Abstract

This study develops and tests operational definitions of relational maintenance strategies appropriate to online public relations. An experiment was designed to test the new measures and to test hypotheses evaluating potential advantages of organizational blogs over traditional Web sites. Participants assigned to the blog condition perceived an organization's "conversational human voice" to be greater than participants who were assigned to read traditional Web pages. Moreover, perceived relational strategies (conversational human voice, communicated relational commitment) were found to correlate significantly with relational outcomes (trust, satisfaction, control mutuality, commitment).

Introduction

Web logs, or blogs, offer a unique channel for developing and maintaining relationships between organizations and publics. The purpose of this study is to triangulate professional literature on online communication, scholarship on relational maintenance strategies and relational outcomes, and quantitative data to explore the potential of blogs as tools for public relations.

The centrality of relationships in building public relations theory is clearly apparent in recent scholarly literature, and this interest is not limited to the ivory tower. Public relations professionals also have taken interest in relationships as the foundation for their work and have sought ways to measure their contributions as such.

Defining Relationships

The term relationship in the context of public relations has been defined as follows:

…the state which exists between an organization and its key publics in which the actions of either entity impact the economic, social, political and/or cultural well-being of the other entity. (Ledingham & Bruning, 1998, p. 62)

Relationships consist of the transactions that involve the exchange of resources between organizations. (Broom, Casey, & Ritchie, 1997, p. 91)

…routinized, sustained patterns of behavior by individuals related to their involvement with an organization… Many online relationships operate in tandem with offline relationships and thus are part of a total organizational-public relationship. (Hallahan, 2004, p. 775)

How can relationships be measured? Many public relations scholars have sought an answer to this question (Broom et al., 1997; Grunig & Huang, 2000; Hallahan, 2004; Hon & Grunig, 1999; Ledingham, 2003; Ledingham & Bruning, 1998). Although they all have slightly different ways of defining the concept, they all seem to agree that the study of relationships can be broken down into antecedents, maintenance processes, and outcomes. This study focuses on blogs as antecedents to people's perceptions of a corporation's relational maintenance strategies. Perceptions of relational maintenance strategies have been posited to lead to better relational outcomes such as trust, commitment, control mutuality, and satisfaction.

Defining Blogs

Merriam-Webster named "blog" their word of the year for 2004, and defined it in their Collegiate Dictionary, Eleventh Edition as "a website that contains an online personal journal with reflections, comments, and often hyperlinks provided by the writer" (www.m-w.com). Blogs also have been defined empirically as "frequently modified web pages in which dated entries are listed in reverse chronological sequence" (Herring, Scheidt, Bonus, & Wright, 2004, p. 1).

Herring et al. (2004) analyzed a random sample of 199 blogs available in May 2003 from the blog-tracking service blo.gs and found a large majority (70%) of them could be classified as personal-journal type blogs. They described the personal-journal genre as comprising blogs "in which authors report on their lives and inner thoughts and feelings" (p. 6). This finding contrasted a widely held conception that most blogs are link-based filters of Web content, what the authors call a filter-type genre. Some 13% of the blogs analyzed were classified as filter blogs.

Defining Organizational Blogs

The current study focuses on organizational blogs as means of communication between organizations and their publics. Sifry's (2004) concept of "corporate bloggers" provides a starting point:

These are people who blog in an official or semi-official capacity at a company, or are so affiliated with the company where they work that even though they are not officially spokespeople for the company, they are clearly affiliated. (Sifry, 2004)

Sun Microsystems' blogs.sun.com is a good example. The site is described in a single sentence on its home page as a space "accessible to any Sun employee to write about anything" (Sun Microsystems, 2005). This description is slightly qualified by Sun President and CEO Jonathan Schwartz's own blog entry in 2005:

As you know, I'm a huge believer in 'opening' Sun to the world—that's what this blogs.sun.com infrastructure is all about—transparency is one our best competitive weapons. But the unauthorized sharing of Sun confidential information is illegal, and against company policy. (Schwartz, 2005)

Microsoft also actively encourages employees to blog at blogs.msdn.com. Microsoft was the single largest host of corporate blogs in October 2004 according to data collected by Sifry, who mused that "that there is still a tremendous opportunity for forward-thinking companies and management to have a significant positive impact on their public perception" by way of hosting blogs (2004).

The blogs hosted by Microsoft and Sun both are more precisely termed "corporate blogs" than "organizational blogs," but they represent a trend in blogging that requires a definition that encompasses all types of organizations, not just large corporations. Although the personal nature of blog entries, as defined by Merriam-Webster and found to be most common by Herring et al. (2004), has led some to question whether the term "corporate blog" is an oxymoron, Herring et al.'s genre analysis helps describe how corporate communication might evolve from prior forms of communication ranging from offline corporate communication to online personal journals. They suggested that online forms of communication such as blogs might be better recognized as hybrid genres than as new forms of communication altogether.

Herring et al. classified a small portion (17%) of the blogs in their sample as either "knowledge blogs," "mixed," or "other" (2004, p. 6). Knowledge blogs are the online equivalent of professional journals in which authors document new knowledge in their professional domains, including research progress, references, and observation. Extending this concept beyond academic and scientific fields, organizational blogs also seem to appear at the intersection of personal reflection and professional communication. Organizational blogs have evolved from both online and offline modes of communication and have characteristics of both personal and professional communication.

Markets as Conversations

In The Cluetrain Manifesto (2001), Doc Searls, editor of Linux Journal, and David Weinberger, fellow at the Berkman Center for Internet and Society at Harvard Law School, examined the competing interests of "business as usual" and online communities that are home to mainly personal-journal and filter-type communications. Searls and Weinberger suggest businesses take a "markets-as-conversations" approach in adapting to the demands of emerging online communities. That is, they recommend veering away from business strategies that treat online communities as targets. This shift in approach from markets as targets to markets as conversations parallels public relations theory that calls for building collaborative relationships with publics rather than targeting them (e.g., Grunig, 2000).

Searls and Weinberger (2001) argue that traditional corporate messaging is less interesting and less relevant to online publics because traditional corporate marketers show little care for the people who make up the markets. Rather, corporate voices sound more like profit-driven machinery than real people engaged in two-way conversations. According to these critics, corporate Web sites that look like corporate brochures (i.e., brochureware) are a visible symptom of an ineffective markets-as-targets communication strategy.

These critics are not PR fatalists, however, and their sharp-witted commentary echoes early mainstream literature on online public relations. Witmer (2000), for example, applied a structuration perspective in recognizing the importance of netiquette as a means of building and maintaining social rules in computer-mediated communication. Holtz (1999) implored public relations people communicating online to be candid:

…real public relations is the process of engaging your constituent publics in a process of two-way symmetrical communication, designed to result in both sides reaching their objectives… But in cyberspace, the requirement to be candid and forthright is doubly important. (p. 230)

This advice is in line with classic texts on nonfiction writing such as William Zinsser's On Writing Well. Zinsser encourages business communicators to use plain talk.

It's what stockholders want from their corporation, what customers want from their bank, what the widow wants from the agency handling her social security. There is a deep yearning for human contact and a resentment of bombast. (2001, p. 173)

Conceptualizing Blogs and the Human Voice

For the purposes of translating these observations into more measurable empirical constructs, the following key characteristics will be used to help quantify the concepts used in this study.

Blogs

Based on Herring et al.'s (2004) study and on the Merriam-Webster definition, widely heralded by bloggers themselves, the conceptual definition of blogs will consist of five characteristics of Web-based postings: 1) frequent updating, 2) reverse chronological order, 3) inclusion of personal journal material, 4) ability of readers to add comments, 5) inclusion of hyperlinks. Perhaps more importantly, reader perceptions of whether posted material is blog content or not blog content must be considered as well.

Organizational Blogs

For the purpose of developing a program of research studying blogs as tools for building relationships between all types of organizations and their publics, and for the purpose of excluding patently hostile blogs from the current analysis, the following adaptation of Sifry's (2004) definition is offered:

Organizational blogs are Web logs that meet three criteria. They are 1) maintained by people who post in an official or semi-official capacity at an organization, 2) endorsed explicitly or implicitly by that organization, and 3) posted by a person perceived by publics to be clearly affiliated with the organization.

Conversational Human Voice

A careful reading of Searls and Weinberger's (2001) chapter, "Markets are Conversations," by researchers with backgrounds in public relations practice, theory, and research yielded several potential indicators of the "human voice" as it might be identified in computer-mediated communications. These included familiar public relations qualities such as being open to dialog, welcoming conversational communication, and providing prompt feedback. They also include characteristics of communication that otherwise might not be associated with traditional corporate communication (which is why Searls and Weinberger go out of their way to highlight them): communicating with a sense of humor, admitting mistakes, treating others as human, and providing links to competitors.

Relational Maintenance Strategies

Stafford and Canary's (1991) taxonomy of relational maintenance strategies provides a theoretical link between strategies found to be effective in interpersonal communications and the relational outcomes of interest in public relations. These relational strategies have been found to be related to perceptions of equitable relationships (Canary & Stafford, 1992) and have subsequently been used as an evaluative tool in a variety of interpersonal communication objectives such as liking, satisfaction, and reduction of apprehension in interethnic relationships (Dainton, Stafford, & Canary, 1994; Messman, Canary, & Hause, 2000; Toale & McCroskey, 2001). More recently, they also have been used to evaluate online interpersonal relationships (Wright, 2004).

Canary and Stafford's (1992) five maintenance strategy factors include the following: positivity (interacting with partners in a cheerful, uncritical manner); openness (directly discussing the nature of the relationship and disclosing one's desires for the relationship); assurances (communicating one's desire to continue with the relationship); social networks (relying on common affiliations and relations); and sharing tasks (performing one's responsibilities). These strategies may be applied to organizational relationships by shifting the focus of the communication strategies to a public, rather than interpersonal, audience (Hon & Grunig, 1999). For example, positivity/cooperation would involve efforts an organization employs to make a relationship more enjoyable for those involved; openness would include providing disclosure regarding the nature of the organization and information of value to audiences; assurances would include communication that emphasizes the value of audience members; social networks would involve an emphasis on common affiliations between organization and publics; and sharing tasks may include asking for public involvement when appropriate.

Evaluating Relational Strategies—Relationship Outcomes

Throughout the last several decades, public relations practitioners have endeavored to quantify their organizational value. A unique contribution of public relations is its ability to build and enhance relationships with key publics via relational strategies.

Organizational relationships, which may include professional, personal, or community relations (Bruning & Ledingham, 1999), can increase organizational effectiveness, reduce the cost of litigation, regulation, boycotts, etc., and may also contribute to an organization's financial well being through shareholder, consumer, and/or donor support. Research in public relations and marketing has recently turned its focus on quantifying this abstract notion of relationship management as a means of contributing to the success and well being of an organization, as well as organizational financial success.

In assessing the effectiveness of relational strategies, a number of fields have developed relationship measures to evaluate the strength and quality of relationships. Within the interpersonal communication discipline, for example, Burgoon and Hale (1987) developed a relational communication scale. The scale involved dependent variable measures such as immediacy/affection, similarity/depth, and receptivity/trust.

Public relations research, meanwhile, has shifted the relationship focus from more interpersonal relational attributes to communication linkages aimed at attaining mutual goals (Broom et al., 1997). While a number of public relations scholars have developed preliminary measures to evaluate the quality of an organization's relationship with publics, including outcomes relating to mutual satisfaction (Ferguson, 1984) and predictions of consumer's behavior in competitive markets (Bruning & Ledingham, 1999), Broom et al. (1997) suggested there was little agreement in defining relationships or how they should be measured.

In response, Hon and Grunig (1999) developed measures to evaluate the success of relationship-building efforts. Through an application of Hon and Grunig's (1999) PR Relationship Measurement Scale, an organization's longer-term relationships with key publics can be evaluated by focusing on the following four indicators of the quality of an organization-public relationship: 1) control mutuality, the degree to which parties agree on issues of power and influence; 2) trust, which includes dimensions of integrity, dependability, and competence; 3) satisfaction, the degree to which parties feel favorably toward each other because positive expectations are met; and 4) commitment, the degree to which parties believe that the relationship is worthwhile to continue.

In addition to these indicators regarding the quality of the organization-public relationship, the status of the relationship also can be evaluated via two relationship type indicators. In an exchange relationship, parties are willing to provide benefits because comparable benefits are expected in return. Conversely, in a communal relationship, both parties provide benefits without expecting anything in return. Rather, they provide benefits out of concern for the welfare of the other (Hon & Grunig, 1999, pp. 18-21). Although some researchers suggest that two-way measurement of both the organization's perceptions and its publics' perceptions is necessary to evaluate organization-public relationships (Broom et al., 1997), there is support for the notion that publics' perceptions of organizations are key indicators of the overall organization-public relationship (Kim, 2001). Kim's argument lends credence to the measures proposed by Hon and Grunig (1999) and the rationale for evaluating online relationship building according to public perceptions.

Research Questions and Hypotheses

The following research questions were posed to develop operational definitions for an index appropriate for evaluating relational maintenance strategies in an online context.

RQ1: What relational strategy measures are most appropriate for evaluating computer-mediated relationships and how can they be measured in the context of online public relations and blogs?

RQ2: What attributes are appropriate for measuring the human voice in computer-mediated public relations and blogs as a relational strategy?

Given the operational definitions obtained from RQ1 and RQ2, the following hypotheses were advanced:

H1: People who read organizational blogs will perceive an organization's relational maintenance strategies as demonstrating more of a conversational human voice than those who read only the organization's traditional Web content or those in a control group who read Web content unrelated to the organization.¹

H2-H5: People who read organizational blogs will perceive an organization's relational maintenance strategies (communicated relational commitment, task sharing, responsiveness/customer service, positivity/optimism) as greater than those who read only the organization's traditional Web content or those in a control group who read Web content unrelated to the organization.

Finally, the following question evaluates online organization-public relationships from a more general perspective:

RQ3: Which relational strategies correlate with which relational outcomes?

Method

Procedure

College students are undoubtedly a key public for software companies, and a Fortune 100 tech-industry company is likely relevant to current students and future communication industry professionals. The experimental manipulations were designed to take advantage of this mutual interest. A total of 42 undergraduate students enrolled in introductory journalism and mass communication courses agreed to participate in this study. Students were awarded class credit for their participation or were provided alternative assignments of comparable length for the same credit.

Students were randomly assigned to one of three experimental groups. The first group was asked to read a sample of content from Microsoft's corporate Web site. This sample was a case study published online by Microsoft that described the business benefits of a Microsoft-developed bank account self-management site. The printed case study was six pages long. The second group was asked to read a selection from Microsoft's corporate blog page, which included in reverse-chronological order the first paragraphs of a dozen "most recent" blogs posted by Microsoft employees at blogs.msdn.com. One expanded blog entry, which included a comments-form, elaborated thoughts by a Microsoft employee about an online scavenger hunt used by Microsoft recruiters to screen applicants. Feedback from blog readers was attached to the printout from the general blog listing. The total printout (blog summaries and expanded individual blog entry with comments) was six pages long.

To manipulate "voice," the experiment was designed to present conditions that differed primarily in content between the traditional corporate communication condition and the blog condition. Although some sacrifices in validity were made by printing the stimulus materials from actual Microsoft Web pages, this made it more likely that any differences found would be a result of conversational voice typical in blogs compared to the more declarative voice of traditional corporate communication (i.e., brochureware). Finally, as a control condition, a third group was asked to read a six-page printout from Web content posted on the DuPont Web site.

Students then were asked to complete a three-part questionnaire. The first section asked students about their past communications with Microsoft, the understandability of the material they read, and how relevant the material seemed (e.g., whether they thought they were members of the intended audience for the material). It also included manipulation check items about the perceived characteristics of the material they read. The next two sections included Likert-type questions written to measure perceptions of Microsoft's use of conversational human voice, relational strategies, and relational outcomes.

Index development

In order to assess the first two research questions—which were aimed at developing and assessing computer-mediated relational strategies—a composite list of relational strategies based on literature in interpersonal communication and public relations was compiled. These strategies were operationalized via modifications of Canary and Stafford's (1992) maintenance strategies and Hon and Grunig's (1999) suggestions for their application in an organization-public context. These included attributes of positivity, openness, assurances, social networking, and task sharing. Additional conversational attributes of relational strategies based on Searls and Weinberger's (2001) notion of markets as conversations were also compiled to operationalize the "human voice."

Given the revisions and additions to the maintenance strategy items and the exploratory nature of the "human voice" items, exploratory factor analyses (principal axis factoring with oblique rotation) were used to help assess the validity of the constructs. Internal reliability of the constructs was also assessed.

To examine the correlation of these strategies with relationship outcomes (RQ3), Hon and Grunig's (1999) PR Relationship Outcome Scale was used.

Results

Phase 1—Determining Operational Definitions for Independent Variables (RQ1 and RQ2)

Exploratory factor analyses and assessments of face validity of the constructs suggested five indices to measure relational strategies appropriate in an online context. The criteria used to determine the number of factors included the scree test and the interpretability of the factor solution. (See Appendix A for a list of resulting index items.) The strategies that were identified follow.

Communicated Relational Commitment

The interpersonal communication constructs of "openness" and "assurances" loaded onto one factor that accounted for 48.38% of the item variance. The index included six items and produced a standardized alpha reliability coefficient of .82. These items emphasize an expressed commitment to building and maintaining a relationship and underscore the nature and quality of the relationship.

Conversational Human Voice

Developed specifically for this analysis, these items were developed to emphasize human communication attributes capable of being conveyed in a computer-mediated context. These 11 items loaded onto a single factor that accounted for 45.70% of the item variance and produced a standardized alpha reliability coefficient of .87.

Task Sharing

Items originally designated under Canary and Stafford's (1992) general heading of "tasks" loaded onto two separate factors—one emphasizing organizational tasks and responsibilities, the other focusing on responsiveness and customer service. The three items emphasizing organizational tasks loaded onto a single factor accounting for 55.51% of the item variance and produced a standardized alpha reliability coefficient of .76.

Responsiveness/Customer Service

These three items, which appear to represent a separate factor in a computer-mediated context, indicate an organization's willingness to respond promptly to customer inquiries and complaints. They loaded onto one factor that accounted for 49.67% of the item variance and produced a standardized alpha reliability coefficient of .73.

Positivity/Optimism

Only two items from the initial composite list of interpersonal communication and traditional public relations relational strategies emphasizing positivity translated into a construct applicable in this computer-mediated context. These two items accounted for 79.66% of the item variance and produced a standardized alpha reliability coefficient of .89.

Phase 2—Hypothesis testing

Applying the operational definitions developed in the initial phase of the study, relational strategies were used to assess potential advantages of blogs over traditional Web pages as tools for building and maintaining relationships online.

H1a: People who read organizational blogs will perceive an organization's relational maintenance strategies as demonstrating more of a conversational human voice than those who read only the organization's traditional Web content.

H1b: People who read organizational blogs will perceive an organization's relational maintenance strategies as demonstrating more of a conversational human voice than those in a control group who read Web content unrelated to the organization.

H2a-5a: People who read organizational blogs will perceive an organization's relational maintenance strategies (communicated relational commitment, task sharing, responsiveness/customer service, positivity/optimism) as greater than those who read only the organization's traditional Web content.

H2b-5b: People who read organizational blogs will perceive an organization's relational maintenance strategies (communicated relational commitment, task sharing, responsiveness/customer service, positivity/optimism) as greater than those in a control group who read Web content unrelated to the organization.

Manipulation Check

Before testing hypotheses, a manipulation check was conducted to verify that participants identified the type of material to which they were exposed as expected (see Table 1). Participants exposed to Microsoft blogs, for example, perceived a significantly greater ability to add comments than those who viewed Microsoft's traditional Web content. Similarly, participants who were exposed to Microsoft blog material also reported significantly higher perceptions of personal journal material as content than those who viewed Microsoft's traditional Web site. Thus, participants were aware of significant differences between Microsoft's blog and Web content in two of the definitional aspects of blogs—personal journaling and the ability to add comments. Reverse chronological order and frequent updating were clearly apparent in the blog stimulus materials. The final definitional aspect of blogs, the inclusion of hyperlinks, was available in all three conditions. Consistent with expectations, those in blog conditions also identified what they read as indeed "including blog content" compared to those who read traditional Web material. This last item indicates that, in general, this group of undergraduates was capable of identifying blog content when it was presented in a printed format.

Hypotheses 1a & 1b—Conversational Human Voice

"Conversational human voice," which was operationalized as the average of 11 Likert-type items with responses ranging from 1 = strongly disagree to 7 = strongly agree, was significantly greater in the Microsoft blog condition (M = 5.16, SD = .86) than in the Microsoft traditional Web page condition (M = 3.83, SD = 1.11). This difference was significant at p ≤ .001. "Human voice" also was significantly greater in the Microsoft blog condition (M = 5.16, SD = .86) than in the control group condition (M = 3.98, SD = .75). This difference was significant at p ≤ .001. Hypotheses 1a and 1b were supported.

Hypotheses 2a & 2b—Communicated Relational Commitment

"Communicated relational commitment," which was operationalized as the average of six Likert-type items with responses ranging from 1 = strongly disagree to 7 = strongly agree, was not significantly different in the Microsoft blog condition (M = 4.44, SD = .97) than in the Microsoft traditional Web page condition (M = 5.12, SD = 1.20). Perceived "communicated relational commitment" also was not significantly different in Microsoft blog condition (M = 4.44, SD = .97) than in the control group who read unrelated Web content (M = 4.83, SD = .69). Hypotheses 2a and 2b were not supported.

Hypotheses 3a & 3b—Task Sharing

"Task sharing," which was operationalized as the average of three 7-point Likert-type items, was not significantly greater in the Microsoft blog condition (M = 4.21, SD = .67) than in the Microsoft traditional Web page condition (M = 4.95, SD = 1.05). Rather, the traditional Web page condition was perceived as demonstrating significantly greater task sharing strategies than the blog condition, with p ≤ .02. The control group condition (M = 4.67, SD = .55) also was perceived as demonstrating significantly greater task sharing strategies than the Microsoft blog condition (M = 4.21, SD = .67), with p ≤ .03. Hypotheses 3a and 3b were not supported.

Hypotheses 4a & 4b—Responsiveness/Customer Service

"Responsiveness," which was operationalized as the average of three 7-point Likert-type items, was not significantly greater in the Microsoft blog condition (M = 4.79, SD = 1.11) than in the Microsoft traditional Web page condition (M = 4.54, SD = 1.27). Perceived "responsiveness" also was not significantly greater in Microsoft blog condition (M = 4.79, SD = 1.11) than in the control group who read unrelated Web content (M = 4.47, SD = 1.25). Hypotheses 4a and 4b were not supported.

Hypotheses 5a & 5b—Positivity/Optimism

"Positivity/optimism," which was operationalized as the average of two 7-point Likert-type items, was not significantly different in the Microsoft blog condition (M = 5.25, SD = .96) than in the Microsoft traditional Web page condition (M = 5.75, SD = 1.33). Perceived "positivity/optimism" also was not significantly different in Microsoft blog condition (M = 5.25, SD = .96) than in the control group who read unrelated Web content (M = 5.63, SD = .79). Hypotheses 5a and 5b were not supported.

Phase 3—Further Analyses/General Perspectives

From a more general perspective regarding online organization-public relationships, the following question was asked:

RQ3: Which relational strategies correlate with which relationship outcomes?

Correlation coefficients were computed among the five relational strategies variables, including "conversational human voice," and the six relationship outcome measures developed by Hon and Grunig (1999), which include four relationship outcome measures and two relationship type indicators. The results of the correlation analyses presented in Table 2 show that of the five computer-mediated relational strategies, three strategies correlated significantly with each of the four relationship outcome measures. These include communicated relational commitment, human voice, and positivity/optimism. The relational strategy of task sharing, meanwhile, correlated significantly with all outcomes except control mutuality. Responsiveness/customer service correlated significantly only with the relationship outcome measures of trust and commitment.

Overall, it appears that the types of perceived relationships were not strongly correlated with online relational strategies. Communal relationships correlated significantly only with the relational strategy of communicated relational commitment, and exchange relationships correlated significantly only with the task sharing relational strategy.

Discussion

This study identified and operationalized several computer-mediated relational strategies that relate directly to relational outcomes. Perhaps most methodologically useful among these constructs are conversational human voice and communicated relational commitment. Both show promise in terms of face validity, internal reliability, internal validity, and external validity. The index for conversational human voice was built using a list of ingredients informed by an array of online, professional, and first-hand observations that were articulated particularly well by Searls and Weinberger (2001). The first attempt in analyzing the 11 items loaded onto a single factor. This added a dimension of internal validity to the face validity as judged by the researchers. The Cronbach's alpha test then added confidence in the internal reliability of the index items.

Of course, the external validity and utility of the human voice in CMC research depends greatly on how it relates to other constructs of interest. Conversational human voice was significantly greater in the blog condition than in non-blog conditions, showing that voice is an important quality of this channel of communication. Furthermore, as a relational maintenance strategy, voice correlated positively and significantly with all four of the key relational outcomes measured: trust, satisfaction, control mutuality, and commitment. In other words, conversational voice, which was perceived to be greater in the blog condition than in the non-blog condition, appears to be an appropriate relational maintenance strategy in online contexts. From an online public relations perspective, this finding yields some initial support for the notion that organizational blogs may have value in efforts aimed at building and maintaining relationships. In the context of this study of blogs and online public relations, the concepts of "openness" and "assurances" seem better treated as a single factor, which was labeled communicated relational commitment. The index of these six items was validated in the same manner as the human-voice construct above, but blogs did not carry any apparent advantage in relaying communicated relational commitment.

Given the potential for overlap between the human voice factor and the communicated relational commitment factor, a second factor analysis (principal axis) was computed with all 17 variables representing both constructs. The factor loadings illustrated the differences between these two indices (Appendix B), and hence give some evidence of discriminant validity.

While positivity/optimism shows promise when one considers the strong correlations between this quality and the four relational outcomes, its two-item index suggests that different dimensions of this construct are worth seeking, especially dimensions that may inform the practice of online public relations more than merely recommending that practitioners be cheerful and optimistic in their computer-mediated communications.

Finally, the three-item indices for task-sharing and responsiveness/customer service provide good places to continue exploring areas of overlap between online public relations and related management and marketing functions. The link between task sharing and exchange relationships, and the correlations between responsiveness/customer service and the relational outcomes of trust and commitment illustrate the importance of integrated approaches to understanding where public relations fits into an organization's online strategies.

Only conversational human voice was perceived higher in the blog condition than in the non-blog condition, which perhaps underscores the unique value of organizational blogs to online public relations. Although it appears that both organizational blogs and Web sites have the ability to communicate other relational maintenance strategies, particularly relational commitment and positivity/openness, blogs may be particularly well suited for conveying a "human voice" in a computer-mediated context. Blogs may therefore serve as valuable tools in an organization's array of online communication tactics by conveying some sense of human attributes existing behind an organizational façade.

Limitations and Directions for Future Research

The results of this study should be considered with three categories of issues in mind: experimental control issues, sampling issues, and statistical issues.

Experimental Control Issues

Manipulating "blogliness" while maintaining some degree of external validity meant accepting a trade-off between experimental control and authenticity of stimulus materials. Providing printouts of the Web-based material, for example, meant that participants in this experiment did not see the information as they would online, with the full effect of graphics, links, and interactive features. But using printouts ensured that the researchers knew exactly what the students were exposed to.

Likewise, choosing comparable samples of online material to represent the universes of blog and non-blog material required a challenging judgment call on the part of the researchers. Additional analyses showed that participants who read the blog material perceived it to be significantly less understandable and marginally less relevant to them than those who read non-blog material. These analyses will be helpful in choosing materials for future research.

The likely effect of these issues, however, would work against any apparent advantage for blogs. That is, blog materials were more constrained by any perceived differences in intended audience (much of the blogs.msdn.com content is written for current and aspiring software developers) and by being presented in print. This might be interpreted as adding weight to the advantages that were found for blogs in terms of human voice. Nonetheless, these extra analyses suggest that future experimental research will benefit from the use of better-developed stimulus material. In order to assess perceptions of the conversational human voice, computer-mediated communication, and online relational strategies, a more realistic approach might involve allowing participants to navigate organizational Web sites and blogs in a computer lab. Given the frequency of organizational Web site updates and blog entries, the actual experiment likely would need to be undertaken in one sitting, however.

Sampling Issues

The selective sample of participants in this study implies the usual caveats of experimental design vis-à-vis broader-scale methods with more representative samples. The issue of how well these findings generalize to larger populations must be considered. As a study in online public relations, this experiment was designed to take advantage of the likelihood that college juniors and seniors would meet two important criteria: 1) They see the operations of a company like Microsoft as relevant, and 2) they are considered a key public for such a company. In any case, future research should examine other publics beyond college students.

Statistical Issues

Although sampling size is less of an issue in questioning the support found for the human-voice hypothesis, it does make it more difficult to "reject the null" for hypotheses that weren't supported. That is, the smaller sample size means greater likelihood of a Type II error. In many cases, though, the compared means for indices used in hypothesis testing were actually the reverse of what was predicted, indicating fairly strong evidence that these hypotheses will be hard to support in any context.

The small sample size also limits the array of statistical tests available. The factor analyses, reliability tests, and table of correlations help support claims about the internal validity of the factors and highlight interesting relationships. But exploring the predictive value of these constructs and directional relationships with structural equation modeling/path analysis will require a larger sample size.

Conclusion

The observed advantage of blogs in conveying a conversational human voice echoes what popular and professional literature on the nature of blogs already tells us. Blogs are a good place to speak candidly with a conversational style (e.g., "invite people to a conversation"), and this conversational style may be an important part of the process of building and maintaining computer-mediated relationships. Among the most important findings of this study are that 1) blogs were perceived as more conversational than organizational Web sites, and 2) this conversational human voice correlated positively with other previously-identified relationship outcomes. The perceived personal nature of organizational blogs, in this case, is related to relationship indicators. But when public relations people want to discuss their commitment to the relationship between their organizations and publics, or to communicate the organization's desire to build relationships (identified as communicated relational commitment), this is likely done just as well where people would expect to find such information, in the more traditional venues of case studies and corporate Web pages. Although blogs allow people representing organizations to speak candidly, blogs may not be the best venue for "PR" messages intended to talk up an organization's commitment to its public relationships.

As Searls and Weinberger expounded "… the best of the people in PR are not PR Types at all… they're the company's best conversationalists" (2001, p. 90). This is quite a conundrum for the practice and theory of public relations. Answering the riddle requires an understanding of which relational strategies should be (and should not be) used in which forums. At the heart of the matter are the questions of who public relations people are and how they communicate to build and maintain solid relationships, online and elsewhere. This paper offers valuable theoretical tools for pursuing this line of inquiry. To enter the market conversation, Web sites need to have a voice, express a point of view, ignite a dialogue, and give access to helpful people (Searls & Weinberger, 2001). It all starts with having a voice. Expressing a point of view in a personal tone in a blog is likely a good way to get a conversation started. The actual interaction between organizations and their publics that follows is a promising focus of future research.

Notes

1. Operational definitions of effective perceived relational strategies were developed based on results from RQ1 and RQ2.

References

Broom, G. M., Casey, S., & Ritchey, J. (1997). Toward a concept and theory of organization-public relationships. Journal of Public Relations Research, 9 (2), 83-98.

Bruning, S. D., & Ledingham, J. A. (1999). Relationships between organizations and publics: Development of a multi-dimension organization-public relationship scale. Public Relations Review, 25 (2), 157-58.

Burgoon, J. K., & Hale, J. L. (1987). Validation and measurement of the fundamental themes of relational communication. Communication Monographs, 54 (1), 19-41.

Canary, D. J., & Stafford, L. (1992). Relational maintenance strategies and equity in marriage. Communication Monographs, 59 (3), 243-67.

Dainton, M., Stafford, L., & Canary, D. J. (1994). Maintenance strategies and physical affection as predictors of love, liking, and satisfaction in marriage. Communication Reports, 7 (2), 88-98.

Ferguson, M. A. (1984, August). Building Theory in Public Relations: Interorganizational Relationships. Paper presented at the annual meeting of AEJMC, Gainesville, FL.

Grunig, J. E. (2000). Collectivism, collaboration, and societal corporatism as core professional values in public relations. Journal of Public Relations Research, 12 (1), 23-48.

Grunig, J. E., & Huang, Y. (2000). From organization effectiveness to relationship indicators: Antecedents of relationships, public relations strategies, and relationship outcomes. In J. A. Ledingham & S. D. Bruning (Eds.), Public Relations as Relationship Management: A Relational Approach to the Study and Practice of Public Relations (pp. 23-53). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc.

Hallahan, K. (2004). Online public relations. In H. Bidgoli (Ed.), The Internet Encyclopedia, Vol. 2 (pp. 769-783). Hoboken, NJ: John Wiley.

Herring, S. C., Scheidt, L. A., Bonus, S., & Wright, E. (2004). Bridging the gap: A genre analysis of weblogs. Proceedings of the 37th Hawai'i International Conference on System Sciences (HICSS-37). Los Alamitos: IEEE Computer Society Press.

Holtz, S. (1999). Public Relations on the Net: Winning Strategies to Inform and Influence the Media, the Investment Community, the Public, and More! New York: AMACOM.

Hon, L. C., & Grunig, J. (1999). Guidelines for Measuring Relationships in Public Relations. Gainesville, FL: The Institute for Public Relations. Retrieved March 30, 2005 from http://www.instituteforpr.com/pdf/1999_guide_measure_relationships.pdf

Kim, Y. (2001). Searching for the organization-public relationship: A valid and reliable instrument. Journalism and Mass Communication Quarterly, 78 (4), 799-816.

Ledingham, J. A. (2003). Explicating relationship management as a general theory of public relations. Journal of Public Relations Research, 15 (2), 181-198.

Ledingham, J. A., & Bruning, S. D. (1998). Relationship management in public relations: Dimensions of an organization-public relationship. Public Relations Review, 24 (1), 55-65.

Merriam-Webster Collegiate Dictionary. (2003). (11th ed.) Springfield, MA: Merriam-Webster, Inc. Retrieved December 23, 2005 from http://www.m-w.com

Messman, S. J., Canary, D. J., & Hause, K. S. (2000). Motives to remain platonic, equity, and the use of maintenance strategies in opposite-sex friendships. Journal of Social and Personal Relationships, 17 (1), 67-94.

Sifry, D. (2004). Sifry's Alerts: October 2004 State of the Blogosphere. Retrieved March 30, 2005 from http://www.sifry.com/alerts/archives/000390.html

Schwartz, J. (2005). A Simple Request to Sun's Employees. Retrieved October 31, 2005 from http://blogs.sun.com/roller/page/jonathan?entry=a_simple_request_to_sun

Searls, D., & Weinberger, D. (2001). Markets are conversations. In R. Levine, C. Locke, D. Searls, & D. Weinberger (Eds.), The Cluetrain Manifesto: The End of Business as Usual (pp. 75-114). New York: Perseus Publishing.

Stafford, D. J., & Canary, L. (1991). Maintenance strategies and romantic relationship type, gender and relational characteristics. Journal of Social and Personal Relationships, 8 (2), 217-42.

Sun Microsystems. (2005). blogs.sun.com. Retrieved October 14, 2005 from http://blogs.sun.com

Toale, M. C., & McCroskey, J. C. (2001). Ethnocentrism and trait communication apprehension as predictors of interethnic communication apprehension and use of relational maintenance strategies in interethnic communication. Communication Quarterly, 49 (1), 70-83.

Witmer, D. F. (2000). Spinning the Web: A Handbook for Public Relations on the Internet. New York: Longman.

Wright, D. K. (2001). The Magic Communication Machine: Examining the Internet's Impact on Public Relations, Journalism, and the Public. Gainesville, FL: The Institute for Pubic Relations. Retrieved March 30, 2005 from http://www.instituteforpr.com/pdf/Worldcom%20Report.pdf

Wright, K. B. (2004). On-line relational maintenance strategies and perceptions of partners within exclusively Internet-based and primarily Internet-based relationships. Communication Studies, 55 (2), 239-253.

Zinsser, W. K. (2001). On Writing Well: The Classic Guide to Writing Nonfiction (25th Anniversary ed., 6th ed.). New York: HarperCollins.

Appendix A—Relational Maintenance Strategy Index Items

Introductory Statement

The following items concern things that an organization might do to maintain relationships with its publics. For each of the following items, please indicate the extent to which you perceive that Microsoft's online communication does each of these in order to maintain a relationship with you and others. Please indicate your level of agreement on a scale of 1 to 7, where 7 means that you strongly agree and 1 means that you strongly disagree that Microsoft employs these strategies. ²

Conversational Human Voice (11 items)

• Invites people to conversation.
• Is open to dialogue.
• Uses conversation-style communication.
• Tries to communicate in a human voice.
• Tries to be interesting in communication.
• Uses a sense of humor in communication.
• Provides links to competitors.
• Attempts to make communication enjoyable.
• Would admit a mistake.
• Provides prompt feedback addressing criticism with a direct, but uncritical manner.
• Treats me and others as human.

Communicated Relational Commitment (6 items)

• Attempts to demonstrate they are committed to maintaining the relationship.
• Communicates the organization's desire to continue to maintain and/or build a relationship with me and others.
• Stresses commitment to me and others.
• Implies that our relationship has a future/is a long-term commitment.
• Directly discusses the nature of the organization.
• Emphasizes the quality of our relationship.

Task Sharing (3 items)

• Performs their organizational responsibilities.
• Directly addresses organizational responsibilities.
• Does not shirk duties.

Responsiveness/Customer Service (3 items)

• Answers telephone calls from me and others.
• Answers emails from me and others.
• Is willing to address complaints or queries, rather than taking negative reactions.

Positivity/Optimism (2 items)

• Present themselves as cheerful and optimistic regarding the future.
• Uses a positive, optimistic tone.

2. Relational-maintenance-strategy items were presented in mixed order to prevent any systematic response error.

Appendix B—Factor Analysis for Conversational Human Voice and Communicated Relational Commitment Items

About the Authors

Tom Kelleher is an assistant professor in the School of Journalism and Mass Communication at the University of North Carolina at Chapel Hill. His research interests are online public relations, public relations theory, campaigns, ethics, and teaching/learning with online media.
Address: Carroll Hall, Campus Box 3365, UNC-Chapel Hill, Chapel Hill, NC 27599-3365 USA

Barbara M. Miller is a Ph.D. student and Park Fellow in the School of Journalism and Mass Communication at the University of North Carolina at Chapel Hill. Her research interests are online public relations, relationship building, and corporate social responsibility.
Address: Carroll Hall, Campus Box 3365, UNC-Chapel Hill, Chapel Hill, NC 27599-3365 USA