagree," and "7" as "strongly agree."

The statistical results show that these items on the perceived CSR scale is very

reliable (alpha = .92).

Table 3-4. Liker-type items for perceived CSR
CALCO believes in environmental commitment.

CALCO is likely to follow environment-friendly rules and policies.

CALCO is highly concerned about environmental issues.

CALCO is highly involved in community activities.

CALCO is genuinely concerned about public welfare.



Attitudes

A 7-point semantic differential scale with six items was used to evaluate

participants' attitudes toward CALCO. Priester and Petty's (2003) attitude scale is

applied with a slight change in wording for evaluating participants' attitudes toward the

company in this study. The items of attitude scale included in this study were negative-

positive, harmful-beneficial, foolish-wise, unfriendly-friendly, bad-good, and









unfavorable-favorable (see Table 3-5). The following instructions were given for

participants: "Please evaluate how you feel about CALCO by circling a number on each

of the scales below. If you feel that you have no reaction, please circle the number 4 to

indicate your neutrality."

The statistic results of the items applied in this study turned out to be very reliable

(alpha = .95).

Table 3-5. Semantic differential type items for attitude toward the company
Negative 1 2 3 4 5 6 7 Positive
Harmful 1 2 3 4 5 6 7 Beneficial
Foolish 1 2 3 4 5 6 7 Wise
Unfriendly 1 2 3 4 5 6 7 Friendly
Bad 1 2 3 4 5 6 7 Good
Unfavorable 1 2 3 4 5 6 7 Favorable


Manipulation Check

Involvement

The manipulation check for involvement remained the same as in the pilot study. In

this study, participants' levels of involvement were controlled according to the different

locations of the company. For the low involvement condition, CALCO was presented as

a lumber company in Taiwan, whereas for the high involvement group CALCO was

presented as a lumber company located in Florida. It was assumed that participants would

feel more involved with the issue presented in the messages if the company was located

in Florida.

To evaluate the manipulation of involvement, four questions were asked to

examine participants' involvement level: (1) How important are environmental issues in

Taiwan (or Florida) to you personally?; (2) How important is the deforestation issue in

Taiwan (or Florida) to you personally?; (3) How much are you concerned with









environmental issues in Taiwan (or Florida)?; (4) How much are you concerned with

deforestation issue in Taiwan (or Florida)? (see Table 2).

Message sidedness

The manipulation check for message sidedness also remained the same as in the

pilot study (see Table 3). For the manipulation check of message sidedness, we used two

items from Trent and Greer's (2001) with a slight change in wording: (1) CALCO's

statement shows two-sided arguments (2) CALCO refutes opponents' arguments.

Participants show their agreement with 7-point rating scales anchoring as follows: "1" as

"strongly disagree," and "7" as "strongly agree."














CHAPTER 4
FINDINGS

The analyses of this study contained four parts. First section is the descriptive

statistics of the present participants in this experiment. Second part contains the

manipulations checks of the main experiment and reliability of dependent measures.

Third, the hypotheses and the research question are addressed, using t-test, One-Way

ANOVA, post-hoc analysis, Pearson correlation and regression analyses (hierarchical

regression). Finally, the path analysis was performed to measure the causal relationships

between (1) the effects of message sidedness and involvement level on perceived CSR

and (2) the impact of perceived CSR on attitude toward the company.

Profile of Participants

All respondents used for analysis in this experiment were college students at the

University of Florida. A total number of 205 students participated in this study. Of the

total 205 participants, 68 (33.2%) were males, 113 (55.1%) were females and 24 (11.7%)

did not report.

Of those who responded to the question of ethnicity, 114 (55.6 %) were White-Non

Hispanic, 30 (14.6%) were Hispanic American, 16 (7.8%) were Asian American, 14

(6.8%) were African American, and 7 (3.1%) reported other. Participants' age ranged

from 18 to 40. However, over 90% of participants were in the 18-24 age groups. This

study was conducted in two undergraduate classes. Therefore, educational level and the

age of participants appeared to be of traditional college students.









205 participants were randomly assigned to one of the seven groups. 21 participants

were assigned in the control group, and the remaining 184 participants were assigned in

the following treatment groups:

1) 31 in the high and one-sided message group, 2) 30 in the low and one-sided

message group, 3) 31 in the high and nonrefutational two-sided message group, 4) 31 in

the low and nonrefutational two-sided message group, 5) 30 in the high and refutational

two-sided message group, 6) 31 in the low and refutational two-sided message group (see

Table 4-1).

Table 4-1. The number of participants in each cell
Involvement One-Sided NR R Control Total
vs. Message Two-Sided Two-Sided
Low 30 (32.6%) 31 (33.7%) 31 (33.7%) 92 (100%)
High 31 (33.7%) 31 (33.7%) 30 (32.6%) 92 (100%)
Control 21 (100%) 21 (100%)
Total 61 (29.8%) 62 (30.2%) 61 (29.8%) 21 (10.2%) 205 (100%)
Note. 1. NR stands for "nonrefutational." R stands for "refutational."

Results of Manipulation Checks

Involvement

Consistent with the results from the pilot test, the manipulation check of

involvement worked successfully. The four indicators used to check the effectiveness of

the involvement manipulation yielded converging and supportive results. Table 4-2

shows the mean scores for two groups that pertained to either a high- or a low-

involvement condition were significantly different (M low = 3.89, SD = 1.62; M high =

5.08, SD = 1.32; t = 5.39, df = 179, p < .001).

In other words, the manipulation and the measurements of involvement level

applied in this study which closely paralleled by the study of Petty, Cacioppo, and

Schumann (1983) work successful.









Table 4-2. The manipulation check of involvement
Measure Treatment N Mean SD t df
Involvement Low 90 3.89 1.62 5.39*** 179
High 91 5.08 1.32
Note. Items in the involvement scale were measured on a 7-point scale ranging from 1
(not at all) to 7 (very much).
*** p< .001

Message Sidedness

For the manipulation check of message sidedness, the revised background story

turned out to be a success. After the news article about clearcutting issue was replaced by

a CALCO's "About Us" Web page, the manipulation check of message sidedness worked

successfully. The original background story about Julie "Butterfly" Hills' tree-sitting

protests might have delivered a strong message to the experiment group that possibly had

overridden the message sidedness effect. Therefore, it was difficult to find significant

differences between participants' attitude toward the company even after exposure to

CSR campaigns advocating the company's good citizenship and its corporate social

performance. Therefore, a brief introduction to the company's history was substituted for

the news article.

Message sidedness resulted in a significant mean difference between three

treatment groups. As researcher's expectation, the mean score of the message sidedness

index for the one-sided message (M=3.44, SD=1.06) turned out to be the lowest among

three sided message groups. Table 4-3 indicates that the mean score for a refutational (R)

two-sided message was significantly higher than that of a nonrefutational (NR) two-sided

message and that of a one-sided message.









Table 4-3. The manipulation check of message sidedness
Measure Treatment N Mean SD F df
Message One-Sided 61 3.44 1.06 23.67 2/181
Sidedness NR Two-Sided 62 3.90 1.03
R Two-Sided 61 4.85 1.33
Note. 1. NR stands for "nonrefutational." R stands for "refutational."
2. Items in the message sidedness scale were measured on a 7-point scale ranging from 1
(strongly disagree) to 7 (strongly agree). Question Items used in the message sidedness
were "CALCO's statement refutes the opponent's arguments," and "CALCO's statement
shows two-sided arguments."
*** p< .001

Reliability Checks for Dependent Measures

In order to establish the reliability of dependent measures used in this study,

reliability analyses were conducted for each of construct used in this study including

perceived corporate social responsibility (CSR) and attitude toward the company.

Cronbach's alpha was computed to evaluate if the items within each index have high

internal reliability. Alpha represents a coefficient that demonstrates how well the items

measuring the same characteristic correlate with one another (Hon & J. E. Grunig, 1999).

In general, reliability coefficients over .90 are considered "excellent," over .80 are "very

good," and over .70 are considered "adequate" (Kline, 1998).

The index of perceived CSR includes five questions relating to CALCO's

environmental policies. Participants' attitudes toward CALCO were evaluated by a 7-

point semantic differential scale with six items. Reliability tests confirmed that the

indexes of overall attitude toward the company (Cronbach's a =.93) and perceived CSR

(a = .90) applied in this research was appropriate. Therefore, the statistical analysis

indicated that the dependent measures used in this experiment had a high internal

reliability.









Test of Hypotheses

Cognitive Elaboration

Hypothesis 1: Participants in a high involvement condition will generate more

message-related thoughts than those in a low involvement condition. (Predicted

main effect)

Based on the literature review, it was assumed that message sidedness and

involvement have main effects on cognitive elaboration. Contrary to the expectations, the

main effects of involvement and message sidedness were not significant. Hypothesis 1,

"participants in a high-involvement condition would generate more message-related

thoughts than those in a low-involvement condition," was rejected.

A t-test was conducted to test the main effect of issue involvement level on

cognitive elaboration. As Table 4-4 demonstrates, no statistical differences of message-

related thoughts between a high and a low involvement group were found (p= .398).

Thus, Hypothesis 1 was rejected. Contrary to previous research, this study did not find

more cognitive elaboration among participants of the high-involvement condition. These

results showed that regardless of whether the participant is highly involved or less

involved with the issue, it is likely that he or she would generate the same amount of

message-related thoughts about to the issue discussed in the messages.

Table 4-4. A t-test of mean difference of message-related thoughts generated between the
high and the low involvement group
Measure Treatment N Mean SD t Df
Involvement Low 91 3.42 2.21 .190 181
High 92 3.36 1.99
Note. 1. Participants were requested to generate message-related thoughts after reading
experimental stimuli.
2. p= .398









Hypothesis 2: A two-sided refutational message will induce more message-related

thoughts than a one-sided or a nonrefutational two-sided message. (Predicted main

effect)

To test Hypothesis 2, a One-Way Analysis of Variance (ANOVA) test was run (see

Table 4-5). In this experiment, no statistically significant differences were found between

the numbers of message-related thoughts from different message sidedness groups

(p=. 134). Thus, Hypothesis 2 was also rejected. For cognitive elaboration, there were no

significant effects attributed to involvement and message sidedness.

Table 4-5. A t-test of mean difference of message-related thoughts generated between
three groups of sided messages
Measure Treatment N Mean SD F df
Attitudes One-Sided 61 3.70 2.22 2.03 2/180
NR Two-Sided 62 2.97 2.10
R Two-Sided 61 3.49 1.91
Note. 1. NR stands for "nonrefutational." R stands for "refutational."
2. Items in the attitude scale were measured in a seven-point semantic differential scale.
3. p= .134

Main Effects of Message Sidedness on Perceived CSR

RQ1: How does message sidedness influence people's perception about corporate

social responsibility of a company? What will lead to more positive perceived CSR

between a two-sided message and a one-sided message?

None of the previous studies from the literature of CSR and persuasion research

had empirically tested the persuasiveness of a CSR campaign based on the effect of

message sidedness. Therefore, a research question was proposed. To answer this research

question, a one-way ANOVA test with perceived CSR as a dependent variable was

undertaken. The test yielded a significant message sidedness effect, F (3, 201) = 15.74, p

< .001.









The results in Table 4-6 and Table- 4-7 show that there were significant mean

differences between the treatment group and the control group. The mean score of overall

perceived CSR for the control group (M=4.05) was significantly lower than those of

treatment groups (M One-sided = 5.39; MNR Two-sided = 4.61; MR Two-sided = 4.86,

Scheffe post-hoc test, p <. 05). In other words, participants' perceptions about the

company's corporate social performance were significantly different from that in the

control group who did not receive CSR campaigns.

Table 4-6. Perceived CSR by message sidedness
Message Sidedness
Control (N=21) One- NR R
Sided Two- Two-
(N=61) Sided Sided
(N=62) (N=61)
Overall Perceived CSR 4.00 5.51 4.61 4.86
(.62) (.97) (.88) (1.15)
Note. 1. NR stands for "nonrefutational." R stands for "refutational."
2. Standard deviation in parentheses.
3. The perceived CSR index shows the average of five items on a 7-point scale, with
higher value indicating higher perceived CSR.

Table 4-7. F-test of perceived CSR by message sidedness
Df MS F
Overall Perceived CSR Between groups 3 15.01 15.74*
Within groups 201 .95
Note. 1. NR stands for "nonrefutational." R stands for "refutational."
2. *** p< .001

In particular, significant mean differences were also found between the scores of

perceived CSR among the three message sidedness groups. Table 4-8 shows the mean

score of perceived CSR for the one-sided message group is significantly higher than that

of two-sided message groups. It is noteworthy that participants of all the three groups

received the same CSR message(see Appendix C & D) while only that in two-sided

message groups received additional arguments on the clearcutting issue (see Appendix E









& F). Moreover, participants in nonrefutational two-sided message groups rated the

lowest score on perceived CSR among all the three groups. Participants in the refutational

two-sided message groups, on the other hand had a lower score on perceived CSR than

the one-sided message group.

However, the finding is contrary to previous research. These results indicated that

the one-sided CSR message is the most persuasive in influencing participants'

perceptions about the company's corporate social performance.

Table 4-8. Perceived CSR by message sidedness
Sidedness Subset for Alpha .05
N 1 2 3
Control 21 4.00
NR Two-Sided 62 4.61
R Two-Sided 61 4.86
One-Sided 5.51
Note. 1. NR stands for "nonrefutational." R stands for "refutational."
2. Displayed are means for groups in homogeneous subset.
3. Scheffe post-hoc test was performed; p < .05.

Interaction Effect on Perceived CSR

As results of Table 4-9 and Table 4-10 show, a significant interaction effect was

found only when the participant is in the low involvement condition. The results

indicated that perceived CSR was more positive in a one-sided condition (M=5.85,

SD=.82) than in the remaining conditions (Mfor NR TWO = 4.57, Mfor R TWO=4.70)

under a low involvement condition, however the mean difference between the

refutational two-sided group (M = 5.02) and the one-sided group (M = 5.18) decreased

under a high involvement condition. The mean score of the nonrefutational two-sided

group remained relatively unchanged by the two involvement levels. In short, the









interaction effect of involvement and message sidedness on perceived CSR was

statistically significant, F (2,178) = 4.17, p < .05.

Table 4-9. Interaction of involvement and message sidedness on perceived CSR
Message Sidedness Low Involvement High Involvement
N M(SD) N M(SD)
ONE-SIDED 30 5.85 (0.82) 31 5.18(1.01)
NR TWO 31 4.57 (0.89) 31 4.65 (0.88)
RTWO 31 4.70 (1.14) 30 5.02(1.09)
Note. 1. NR stands for "nonrefutational." R stands for "refutational."
2.p <.05

Table 4-10 Tests of between-subject effects (dependent variable: perceived CSR)
MS df F
Intercept 4591.72 1 4656.13 ***
Involvement .38 1 .38
Message Sidedness 13.40 2 13.59 ***
Involvement* Message Sidedness 4.11 2 4.17 *
Error .99 178

Note. 1. < .05; *** p <.001
2. R2 =.17

The Figure 4-1 clearly shows the differences in the interaction effects of message

sidedness on perceived CSR when the participant is in the low involvement and the high

involvement conditions. These findings indicated that a one-sided CSR message appeared

to be most persuasive in influencing people's perceptions about a company's corporate

social performance. Particularly, the persuasion effect of a one-sided CSR message is

more obvious when the message recipient is in a low-involvement situation. However,

the advantage of persuasiveness for one-sided message decreased when involvement was

high.







58




An interaction of message sidedness and involvement on perceived CSR



6.4
6.2
6
5.8
5.6
5.4 One-sided
5.2 R Two-sided
SNR two-sided

4.8
4.6 _
4.4
4.2
4
Low High
involvement


Figure 4-1. Interaction of involvement and message sidedness on perceived CSR.

Attitudes toward a Company

Hypothesis 3: Participants exposed to a refutation two-sided message would show

more positive attitudes than those who were exposed to a one-sided or a non-

refutational two-sided message. (Predicted main effect)

Another important dependent variable that was widely studied by researchers was

participants' attitude toward a brand or a company. To test this Hypothesis 3, a one-way

ANOVA test was applied to the overall attitudes toward CALCO, followed by the

Scheffe homogeneous subset comparison procedure upon significance level at c= .05.

Table 4-11 and Table 14-2 demonstrate that results from the ANOVA test turn out to be

contrary to the prediction of Hypothesis 3.









Contrary to Hypothesis 3, it was found that the one-sided message group showed

the most positive attitude toward CALCO with M=5.82, SD=1.03. Results from the one-

way ANOVA test revealed a significant main effect of message sidedness on attitudes

toward CALCO, F (3, 200) = 20.56, p < .001.

Table 4-11. Attitudes toward CALCO by message sidedness
Message Sidedness
Control One-sided NR TWO R TWO
(N=21) (N=60) (N=62) (N=61)


Attitudes Toward CALCO 4.18 (.63) 5.82 (1.03) 4.73 (.96) 4.75 (1.11)

Note. 1. NR stands for "nonrefutational." R stands for "refutational."
2. "Two" indicates "two-sided."
3. Standard deviation in parentheses.
4. The attitude index shows the average of six items with a 7-point semantic differential
scale, with higher value indicating more positive attitudes toward CALCO.

This finding again suggested that one-sided CSR messages have an advantage in

persuasiveness over refuational two-sided messages and non-refutational two-sided

messages. According to the findings of the current study, one-sided CSR messages

appear to be more persuasive in communicating CSR activities, in terms of generating

positive attitudes toward the company and inducing higher perceptions of the company's

corporate social performance. In other words, a one-sided CSR campaign presenting

simply the positive information related to the social activities of the company generated

more positive attitudes toward the company than a two-sided CSR campaign.

Table 4-12. F-test for attitudes toward CALCO by message sidedness
Df MS F
Attitudes toward CALCO Between groups 3 20.77 20.56***

Within groups 200 1.01
Note. 1. NR stands for "nonrefutational." R stands for "refutational."
2. *** p <.001









Since the main effect for control of message sidedness on attitudes toward the

company was significant, a subsequent post-hoc analysis with Scheffe's homogeneous

subset test showed that the effects of message sidedness on the attitudes toward CALCO

could be attributed to the differences between the one-sided message group and the two-

sided message groups (see Table 4-13). The mean attitude score of the one-sided message

group (M = 5.82) was significantly higher than that of both the nonrefutational two-sided

message group (M= 4.73) and the refutational two-sided message group (M = 4.75) under

p <.05.

Table 4-13. Attitudes toward CALCO by message sidedness
Sidedness Subset for Alpha .05
N 1 2
Control 21 4.18
NR TWO 62 4.73
RTWO 61 4.75
One-sided 5.82
Note. 1. NR stands for "nonrefutational." R stands for "refutational."
2. Displayed are means for groups in homogeneous subset.
3. Scheffe post-hoc test was performed; p < .05.

These findings indicated that significant differences were found between the one-

sided message and the two-sided message (refutational two-sided and nonrefutational

two-sided) in terms of the effects on perceived CSR and attitudes toward the company.

However, no significant differences were found between the refutational two-sided

message and the nonrefutational two-sided message. The results of the current study

might explain why the mainstream of message sidedness research mainly focused on one-

sided and two-sided messages, instead of comparing two types of two-sided messages.

Interaction Effect on Attitude toward a Company

Hypothesis 4: Participants in a low involvement condition will show more positive

attitudes n i/h a one-sided message while those in a high involvement condition









would express more positive attitudes i i/h a refutational two-sided message.

(Predicted interaction effect on attitude)

In Hypothesis 4, an interaction effect of message sidedness and involvement on

attitude toward CALCO was predicted. It was more specifically hypothesized that

participants in a low involvement condition would show more positive attitudes to a one-

sided message while those in a high involvement condition would express more positive

attitudes to a refutational two-sided message.

Table 4-14 and Table 4-15 revealed that there was no significant interaction effect

of message sidedness and involvement on attitudes toward the company. Therefore,

Hypothesis 4 was rejected. There was only a main effect of message sidedness on attitude

toward the company as confirmed in the earlier analysis. The one-sided message resulted

in more positive attitude toward CALCO both in a high involvement and in a low

involvement condition than other remaining conditions.

Table 4-14. Interaction of involvement and message sidedness on attitudes


Message Sidedness Low Involvement High Involvement
N M (SD) N M (SD)
One-Sided 30 5.93 (1.16) 30 5.71 (0.90)
NR Two-Sided 31 4.45 (0.98) 31 5.01 (0.97)
RTwo-Sided 31 4.70(1.27) 30 4.75(1.11)


Note. 1. NR stands for "nonrefutational." R stands for "refutational."
2. "Two" indicates "two-sided."
3. Standard deviation in parentheses.
4. The attitude index shows the average of six items with a 7-point semantic differential
scale, with higher value indicating more positive attitudes toward CALCO.









Table 4-15. Tests of between-subject effects (dependent variable: attitudes)
MS df F
Intercept 4758.61 1 4474.25 ***
Involvement 1.00 1 .94
Message Sidedness 23.31 2 21.92 ***
Involvement* Message Sidedness 2.39 2 2.25
Error 188.25 177
Note. 1. *** p < .001 2. R2= .22

A sub Analysis between Two Sample Groups

The participants of the present experiment were recruited from two undergraduate

class from two different colleges: the College of Agricultural and Life Sciences and the

College of Journalism and Communications. It is appropriate to assume that there might

be some differences between the two groups of participants because of the majors of

students in the College of Agricultural and Life Sciences are related to environmental

issues and therefore assumed to be more environmental concerned. A sub analysis was

run between two groups of participants to test the above four hypotheses. However, the

results did not indicate any significant statistical differences.

A Correlation between Perceived CSR and Attitudes toward a Company

Hypothesis 5: Higher perceived CSR would be positively associated i iih more

positive attitude toward CALCO.

As a test of this hypothesis, a correlation analysis was conducted. In order to

determine any significant relationships between perceived CSR and attitudes toward the

company, the Pearson correlation was undertaken. Pearson correlations coefficient is a

measure of association which caries from -1 to +1, while 0 indicating no linear

relationship, -1 indicating a perfect negative linear relationship and +1 indicating a

perfect linear relationship (Garson, 2005).









In the Table 4-16, the Pearson correlation coefficient shows that there is a positive

correlation between perceived CSR and attitude toward a company. Thus, Hypothesis 5

was supported. This finding indicated that perceived CSR was highly associated with

attitudes toward the company. In other words, when a person perceives a company as

socially responsible, it was likely that he or she would have more positive attitude toward

the company.

Table 4-16. A Pearson Bivariate Correlation between perceived CSR and attitudes
Perceived CSR Attitudes
Perceived CSR 1.00
Attitudes .697 ** 1.00
Note. 1. ** p <.01 (2-tailed).

A Hierarchical Regression Analysis between Perceived CSR and Attitudes toward a
Company

To look into the relationship between perceived CSR and attitude toward the

company, a hierarchical regression analysis was performed (see Table 4-17).

"Attitudes toward CALCO" was the dependent variable for this hierarchical

regression model. In this hierarchical regression analysis, three blocks of variables were

entered into the hierarchical regression analysis

The three blocks of variables consisted of "participants' demographics,"

"experimental variables (i.e., message sidedness and involvement)," and "mediators."

Participants' demographics consisted of age, gender, and ethnicity: gender and ethnicity

were dummy-coded (Ml). The experimental variables of involvement and messages

sidedness were also entered as dummy variables (M2). Finally, in order to examine

mediating impacts of cognitive elaboration and perceived CSR on the dependent variable,

the total number of message-related thoughts and perceive CSR were entered into the

third step (M3). The regression model helped verify the hypothesis by comparing the










standardized coefficient (fl) and R2. Standardized regression coefficients (/) and R-

square changes (AR2) in explained variance were examined.

Table 4-17. Hierarchical regression analysis for proposed model
Attitude to\\ ard the coimpani as a DV_1
Ml M2 M3
Std. Std. Std.
Error Error Error
Demographics
Age -.01 .08 .08 .07 .00 .06
Gender (female)b -.01 .08 .01 .07 -.02 .06
Ethnicity (White)b -.00 .08 -.01 .07 .01 .05
Experimental
Variables
One-sidedb .52*** .08 .21** .07
Refutational Two- .08 -.02 .07
sided .13 .08 -.02 .07
sidedb
Involvement: Highb .05 .07 .09 .05
Mediators
Cognitive .02 .05
Elaboration
Perceived CSR .63*** .06
Constant 5.11** 3.80** 1.50**
R2 .00 .21 .54
Adjusted R2 -.02 .18 .52
AR2 .00 .21 .33*
F for Regression .02 7.68*** 25.13
Notes: Standardized regression coefficients (f,) are shown; Std. Error means standard
errors.
a. DV = dependent variable
b. dummy coded
Mean for the items means the average score for the 7 items in the Fortune reputation
index.
p<.01; *p< .01

In the regression analysis, three demographic variables were entered first. No

significant influence was found from entering this block (i.e., Ml) of variables. Since

these demographics were considered as control variables in this experiment, it could be









interpreted that individual differences that may hurt the results of this experiment were

strongly controlled. The second bloc of the analysis (M2) observed a significant impact of

message sidedness (i.e., one-sided message) on the dependent variables of attitudes

toward CALCO (f= .52, p <. 001). The refutational two-sidedness entered as a dummy

variable did not contribute to the dependent variable, nor did involvement manipulation.

The experimental variables entered in step 2 accounted for 21 percent of the variance (p <

.001).

In the final step of analysis (M3), the entire variables were entered to look into a

relative contribution of each variable to the dependent variable. The beta coefficient of

message sidedness (8 = .21, p < 01) was reduced but it was still significantly associated

with attitudes toward CALCO, the dependent variable. Therefore, the one-sided message

could be interpreted as a reliable contributor to positive attitudes toward CALCO. The

results also showed that perceived CSR could be an effective predictor for achieving

positive attitudes toward CALCO (/ = .63,p < 001). Therefore, Hypothesis 5 that

predicted a positive relationship between perceived CSR and attitudes toward the

company was corroborated.

In order to be perceived as socially responsible, the findings of this study suggest

that companies should emphasize not only practicing CSR but also communicating its

corporate social performance to their stakeholders. In brief, the findings of this study

demonstrate the effectiveness of CSR campaigns on enhancing people's perceptions of

the corporate social performance and attitude toward the company. More specifically, in

terms of CSR communication strategy, one-sided messages presenting supportive









arguments for the company's social performances are more persuasive in influencing

people's perceptions in a desired direction regardless of involvement level.

Path Analysis

The researcher proposed a cognitive processing model of a CSR message in this

study (see Figure 4-2). However, no significant difference of the numbers of message

related thoughts was found between treatment groups. The assumption related to the main

effects and interaction effects of message sidedness and issue involvement on

participants' levels of cognitive elaboration was rejected. Therefore, in order to revise the

cognitive processing model of a CSR message, an additional path analysis was

undertaken to explore the causal relationships between message sidedness, involvement,

perceived CSR and attitude toward the company.

Moreover, in response to the research call for effective communication methods of

CSR (Clark, 2000), it is essential to demonstrate the impact of communicating CSR

activities or commitment on achieving organizational goals. Based on the notion that

effective CSR communication should have positive impact on organizational goals, the

present study attempts to explore the causal relationships between the perceptions of a

company's corporate social performance and the attitudes toward the company by

conducting path analysis.

Path analysis is a statistical procedure consisting of a series of regression analyses.

The key advantage of using the path analysis approach in this research study is that the

researcher can demonstrate the causal relationships between variables so that in turn the

theories and relationships among variables can be clarified and strengthen (Agresti &

Finlay, 1997).









The equations for a five-variable path model were depicted in Figure 4-3. Based on

the equations, the researcher tested direct and indirect effects of exogenous variables and

different patterns of CSR messages and involvement level on both attitude toward the

company and perceived CSR. It is assumed in this study that perceived CSR has a

mediating effect on attitude toward the company.

The results of the path analysis suggest that the one-sided message has both

significantly direct effects on perceived CSR (P=.38, p< .001) and on attitude toward the

company (P=.43, p< .001). Refutational two-sided message also has a direct effect on

perceived CSR (P=. 18, p< .05), but only marginal impact on attitude toward the company

(P=. 10, p< .10).The path analysis also indicated that a significant direct effect between

perceived CSR and attitude toward the company (P=.65, p< .001). That is, the higher the

person perceived a company's corporate social activities, the more likely to the person

will induce positive attitudes toward the company.

However, Figure 4-3 shows that involvement level did not link to either perceived

CSR or attitude toward the company. In other words, according to the results of the path

analysis, involvement level does not affect people's perceptions of a company's corporate

social performance and their attitudes toward the company.

The most important findings in the causal model were related to the power of

perceived CSR in predicting attitude toward the company. Consistent with the hypothesis

of this study, it was found that perceived CSR was highly associated with attitude toward

the company. The higher participants perceived the corporate social performance of a

company, the more they had positive attitudes toward the company. In this regard,

communicating a company's social performance is as important as conducting corporate









social activities. Therefore it is needed to address a communication management

perspective of corporate social responsibility in order to include CSR as part of a business

strategy.

Another interesting finding from this path analysis is the effects of message

sidedness on perceived CSR and attitude toward the company. The path coefficients of

message sidedness in the path model show that a one-sided message has a direct impact

on perceived CSR and attitude toward the company. In other words, a one-sided CSR

campaign which presents only positive information has a significant advantage over a

two-sided CSR campaign in influencing message recipients' perceptions of the

company's social performance and their attitudes toward the company.

The involvement variable however, has no direct effect on perceived CSR nor has it

on attitudes toward the company. Regardless of involvement level, a CSR campaign has

the same effect on message recipients' perceived CSR and their attitudes toward the

company. Therefore, according to the results of the path analysis, a revised cognitive

processing model of a CSR message is presented in Figure 4-4. The effect of message

sidedness is highly associated with both perceived CSR, and attitude toward the

company.

In brief, the results of the path analysis show that five paths, all direct, are

statistically significant (see Figure 4-3): from a one-sided message to perceived CSR

(P=.38, p< .001), from a one-sided message to attitude toward the company (P=.43, p<

.001), from perceived CSR to attitude toward the company (P=.65, p< .001), from

refutational two-sided message to perceived CSR (P=. 18, p< .05), from refutational two-

sided message to attitude toward the company (P=. 10, p< .10).



























Figure 4-2. The original cognitive processing model of a CSR message proposed in this
research.


Figure 4-3. The path diagram based on the path analysis.

Note. 1. Message sidedness and involvement were dummy coded. 2. t p < .10, p < .05,
*** p< .001 3. n.s. represents nonsignificant






70




Message Perceived CSR Attitude toward
Sidedness No P- the Company



Figure 4-4. The revised cognitive processing model of a CSR message proposed in this
research.














CHAPTER 5
DISCUSSION

Over View of the Present Study

The present study represents an innovative experiment of a cognitive processing

model of a CSR message. The purpose of this study is to test empirically the persuasive

effect of a CSR campaign in influencing participants' attitudes toward a company and

perceived CSR regarding involvement level and the effect of message sidedness. In

particular, the relationships of a CSR campaign on attitudes toward the company via

message recipients' perceptions of the company's social performance were verified in

this study.

By testing the persuasiveness of a CSR campaign and proposing a cognitive

processing model, this study sheds light on the research of CSR communication and

persuasion. Even though previous research has discussed CSR from the issues

management perspective (Carroll, 1991; Clark, 2000; Esrock & Leichty, 1998; Heath,

1998; Heath & Ryan, 1989; L'Etant, 1994; Wartick & Cochran, 1985; Wartick & Rude,

1986), the discussions about the effects of a CSR campaign on public perceptions of a

company's social performance were missing.

To fill this research lacuna, this study explored the persuasiveness of a CSR

message in influencing public perceptions of a company's social performance and

attitudes toward the company. The present study asked whether presenting a one-sided

CSR campaign or a two-sided CSR campaign is more persuasive in advocating a

company's corporate citizenship and addressing the controversial issues related to the









fundamental practice of the business. The two-sided CSR campaigns were further broken

down into two different forms: 1) a nonrefutational two-sided campaign in which the

opposing messages were shown without additional refutations, and 2) a refutational two-

side message in which the opposing arguments were shown with refutational

explanations.

From the public relations perspectives, this study contributed to explore the

effectiveness of CSR campaigns as proactive tools of issues management. The main

objective of the present study was to test whether attitude toward the company can be

influenced by people's perceptions of the company's corporate social performance.

Moreover, this study also explored the functions of a CSR campaign not only in

advocating the corporate citizenship of the company but also preventing crisis situations

by actively addressing controversial issues with external stakeholders.

Little research had empirically examined the effectiveness of a CSR campaign. The

present study attempted to explore this uncharted research area with relevant theories in

persuasion, such as the cognitive response research and the ELM. In addition, this study

tested the causal relations between perceived CSR and attitude toward the company.

The experimental design of this study can provide some practical implications to

public relations practitioners who have to communicate with active publics concerning a

critical issue. According to Grunig's (1997) situational theory, the active public is

identified as a high involvement group. It is assumed in this study that if participants are

highly involved with the issue, they would generate more message-related thoughts

particularly when they were exposed to a refutational two-sided message.









In addition, another important assumption related to involvement level was

message recipients' cognitive elaboration of the CSR messages. Based on the ELM, it

was reasoned in this study that an individual's perceptions about a company's corporate

social performance and attitudes toward the company would be mediated by his or her

motivation to process the message and the issue presented in CSR campaigns. However,

findings in this study did not support the assumption that message sidedness and issue

involvement have impact on participants' cognitive elaboration level of processing a CSR

message.

Even though there were no significant differences between the numbers of

recipients' message-related thoughts, it was found in this study that a one-sided CSR

campaign has an advantage of persuasiveness over both refutational two-sided and

nonrefutational two-sided CSR campaigns. The findings in this study indicated that

presenting only supportive messages about a company's corporate social activities will

induce more positive attitudes toward the company. Regardless of the involvement level,

people are likely to form a positive attitude toward a company when the company is

considered as socially responsible.

Contrary to previous findings on the effect of message sidedness, it was found in

this study that a one-sided CSR message had generated more positive attitudes toward the

company and more positive perceptions of the company's corporate social achievements

than two-sided one. In addition, this study also found that the perceived CSR was

positively associated with attitudes toward the company (r = .697), implying that the

perceived CSR might play a mediating role in the process of attitude formation when a

person was exposed to a CSR advocacy message.









Overview of the Hypotheses and Research Question

In the following section, each of the five hypotheses and one research question is

discussed in detail based on the results of this study.

Hypothesis 1: Participants in a high involvement condition will generate more

message-related thoughts than those in a low involvement condition

The first hypothesis investigated the main effect of involvement level on cognitive

elaboration of the CSR messages presented in the experiment. To test this hypothesis, a t-

test was conducted. However, the results showed no statistical differences of message-

related thoughts between a high and a low involvement group (p= .398). Thus,

Hypothesis 1 was rejected.

Even though a successful manipulation check of involvement was obtained in this

study, participants in high involvement groups did not demonstrate higher level of

cognitive elaboration as previous research has indicated. The results showed that

regardless of participants' involvement level, the numbers of message-related thoughts

generated by participants in high involvement conditions were not significantly different

from that in low involvement conditions.

One explanation of these results is that the manipulation of involvement used in

this experiment did not affect respondents' motivation to process the messages presented

in the stimuli booklets.

In this study, participants' level of involvement was manipulated by the location of

the lumber company presented in the message stimulus. For high involvement groups, the

lumber company presented in the stimuli was located in Florida, whereas the lumber

company was located in Taiwan for low involvement groups. Involvement level was









measured using four questions regarding participants' concerns about the environmental

issues in Florida or in Taiwan.

Even thought the manipulation check of involvement worked successfully in this

experiment, it did not show that participants' motivation to cognitively processing the

CSR messages was also manipulated by the stimuli. Therefore, it is appropriate to assume

that the amounts of effort participants invested in reading the CSR messages and

answering the questionnaire were not influenced by the involvement manipulation

presented in the stimuli materials. Participants in the high involvement group were not

necessary more motivated to process the stimuli even though their responses to the

manipulation check were significantly different from that in the low involvement group.

Therefore, to test the cognitive elaboration, future research should consider other

approaches to enhance participants' motivation in responding experimental materials.

Hypothesis 2: A two-sided refutational message will induce more message-related

thoughts than a one-sided or a nonrefutational two-sided message.

To test Hypothesis 2, a One-Way ANOVA test was run. In this experiment, no

statistically significant differences were found between the numbers of message-related

thoughts from different message sidedness groups. Thus, Hypothesis 2 was also rejected.

For cognitive elaboration, there were no significant effects attributed to involvement and

message sidedness.

However, a previous research by Hale et al. (1991) found that more positive

cognitions were generated by refutational two-sided messages than nonrefutational two-

sided messages and one-sided messages. Contrary to previous study, this experiment

found that one-sided CSR message had induced more message related thoughts than two-









sided CSR messages. One explanation may be that the total number of message-related

thoughts was considered for measuring cognitive elaboration in this study, instead of the

number of positive and negative thoughts.

In conclusion, these results failed to support the main effects of message

sidedness and involvement on cognitive elaboration. Further analysis is needed to revise

the cognitive elaboration model of a CSR message proposed in the beginning of this

study. Therefore, path analysis, regression analysis and Pearson correlation analysis were

conducted in this study and discussed in the following section to explain the causal

relationships between the three variables in the revised cognitive elaboration model of a

CSR message.

RQ1: How does message sidedness influence people's perception about corporate

social responsibility of a company? What will lead to more positive perceived CSR

between a two-sided message and a one-sided message?

To answer this research question, a one-way ANOVA test with perceived CSR as a

dependent variable was undertaken. The test yielded a significant message sidedness

effect and the results showed that participants' perceptions about the company's

corporate social performance were significantly influenced by CSR campaigns. In

particular, these results indicated that the one-sided CSR message is the most persuasive

in influencing participants' perceptions about the company's corporate social

performance.

A significant interaction effect was also found in this study. However, the

interaction effect was limited to the low involvement condition. The results indicated that









one-sided CSR message had induced more positive perceived CSR in the low

involvement condition.

In conclusion, the present study confirmed previous studies that demonstrated a

relationship between CSR communication and perceptions about a company's corporate

social performance. The CSR campaign designed for the present experiment appears to

represent the majority of CSR initiatives and activities, because the three areas of CSR

activities chosen for research, environment, employees and community, are the focus of

CSR initiatives and activities for the majority of today's corporations.

Hypothesis 3: Participants exposed to a refutation two-sided message would show

more positive attitudes than those who were exposed to a one-sided or a non-

refutational two-sided message. (Predicted main effect)

To test this Hypothesis 3, a one-way ANOVA test was applied to the overall

attitudes toward the company. Results revealed a significant main effect of message

sidedness on attitudes toward the company (p < .001). However, contrary to the

prediction, this finding again suggested that one-sided CSR messages have an advantage

in persuasiveness over refuational two-sided messages and non-refutational two-sided

messages.

Hypothesis 4: Participants in a low involvement condition will show more positive

attitudes i/ i/t a one-sided message while those in a high involvement condition will

express more positive attitudes i/ i/h a refutational two-sided message. (Predicted

interaction effect on attitude)

The results indicated that no interaction effect of message sidedness and

involvement on attitudes toward the company was found. Therefore, Hypothesis 4 was









rejected. There was only a main effect of message sidedness on attitude toward the

company as confirmed in the earlier analysis.

Two main effects of message sidedness were found on perceived CSR and

attitude toward the company, while only one interaction effect between message

sidedness and perceived CSR was found when participants' involvement levels were low.

In conclusion, the results to above hypotheses showed that (1) generally the effect of

message sidedness on perceived CSR and on attitude toward the company were

significant, (2) contrary to previous research, one-sided CSR messages appear to have an

overall advantage of persuasiveness in influencing people's perceptions about a

company's social performance and their attitude toward the company, and (3) the

relationships between message sidedness and involvement on cognitive elaboration were

do found. Based on these findings, it can be suggested that the proposed cognitive

elaboration model of processing a CSR message can be revised by taking out the

cognitive elaboration.

Hypothesis 5: Higher perceived CSR will be positively associated i iili more

positive attitude toward CALCO.

As expected, the Pearson correlation coefficient shows that there is a positive

correlation between perceived CSR and attitude toward a company. This finding

indicated that perceived CSR was highly associated with attitudes toward the company. A

hierarchical regression analysis was performed to look into the relationship between CSR

messages, perceived CSR and attitude toward the company. The regression analyses

tested whether CSR messages and perceived CSR had significant effect on predicting









attitudes toward the company. According to the results, the one-sided message could be

interpreted as a reliable contributor to positive attitudes toward CALCO.

Using the path analysis, the researcher examined the effects of message

sidedness and involvement had on perceived CSR and attitude toward the company. As

expected, the analysis showed that both message sidedness had both direct effects on

perceived CSR and attitude toward the company, while no significant effects were found

between involvement and other variables. The results of path analysis also indicated the

direct relationship between perceived CSR and attitude toward the company is

statistically significant.

Path analysis confirmed five paths, all direct, are statistically significant: one-sided

message had two direct effects on perceived CSR and attitude toward the company, while

refutational two-sided message also had two direct effects on perceived CSR and attitude

toward the company, and refutational two-sided message had a direct effect on attitude

toward the company. Consequently, the present study proposed a model of relationship

between message sidedness, perceived CSR and attitude toward the company as the final

model.

In conclusion, the present study demonstrated that (1) CSR messages (one-sided

message and refutational two-sided message) were significantly associated with

perceived CSR, (2) perceived CSR was significantly associated with attitude toward the

company.

Although most of hypotheses were rejected except for the last one hypothesized

on the relationship between perceived CSR and attitude toward the company, some of

interesting and information were found. First, regardless of people's involvement level









with issues, the one-sided CSR message has an overall advantage of persuasiveness in

influencing people's perceptions about a company's corporate social performance and

their attitudes toward the company. Contrary to previous research, the results of the

present study reflected the conflicting findings of the mainstream research in message

sidedness.

Second, perceived CSR was found to have a significant effect on predicting

attitude toward the company. The present study had verified the relationship between a

CSR message, perceived CSR and attitude toward the company. The results of this study

suggest that CSR communication is highly associated with the attitude toward the

company. In particular, a one-sided CSR message presenting only positive information

about the company's social commitment and activities are most persuasive in positively

influencing the attitude toward the company.

Despite these unexpected results related to the message sidedness effect, the

present study supports the proposition that perceived CSR can function as an important

predictor of attitude toward the company. The present study concluded that CSR

campaigns had a directly effect on people' perceptions about the company's social

performance, while these perceived CSR also directly influence attitudes toward the

company.














CHAPTER 6
CONCLUSION

It has been a crucial topic in public relations research to explore the ways in which

public relations can contribute to organizational goals.

Clark (2000) indicated a set of similarities between CSR and public relations and

suggested that both disciplines are seeking to enhance the quality of the relationship of an

organization among key stakeholder groups. However, for both fields to solidify a

permanent place as a management function, more research into the value of their efforts

is needed.

Moreover, recent developments in the fields of CSR and public relations suggest

that reputation is an important area for future research to measure the value of both fields

(Clark, 2000). According to Fombrun (1996), organizations can build a sound reputation

through effective communication. Previous research in communication has looked

specifically into how best to communicate the social responsibility of an organization (J.

E. Grunig, 1986; Ledingham & Bruning, 1999). However, more specifically, a question

about the chosen messages presented in CSR campaigns and how they affect the

reputation or perception of an organization as responsible remain (Clark, 2000, p.376).

To overcome the research gap, the present study explored empirically the

persuasive effects of a CSR message on attitude toward the company and perceived CSR.

The effect of message sidedness and involvement level were also tested as a way of

designing an effective CSR communication strategy.









The results of this study indicated that presenting a one-sided CSR campaign which

presented only positive information about a company's social activities was more

persuasive in influencing public perceptions about the company's social performance.

Moreover, this study verified the relationships between the CSR message, perceived CSR

and attitude toward the company. The results of this study demonstrated that there were

direct effects between the CSR message, perceived CSR and attitude toward the

company.

The results of the present study not only responded to Clark's (2000) research call

on the effects of CSR messages but also verified the communication management

functions on CSR. It was showed in this study that an effective CSR campaign can

influence public perceptions about a company's social performance and induce positive

attitude toward the company. The present study had build on the knowledge of CSR

communication, cognitive response, and persuasive effect of message sidedness on CSR

activities.

Discussions about the Effect of Message Sidedness

A couple of previous studies found that a one-sided message had more persuasive

effect than a refutational two-sided message (e.g., Halverson, 1975; Smith et al, 1994).

O'Keefe (1998) had argued that the conclusion of the message sidedness effect should be

interpreted carefully by considering the message topics. He argued that: "given the

different effects of nonrefutational forms in advertising and nonadvertising contexts, one

should not too easily assume that refutational forms will function identically in the two

circumstances" (p. 238).

Even thought an abundant number of studies have shown superior persuasiveness

of refutational two-sided messages, other studies have found that one-sided messages can









be more persuasive under certain conditions, implying the existence of moderators in the

message sidedness effect. For future research on message sidedness, the possible

moderators of the effect of message sidedness were discussed.

From the very beginning of sidedness research, a number of possible moderators

have been proposed. Hovland et al. (1949, p.225), for example, suggested that the

audience's educational level is an important determinant of the consequences of

sidedness variations. Several variables such as the message recipient's initial attitude

toward the source's position, topic familiarity (Allen, 1991), and intelligence (Hovland et

al., 1949; Allen et al., 1990) are commonly mentioned as moderators of the persuasive

effects of sidedness variations. Other proposed moderators have included perceived

source motivation (Pechmann, 1990), and exposure to subsequent opposing

communications (Lumsdaine & Janis, 1953). Specifically, one-sided messages were

posited to be more effective than two-sided messages when the recipient initially agreed

with the source's position, while being unfamiliar with the topic, or unintelligent (Allen

et al., 1990).

The following sections are discussions about the moderators of the effect of

message sidedness. Even though these variables were controlled or in this experiment,

previous studies had indicated that they may have an impact on the persuasiveness of the

message sidedness which required some attention for future research.

Credibility Perceptions

Credibility perceptions are very important to the interpretation of the findings from

this study. According to Allen et al.'s (1990) meta-analytic review, the credibility

perception about the communicator was reported to play an important role in a persuasive

process. In another comprehensive meta-analysis of the message sidedness effect by









O'Keefe's (2000), the credibility perceptions were discussed in terms of a mediator in a

persuasive process.

The variable of source credibility was controlled in this experiment. All the

participants received CSR messages from the same source: the CALCO's company Web

site. Previous research indicates that refutational two-sided messages might evoke closer

message scrutiny, and the persuasive effect was discounted because the message source

could have been perceived to be less credible. In other words, participants may discount

the message if it is delivered from a less credible source or because the message is

considered as the one with a self-serving bias.

Regarding the impact of message-sidedness on persuasion, Chebat, Filiatrault,

Laroche, and Watson (1988, p.619) had an important finding. They found that the

attention span of negatively oriented recipients was more limited when the source was

less credible and the message was two-sided. Therefore, an unexpected finding regarding

the message sidedness in this study might be attributed to the impact of source credibility.

Pre-Attitudes toward an Issue

Previous research found that a one-sided message could be more persuasive when

the audience was positively predisposed (Etgar & Goodwin, 1982). As discussed

previously, a one-sided message is more persuasive to message recipients with positive

initial attitudes toward an issue while a two-sided message is more persuasive to those

who initially have negative attitudes.

In this experiment, some participants in the experimental conditions reported that

the background information about CALOC which is delivered on CALCO's "About Us"

page generated positive initial attitudes toward the company. Even though we tried to

deliver a brief company introduction, we had not anticipated that a well-designed









corporate Web page would have such an effect in forming a positive pre-attitude toward

the company.

Fear Arousal

Smith et al.'s (1994) study on the persuasive message regarding organ donation

found that a one-sided message was more persuasive than a refutational one. Their study

showed the effects of prior thoughts and intent on outcomes associated with persuasion.

They found that refutational messages induced significantly higher fear and anxiety in the

segment of audience who are lacking in prior thought and intent to donate an organ.

In other words, the decreased persuasive effects of refutational two-sided messages

in the present research might also be associated with unintended consequences such as

the recipients' avoidance of negative discussion. It was possible that participants would

have perceived the refutational messages to be self-serving, which in turn generates

negative attitudes toward the message itself and consequently toward the company.

Application

The current study responded to the research call of developing an effective method

in communicating CSR. The findings in this study suggest that a one-sided CSR message

may have an advantage in communicating the company's corporate social performance as

opposed to a two-sided one. More importantly, findings in this study also confirmed that

communicating corporate social responsibility enhanced people's perceptions about the

company's social performance and also generated more positive attitudes toward the

company advocated. From the positive role of a CSR campaign in generating positive

image and attitudes toward company, the CSR could be recommended as an effective tool

of building a long-term relationship between the public and an organization. Further

research examining the effect of CSR on building relationships is needed.









However, the conclusion of this study should be overstated because of several

limitations mentioned previously. Regarding the level of issue involvement, the

participants recruited in this study may mirror the general public who do not particularly

concern about environmental issues. The results of this study did not suggest that all CSR

campaigns should be presented as one-sided. On the contrary, for public relations

practitioner to design an effective CSR communication strategy, it is essential to consider

the characteristic of targeted audiences. For example, if the CSR campaign is targeted to

communicate with environmental active audiences, then the conclusion of this study

should be applied with caution. Future research is needed to explore the impact of CSR

campaign on different audiences

Limitations

This study examines only a small part of the process and the effects of CSR

communication with a limited group of people and limited stimulus message. The results

of this experimental study offer interesting and useful findings for public relations

practitioners and researchers. However, any generalization from the findings of this study

should be made with caution.

Despite a couple of limitations, this study could be improved with some

considerations. In this research, the total number of message-related thoughts was

considered for measuring cognitive elaboration, instead of the number of positive and

negative thoughts. Several studies in this cognitive response paradigm suggest that

valence of cognitive elaboration be considered. In the current research, the researcher did

not test whether a refutaitonal two-sided message generates more positive message-

related thoughts or it leads to negative cognitive elaboration. Future research can test the









cognitive process model of a CSR message with the consideration of valence of cognitive

elaboration.

Suggestions for Future Research

Knowledge Level

Previous research in persuasion has indicated that participants' educational levels

have an influence on their responses of a one-sided or a two-sided message (Chu, 1967;

Hovland, Lumsdanie, & Sheffield, 1949).

It has been found that lower education people would be more likely to respond to

one-sided messages. However, controlling for the educational level, the preset study still

found that one-sided CSR messages had an advantage of influencing attitude toward the

company and perceived CSR. The clearcutting issue discussed in the stimuli materials

was not a common topic in Florida areas. Therefore, it was assumed that most

participants in this study were not familiar with the issue and had a low knowledge level.

That is, because of the lower knowledge level about the clearcutting issue, participants in

this study were more likely to respond to a one-sided message, regardless of their

educational level.

Message Length

Apparently, another consideration is the length of message in this study. The

lengths of two-sided CSR messages are much longer than the one-sided CSR message.

Previous research suggested that the message length has an impact on the persuasive

effect of message sidedness. It was found that low involvement people are more likely to

respond to short message length. Therefore, for future research, it is suggested that the

length of messages should be controlled.









Source Credibility

As mentioned earlier, the source perception of the CSR message may have

influenced the message effect. Tit is possible that the audiences will discount the

corporate source, presumably because they thought that the arguments presented from the

corporate source at best served for self-interests.

Source credibility is a key role in persuasion and has been studied widely by

researchers in different fields, such as marketing, advertising and social psychology.

Future research could advance the current research with a consideration of source

credibility manipulation. For instance, a future research design might use an experimental

study that presents the same message used in the current study from either a high credible

source (e.g., from a third party or a researcher in forestry) or the company.

Attitudes toward Issues

From an issues management perspective, future research may add some

considerations on the interaction between people's attitudes toward an issue and attitudes

toward the company that advocated the issue. An additional dependent variable such as

the attitudes toward the issue may be included in an experimental study via a pre-and

post- test, so that an attitudinal change can be measured.

The present research has shed light on developing an effective communication

strategy on CSR based on the message sidedness effect. However, in order to fully

understand the persuasive effects of a CSR campaign on people's perceived CSR,

attitudes toward the company, and even the long-term relationship building between the

company advocated and stakeholders, additional primary research is needed concerning

the sidedness effect on CSR messages.