«13TH INTERNATIONAL PUBLIC RELATIONS RESEARCH CONFERENCE “Ethical Issues for Public Relations Practice in a Multicultural World” Holiday Inn ...»
Five items were used to measure ethical responsibility: “[Company name] provides full product information to customers” and “[Company name] acts responsibly toward the environment” (α =.841). Discretionary responsibilities were measured by four items, such as “[Company name] contributes resources to not-for-profit organizations” and “[Company name] contributes resources to the local economy (businesses and schools)” (α =.799). All of these items were measured on a 7-point scale, ranging from 1 = “strongly disagree” to 7 = “strongly agree.” Organization-public relationships. Relationships were measured in terms of trust, control mutuality, commitment, and satisfaction. To measure these relationship constructs, Hon and Grunig’s (1999) relationship items were adopted because this scale has been widely used to assess OPRs (Ki & Hon, 2007). Trust was measured using six items, such as “[Company name] treats people like me fairly and justly” and “[Company name] can be relied on to keep its promises” (α = 896). Control mutuality was assessed with four items, such as “[Company name] believes the opinions of people like me are legitimate” and “[Company name] really listens to what people like me have to say” (α =.785).
Behavioral intentions. Behavioral intentions were assessed in terms of purchase, employment, and investment intentions (Sen et al., 2006). Purchase intention was measured by asking participants to indicate their agreement with the following statement: “I would buy products from [company name].” Employment intention was assessed using four items, such as “I would like to seek information about jobs at [company name] in the future” and “I would very much like to work for [company name]” (α =.944). Regarding investment intention, participants were asked to rate their agreement with the following statement: “If I had money to invest, I would invest in [company name].” Responses were made on a 7-point scale ranging from 1 = “strongly disagree” to 7 = “strongly agree.” Composites for CSR and OPR Scales Since both CSR and OPR consisted of four indicators, each of which was measured by multiple items, principle component analysis was conducted to obtain composites for those indicators.
Principle component analysis is commonly used to extract a small number of components that account for maximum variance in a set of observed variables, whereas factor analysis is used to identify dimensional structures in the observed variables and relationships among those underlying dimensions (Meyers, Gamst, & Guarino, 2006; Tabachnick & Fidell, 2007). Principle component analysis creates a composite score for each component by summing observed variables with weights (Tabachnick & Fidell, 2007). Using this method, composites for CSR and OPR indicators were calculated as weighted sums of the observed items rather than using arithmetic means. Each indicator appears to have a unidimensional structure meeting the criterion of eigenvalues exceeding 1.0 (see Tables 1 and 2). The total variance explained by the observed items in each indicator was at least 61%, higher than the rule of thumb percentage (at least 50% or more) for determining the number of factors retained (Harlow, 2005). Within each indicator, all factor loadings of observed items were greater than a loading criterion of.40 (Meyers et al., 2006).
----------------------------------------See Tables 1 and 2
----------------------------------------Results A two-step approach to structural equation modeling was employed to test the proposed model.
In the two-step modeling approach, a confirmatory factor analysis measurement model is first estimated, and then a structural model is tested, whereas measurement and structural models are tested simultaneously in the one-step modeling approach (Anderson & Gerbing, 1998; Kline, 2005). Thus, compared to the one-step approach, the two-step process makes it easier to identify misspecification of a measurement model, which causes poor model fit, and to build a welldefined measurement model prior to testing the structural model (Kline, 2005). In this study, AMOS 18.0 was used to test the measurement model and to estimate the hypothesized structural relationships. Parameters in both models were obtained by maximum likelihood estimation.
582 Measurement Model First, a confirmatory factor analysis was conducted with an initial measurement model where all latent variables are allowed to covary. According to Hu and Bentler (1999), structural equation models can be considered acceptable when the value of the Comparative Fit Index (CFI) equals or exceeds.95, the value of the Standardized Root Mean Square Residual (SRMR) is less than or equal to.08, and the value of the Root Mean Square Error of Approximation (RMSEA) is less than or equal to.06. These goodness-of-fit indices for the measurement model indicate that it fit the data well (χ2 = 393.839, df = 146, p.05; CFI =.966, SRMR =.0417, RMSEA =.059).
Since the model fit of the initial measurement model appeared to be satisfactory, there was no need to respecify the model; thus, the hypothesized structural relations were imposed on the latent variables (i.e., CA, CSR, OPR, company evaluation, and behavioral intentions) in this measurement model to test the proposed structural model.
Structural Model The proposed structural model appeared to fit the data satisfactorily (χ2 = 409.051, df = 147, p.05; CFI =.964, SRMR =.0399, RMSEA =.061) based on the following model fit criteria: CFI ≥.95, SRMR ≤.08, and RMSEA ≤.06 (Hu & Bentler, 1999). However, an examination of the paths in the proposed structural model revealed that a path from CSR to company evaluation was not statistically significant (B =.050, SE =.056, β =.049, p =.380). Thus, this structural model (structural model I) was revised by eliminating the path between CSR and company evaluation.
The revised structural model (structural model II) was tested, and as a result, it also appeared to be an acceptable model (χ2 = 409.785, df = 148, p.05; CFI =.964, SRMR =.0400, RMSEA =.060) (see Table 3). Since the revised model was more parsimonious than the initial structural model and believed to be good enough to analyze the parameter estimates, hypothesis testing was conducted within the context of the revised structural model.
-------------------------------See Table 3
-------------------------------Testing of Hypotheses and Research Question The hypotheses and research question were examined by confirming the presence of statistically significant associations in the predicted directions based on standardized parameter estimates. H1 posited that the perceptions of CSR initiatives are positively associated with favorable relationships between an organization and its publics. The results showed that the participants’ perceptions of CSR had a positive influence on their relationships with the company (B =.498, SE =.043, β =.520, p.05). Thus, H1 was supported.
H2 predicted that CSR perceptions have a more positive influence on OPRs than do CA perceptions. Like the perceptions of CSR initiatives, CA perceptions were identified to have a significant positive relationship with OPRs (B =.357, SE =.042, β =.372, p.05). Consistent with H2, CSR perceptions (β =.520) appear to have a more positive effect on OPRs than do CA perceptions (β =.372). Based on these results, H2 was supported.
H3 posited that favorable OPRs are positively associated with positive evaluations of an organization. To test this hypothesis, a standardized path from OPR to company evaluation was examined. As a result, favorable OPRs turned out to have a positive influence on the evaluations of a company (B =.332, SE =.052, β =.313, p.05). Thus, H3 was supported.
583 H4 examined whether favorable OPRs are positively associated with behavioral intentions toward an organization. The results revealed that favorable OPRs had a positive effect on behavior intentions (B =.686, SE =.091, β =.512, p.05). In other words, as a company maintains more favorable relationships with its publics, the publics are more likely to purchase its products, seek job opportunities there, and invest their resources in the company. Therefore, H4 was supported.
RQ1 asked whether CSR has a direct effect on the publics’ evaluations of an organization and their behavior intentions or if it affects these outcomes through OPRs. Regarding the evaluations of the company, CSR appears to have only an indirect effect on company evaluation through relationships between the company and its publics (see Table 4). This standardized indirect effect was estimated to be.163. As noted before, the direct path from CSR to company evaluation was deleted in the final structural model (structural model II) due to its statistical insignificance (B =.050, SE =.056, β =.049, p =.380). Regarding behavioral intentions, however, the results indicate that CSR had both direct and indirect effects on behavioral intentions. A path between CSR and behavioral intentions was found to be statistically significant, but not in the predicted direction (B = -.214, SE =.078, β = -.167, p.05). Despite this negative direct effect (β = -.167), the standardized indirect effect of CSR on behavioral intentions through OPRs appears to be positive (.293), resulting in the positive total effect of.126. Figure 2 describes graphically the relationships among the latent variables and observed indicators, presenting the standardized estimates of paths in the tested model.
-----------------------------------------------Insert Table 4 and Figure 2 about here
-----------------------------------------------Discussion Since CSR is now considered to be a viable corporate strategy, this study sought to shed light on the role of CSR in relationship management in public relations, using a conceptual framework linking CSR initiatives to well-developed relationships and positive responses to an organization.
In this link through which CSR influences evaluative and behavioral outcomes, OPR may serve as an important mediator. Thus, another purpose of this study was to examine whether CSR performances affect evaluative and behavioral outcomes directly or indirectly through OPRs.
Structural equation modeling was used to test linkages proposed in the conceptual framework.
This study supports the notion that CSR initiatives play an important role in developing positive relationships between an organization and its publics (Marin & Ruiz, 2007). The overall results suggest that the perceptions of a company’s active engagement in CSR are positively related to favorable OPRs, positive evaluations of a company, and positive behavioral intentions, such as purchasing, employment, and investment. These results point to the value of CSR activities as a vital factor in building intangible corporate assets and boosting economic benefits. Through the implementation of distinctive CSR programs, a company can be perceived as a good corporate citizen that is concerned about society’s and its publics’ well-being. Such perceptions may motivate its publics to engage in a relationship with the company, consume its products, seek employment there, and invest in the company. These findings are consistent with those of Sen et al. (2006) who found that a company’s socially responsible behaviors can enhance not only attitudes and identification but also intentions to purchase its products, commit personal efforts (labor) to, and share resources (money) with the company.
584 Since the main function of business is to sustain and increase economic gains, it may not be prudent for corporations to devote all of their discretionary resources to serving societal needs and expectations. However, given the potential of CSR to generate intangible benefits, companies may need to implement CSR activities to promote their good citizenship without hurting their bottom line. Knowing that the primary goal of business is the pursuit of financial success, people may also accept and incorporate both firm-serving and public-serving motives underlying CSR activities. Thus, regardless of whether a company’s commitment to CSR is viewed as stemming from sincere social concerns or profit-generating motives, good CSR performances can be effective in maintaining positive OPRs.
Another noteworthy finding is that CSR’s influence is more positive than CA in developing positive OPRs. In other words, perceptions of a company’s CSR initiatives may be a more influential factor to stimulate the publics’ interest in the company and their desire to develop a long-term relationship with the company. Even though CSR does not compensate for shortcomings in business practices, a company’s devotion to CSR activities may help facilitate OPRs even when its business is suffering. It is notable, that while CSR has a stronger influence on OPR than does CA, CA does have a direct and positive effect on OPR, company evaluation, and behavioral intent.
Perhaps even more importantly, this study found that CSR had only an indirect effect on company evaluation through OPRs, while both directly and indirectly affecting behavioral intentions (i.e., purchase, employment, and investment). This finding highlights an important mediating role of well-managed relationships in the link between CSR initiatives and a company’s evaluation. In other words, perceptions of CSR may lead to increasing positive evaluations of a company only by developing favorable OPRs. Even with regard to the direct and indirect effects of CSR on behavioral outcomes, OPRs appear to be a critical factor that determines the degree to which CSR performances are beneficial to a company. Thus, relationship building is an essential step in maximizing the positive impact of CSR and producing its intended outcomes.