Understanding Public Engagement: Conceptualizing and Measuring its Influence on Supportive Behavioral Intentions

Abstract

The concept of public engagement has suffered from a lack of a clear theoretical definition and sound operationalization. This study conceptualized and operationalized public engagement at the individual level and tested the micro-model of public engagement in relation to key public relations concepts and supportive behavioral intentions. Further, the study tried to bring the affective component into public relations scholarship by introducing engagement as a behavioral motivator that elicits individual publics' supportive behaviors toward an organization. The results showed the proposed 13-item scale of public engagement was sound in reliability and validity and found support for the proposed model as well as the mediation of public engagement.

Many corporate communication campaigns focus on achieving desirable effects on publics such as awareness, knowledge, positive opinions, attitudes, and behaviors (Dozier & Ehling, 1992). In the public relations domain, motivation plays a critical role in connecting public awareness/knowledge to supportive behaviors. Questions regarding motivators of public actions in the form of actual support or opposition are important but have been unexplored. The concept of public engagement has emerged in the contemporary practices of public relations as a means to fill this gap between conceptual relationship assessments and tangible behavioral indicators of public support (e.g., Breakenridge, 2008; Solis, 2010).

The concept of engagement is nothing new. In business and organizational contexts, engagement has become a popular concept that brings positive organizational or marketing outcomes such as increased employee voluntary behaviors, consumer advocacy, financial support, and loyalty. Particularly, with the emergence of social media, publics have increasingly utilized this new form of communication as a critical tool of public engagement with organizations (e.g., Breakenridge, 2008). Hence, the success of organizations in this new public and media environment is largely dependent on organizations' abilities to find ways to effectively and positively engage their stakeholders for meaningful partnerships. This study provides a sound conceptualization of public engagement, grounded in existing discourse on engagement as well as empirically tests measurement of the concept, using a 13-item scale. Further, this study tests engagement as an affective motivational mediator that leads individuals' trust and satisfaction (key antecedents) to be displayed in supportive behavioral intentions for an organization (loyalty and positive word-of-mouth; WOM).

ENGAGEMENT: DEFINITIONAL AND OPERATIONAL AMBIGUITIES

Public engagement has emerged as an important concept in contemporary corporate and strategic communication contexts. Consequently, scholars and professionals in public relations began to view engagement as the ultimate marker or maker of a good organization–public relationship (OPR). Scholars have tried to connect engagement (as online interactivity) to positive outcomes such as increased donations or relationship improvement (e.g., Saffer, Sommerfeldt, & Taylor, 2013; Waters, Burnett, Lamm, & Lucas, 2009). As pioneer scholars in this regard, Kent and Taylor (2002) viewed engagement, the third feature of propinquity—a dimension of dialogue, as a mutually orientated state of mind that is “willing to give their whole selves” (p. 26) to communication encounters among dialogic partners.

Such psychological aspect of engagement, however, has received little attention for theoretical conceptualization in public relations until recently. Uses of the term without a proper definition have caused confusion in the understanding of what engagement means in public relations theories and practices. The conceptual confusion caused by marketers' use of the term engagement to describe any interaction they were having with their customers has limited the utility of the term as many marketers were eager to “spin the program with a variety of metrics like clicks, visits, time spent, downloads, etc.—yet were [past tense added] never able to tie them back to real business results” (Corcoran, 2011, para. 1). Similarly confused, some scholars viewed engagement as the manifest outcomes (e.g., the number of posts on the organization's website) of organizational online dialogic strategies such as features like the discussion board (e.g., Bortree & Seltzer, 2009; Taylor, Kent, & White, 2001). More recently, engagement in the public relations context has been viewed as: the same as relationship such as leaving comments or tweeting about a company (Lovejoy, Waters, & Saxton, 2012), or a process for organizations to involve publics in various aspects of organizational activities through social media tools such as Twitter (Lovejoy et al., 2012), Facebook or interactive Web page features (Sommerfeldt & Taylor, 2011) that mimic dialogic communication by adopting a conversational human voice (Kelleher, 2009; Saffer et al., 2013; Sweetser & Metzgar, 2007).

Closely connected to its definitional ambiguity, engagement measurement has been limited in its operationalization and measurement. Common approaches by social media consultants or marketers equate engagement with the physical manifestations of an engaged state, such as the number of clicks, bookmarking, blogging, friend requests, tweets, or subscribing without really tapping into psychological aspect of being engaged. In organizational literature in particular, contemporary measures of engagement have often been similar to already existing measures of concepts such as commitment and satisfaction (Macey & Schneider, 2008). Such measures are problematic in that factors such as satisfaction, work involvement, or commitment tap only some aspects of engagement, not the whole (Macey & Schneider, 2008).

The lack of a clear theoretical definition and a careful operationalization to measure the concept in a valid and reliable manner has limited the utility of the concept in practice and hindered development of theories and practice on engagement. Most important, an individual public as the agents of experiencing engagement has been completely absent in the discussion of engagement in public relations. Most often, the public as the other dialogic and relational partner is not part of the engagement equation. Further, understanding engagement from the public's perspective enables scholars to discuss engagement in the ongoing OPR scholarship. If the traditional relational perspective has placed individual publics as the center of OPR scholarship, conceptualization and measurement of engagement from the individual level would make it possible and certainly easier to assess the role of engagement in OPR and subsequent effects on positive public behaviors such as donation, loyalty, or positive WOM behavior. Further, this study aims to bring the affective component into public relations scholarship by conceptualizing engagement as an affective motivational mediator that leads individuals' trust and satisfaction (key antecedents) to be displayed in supportive behavioral intentions for an organization (loyalty and positive WOM). The study provides a conceptualization of engagement with three dimensions: affective commitment, positive affectivity, and empowerment, and further suggests and tests a sound measurement of engagement on a 13-item scale.

LITERATURE REVIEW

Part 1: Conceptualization of Engagement as an Affective Motivator

For many decades, affect or emotion had been ignored as an unscientific concept in social and behavioral sciences and “sometimes treated as a unidimensional intervening variable such as drive or activation” (Lazarus, 1991, p. 819). According to Lazarus, the theory of emotion is relational, motivational, and cognitive and “human emotions are often predicated on complex social structures and meanings that define what is harmful and beneficial and, therefore, require judgment, the ability to learn from experience … and to distinguish” (p. 821). As such, the emotive/affective components of the encounters publics experience with an organization are equally crucial for determining that publics' behaviors toward that particular organization. 17th century philosopher Spinoza noted the importance of affects as compelling agents for human actions in that “emotions as affections of the soul of a person either increase the soul's power to act or diminish that power” (De Sousa, 2010, para 5).

Many OPR studies have proposed the direct link between OPR quality and various outcomes such as supportive behaviors (e.g., L. Grunig, J. Grunig, & Dozier, 2002) and positive attitudes and positive reputations (e.g., Bruning, 2000; Hong & Yang, 2009; Ki & Hon, 2007; Yang, 2007). Since the seminal work by Hon and J. Grunig (1999), the view of understanding OPR as perceptions or cognitions has been the dominating paradigm. Bruning's (2002) view of relationship attitudes as “affective orientation (Ki & Hon, 2007, p. 3)” and an important predictor of evaluation and behaviors has not been much explored or expanded. However, various academic fields such as marketing and interpersonal research have been recognizing the importance of affective components as a critical link between relationship quality and behavioral support. In the marketing domain, scholars (e.g., Agustin & Singh, 2005; Fournier, Dobscha, & Mick, 1998) questioned the effectiveness of customer relationship programs by arguing that loyalty from consumers is hard to achieve and predict and have suggested the importance of building strong affective connections or bonds with the customers in order for companies to build strong customer loyalty (e.g., McEwen, 2005; Yim, Tse, & Chan, 2008). Interpersonal literature also acknowledges affective components such as liking as a fundamental characteristic of relationships (Burgoon & Hale, 1984; Stafford & Canary, 1991).

Engagement as a psychological state should embrace related concepts in the form of absorption (e.g., paying high attention to the organization's activities and businesses), attachment (e.g., feeling of belonging to the organization), and/or enthusiasm both in its conceptual and operational level. This contention is echoed in Erickson's (2005) view of engagement: Engagement is “about passion and commitment-the willingness to invest oneself and expend one's discretionary effort” (p. 14) beyond what is required.

Review of existing literature on engagement from organizational management and organizational psychology literature suggests three prominent characteristics of the psychological state of being engaged with an organization: (a) positive affectivity, (b) affective commitment and (c) empowerment. Public engagement in this study is defined as a psychologically motivated affective state that brings voluntary extra-role behaviors, and is characterized by affective commitment, positive affectivity and empowerment that an individual public experiences in interactions with an organization over time.

Engagement as Affective Commitment

Affective commitment “reflects an emotional attachment to, identification with and involvement in an organization” (Meyer & Smith, 2000, p. 320) and has been viewed as one of the important determinants of employee dedication and loyalty (Rhoades, Eisenberger, & Armeli, 2001). Affective commitment implies voluntary desires to commit (Bansal, Irving, & Taylor, 2004) and has been closely linked to the concept of engagement (e.g., Allen & Meyer, 1996; Meyer & Allen, 1991). Some scholars and practitioners have defined engagement in terms of organizational commitment (e.g., Wellins & Conselman, 2005). This emotional and affective connection may result from “a psychological link” (Allen & Meyer, 1996, p. 252) or “a psychological bond” (Gruen, Summers, & Acito, 2000, p. 320) between relationship partners, or willingness “give their whole selves” (Kent & Taylor, p. 26), which makes voluntary defection or disconnection less likely (Gilliland & Bello, 2002; Meyer & Allen, 1997). This study postulates affective commitment is a facet of engagement characterized by emotional bonding and pride that brings additional efforts to sustain that relationship.

Engagement as Positive Affectivity

Characterized by “activated pleasant affect” (Larsen & Diener, 1992, p. 31), positive affectivity is consistent with how Schaufeli, Bakker, Hoogduin, Schaap, and Kladler (2001) defined engagement in terms of a positive affective-motivational state of fulfillment that is typically noted by vigor, dedication, and absorption (i.e., a state of one's attention being engrossed). Some scholars refer to positive affectivity as passion, excitement, (Wellins & Concelman, 2005) or emotional engagement (Fleming, Coffman, & Harter, 2005). Similarly, Macey and Schneider (2008) suggested that positive affectivity connoting “feelings of persistence, vigor, energy, dedication, absorption, enthusiasm, alertness, and pride” (p. 12) is a central facet in the conceptualization and operationalization of engagement. This study postulates positive affectivity is a facet of engagement that is characterized by six positive markers (positive and negative affect schedule [PANAS]) Watson, Clark, & Tellegen, 1988).

Engagement as Empowerment

Closely linked with the concept of power, empowerment has been largely considered equal to delegation or decentralization of decision-making power (Burke, 1986; Kanter, 1983) or “power with relations” (Berger, 2005, p. 16; Lane, 2007). Consequently, empowerment has been used interchangeably with employee or public participation. Based on the theory of intrinsic task motivation, Thomas and Velthouse (1990) argued that to empower is equal to to give power and suggested that empowerment entails a sense of authorization, capacity, and energy, concluding that empowering as to energize best captures the term as defined as a motivational state. Viewing empowerment as a motivational concept, Conger and Kanungo (1988) proposed enabling (empowering) individuals through enhancing self-efficacy. Bandura (1982) argued that the self-perceptions of efficacy are central to human agency, impacting actions, cognitive patterns, and emotional arousal. In other words, an individual's perception of self-efficacy or efficacy judgment will affect how much effort and persistence is necessary for one to exert in a given task and environment.

Empowerment is associated with many positive outcomes. For example, empowered individuals tend to display high efficiency in their achievements, high levels of energy, and effort in learning (Bandura, 1977); initiative and innovative behaviors (Block, 1987; Spreitzer, 1995); and managerial effectiveness (Spreitzer, 1995; Thomas & Velthouse, 1990). Siebert, Wang, and Courtridge (2011) found in their meta-analysis of employee empowerment that psychological empowerment was strongly related to employees' job satisfaction and organizational commitment. Maslach, Schaufeli, and Leiter (2001) defined engagement as an energetic, involved, and efficacious (empowered) state that is the opposite of burnout, cynicism, and inefficacy. This study adopts the view of empowerment as a motivational state that is internal to individuals (Conger & Kanungo, 1988) and proposes empowerment as representative of a motivated facet of engagement, both conceptually and operationally.

Part 2: Proposed Model of Public Engagement

Antecedents of Engagement

Review of engagement literature suggests two of the most important antecedents to engagement are trust and satisfaction.

Trust

Both conceptual and empirical studies found evidence to support the pivotal role of trust as a key determinant of relational commitment (e.g., Garbarino & Johnson, 1999; Tax, Brown, & Chandrashekaran, 1998), loyalty (e.g., Reichheld & Schefter, 2005; Sirdeshmukh, Singh, & Sabol, 2002), and an essential element for strong relationships and sustainable market share (e.g., Urban, Sultan, & Qualls, 2000). Trust and engagement have been linked in extant literature. For instance, trust has been identified as the most important factor that drives civic engagement. Fukuyama (1995) argued that trust engenders spontaneous sociability, which means “the myriad forms of cooperative, altruistic, and extra-role behavior in which members of a social community engage” (Kramer, 1999, p. 583). In Bowling Alone, Putnam (2000) also put emphasis on trust as the determining factor that has reduced civic engagement in American society since the 1960s. In organizational studies, trust in terms of perceived organizational support has been closely linked to higher employee engagement (e.g., Saks, 2006). Therefore, based on the theoretical linkage predicted in the literature, this study proposes a positively correlated relationship between trust and engagement.

H1: Trust leads to enhanced public engagement with organizations.

Prior research in consumer trust shows that trust between consumers and companies, customer loyalty, and positive WOM are positively associated (e.g., Zeithaml, Berry, & Parasuraman, 1996). Therefore, this study proposes that trust is a key antecedent of engagement, which leads to supportive behavioral outcomes such as WOM or loyalty intentions.

H2: Trust leads to supportive behavioral intentions that publics display toward organizations.

Satisfaction

Satisfaction has been extensively studied in customer and organizational management literature as the key to both short- and long-term organizational success (Henning-Thurau, & Klee, 1997) in terms of customer retention and communicative behaviors (positive and negative WOM) and as one of the key outcomes of relationships for decades (Garbarino & Johnson, 1999). Satisfaction and engagement have been closely linked in literature. Some scholars (e.g., Harter et al., 2002) have conceptualized engagement and satisfaction as directly linked (Macey & Schneider, 2008). This study postulates that satisfaction is a key antecedent to engagement in that individuals need to have a satisfying relationship with an organization before they become engaged with the organization.

H3: Satisfaction leads to enhanced public engagement with organizations.

The relationship between satisfaction and supportive behavioral outcomes, such as loyalty (e.g., Schaufeli & Bakker, 2004) and WOM (e.g., Anderson, 1998) are quite extensive in literature. Kotler (1994) summarized the satisfaction-loyalty literature by noting that customer satisfaction is the key to customer retention. Similarly, Ravald and Gronroos (1996) noted that customer satisfaction is one of the leading indicators of customer loyalty and overall customer satisfaction is a better predictor of intentions to repurchase than overall or inferred service quality. Therefore, the following hypothesis is proposed.

H4: Satisfaction leads to supportive behavioral intentions that publics display toward organizations.

Outcomes of Engagement: Supportive Behavioral Intentions

Positive WOM

Defined as “information communication directed at other consumers about the ownership, usage, or characteristics of particular goods and services and/or their sellers” (Westbrook, 1987, p. 261), WOM communication, has recently received renowned interests from scholars and professionals. Some scholars have suggested that WOM is “a dominant force in marketplace” (Mangold, Miller, & Brockway, 1999, p. 73) and the “ultimate test of the customer's relationship” (Bendapudi & Berry, 1997, p. 30).

Since Anderson's (1998) call for more research on antecedents of WOM communication, there have been many efforts by scholars to understand what may lead to increased customer WOM behaviors. Gremler, Gwinner, and Brown (2001) examined the effects of customer-employee relationships on positive WOM and found that the interpersonal relationships between customers and a service provider that are marked by trust resulted in greater positive WOM behavior by customers. Brown, Barry, Dacin, and Gunst (2005) also proposed a comprehensive model of the antecedents of positive WOM and found that customer satisfaction led to positive WOM intentions and behaviors. Hong and Yang (2009) reported positive corporate reputation along with customer satisfaction led to increased WOM intentions among the customers.

Loyalty

The concept of customer loyalty has received extensive interest in marketing history (Knox & Walker, 2001). For several decades, a primary objective of many marketing strategies was to generate loyal customers; the level of customer loyalty to a brand has been used as a whole or partial indicator of success for marketing campaigns and of brand equity (Aaker, 1991; Knox & Walker). As the relationship marketing perspective dominates consumer research, the concept of customer loyalty also has become paramount (Fournier & Yao, 1997) to effective relationship management efforts. That is, the creation of customer loyalty is first and foremost in relationship marketing; as a result, a stable, mutually profitable and long-term relationship is enhanced (Ravald & Gronroos, 1996).

The relationship between loyalty and engagement is evident in organizational research. Salanova, Agut, and Peiro (2005), for example, reported that a supportive organizational climate, which is affected by organizational support and employee work engagement, led to better employee performance and customer loyalty. Our study attempts to explore the essentiality of an affective notion of engagement on the formation of loyalty and positive WOM from members of an organization. The following hypothesis is proposed to test.

H5: Public engagement leads to supportive behavioral intentions such as positive WOM and customer loyalty that publics have toward organizations.

Mediating Role of Engagement

Much OPR research supports a direct link between OPR quality and various outcomes such as positive attitudes and positive reputations (e.g., Bruning, 2000; Hong & Yang, 2009; Ki & Hon, 2007). Now with engagement as an important concept that leads to positive behavioral supports by publics for an organization, this study proposes a research question to examine that the effects of two important antecedents (i.e., trust and satisfaction) on supportive behavioral intentions as mediated by engagement.

Research Question: To what extent does engagement mediate effects of key antecedents of engagement (i.e., trust and satisfaction) on supportive behavioral intentions that publics have toward organizations?

METHOD

Sample and Data Collection Process

The study population was all the donors, long-term subscribers, and single-ticket buyers of a professional theater organization in the United States, during the 2006 to 2010 seasons. This 40-year old, nonprofit theater organization has reported annual revenues of 1 million dollars and the annual production of eight or nine plays. The average number of patrons who come to see performances is around 90,000 per season. The e-mail contact information of these patrons was obtained via the theater's marketing director. The written result report was promised at the end of the study in exchange for such contact information. This study was approved by the IRB prior to the data collection and the informed consent form was attached in the recruiting e-mail with the consent signature waivered with a click of the link to the online survey. The link to the online survey was e-mailed to 4,425 randomly selected patrons. The survey was conducted for a week during the summer of 2010 and generated a random sample of 1084 complete cases (N = 1084) with response rate of 24.50%. Ten randomly selected respondents were awarded monetary support (cash value of $200) that may be used toward their yearly subscription for the next season. Among 976 research participants who answered the gender question, 721 participants (73.90%) were women. ¹ The mode of the participants' age ranged from 46 to 55 (n = 308; 31.50%). The duration of research participants' relationships as donors or patrons was mostly commonly five years or longer (n = 560; 61.50%), suggesting many of the research participants had substantial relationships with the organization studied. ²

Measurements

Engagement

This study defines the concept of engagement with three dimensions: affective commitment, positive affectivity and empowerment (See Table 1 for full items and the descriptive statistics for engagement).

TABLE 1 A Proposed Scale of Public Engagement: Descriptive Statistics and Exploratory Factor Analysis for Measurement Items (N = 1084)

	Factor loading
Item	EM ^a	PA	AC	M ^b	SD	%	α
1. Feel emotionally attached			.66	4.63	1.29	19.11%	.89
2. Feel like a part of family			.87	3.87	1.27
3. Feel a sense of belongings			.86	4.01	1.27
4. Interested		.82		5.48	1.04	26.64%	.89
5. Attentive		.77		5.16	1.14
6. Excited		.85		5.31	1.14
7. Enthusiastic		.85		5.39	1.14
8. Proud		.72		5.56	1.17
9. Can make differences	.76			3.37	1.25	26.81%	.89
10. Determined to develop the org	.79			3.28	1.27
11. Have a control over the org decision-making	.76			2.46	1.23
12. Confident about ability to improve the org	.87			3.00	1.30
13. Collaborate with org	.82			3.26	1.37
Total Variance Explained with overall α						72.56%	.91
Note. Extraction method was Principal Component Analysis with Varimax rotation.
^aEM = Empowerment; PA = Positive affectivity; AC = Affective commitment. % = Variance extracted. ^bSeven-point Likert Scale: 1 (strongly disagree) to 7 (strongly agree).

Affective commitment, conceptualized as emotional bonding and pride that brings additional efforts to sustain OPRs, was measured with three items from Allen and Meyer (1996) and Bansal et al. (2004). The resulting scale led to a Cronbach's alpha of .89.

Positive affectivity, conceptualized as elevated emotional tone of the engagement state (i.e., attention and energy), was measured with five items (Watson et al., 1988). All five items loaded on one factor, which explained about 70.43% of shared variance. The resulting scale led to a Cronbach's alpha of .89.

Finally, empowerment, conceptualized as self-efficacy and impact, was measured with the modified five-item scale from Spreitzer's (1995) empowerment scale and Bandura (1977) and Gist's (1987) self-efficacy. The resulting scale led to a Cronbach's alpha of .89.

To summarize, the proposed 13-item scale of engagement performed well with a Cronbach's alpha of .91. The proposed 13-item scale of engagement clearly retained three theoretical dimensions in exploratory factor analysis as shown in Table 1. The minimum standardized factor loading was .66 in the item of feel emotionally attached; standardized factor loadings range from .66 to .87. In terms of validity, approximately 73% of total variance was extracted by the proposed measurement system, suggesting this scale has sound explanatory power in explicating engagement (see the details on the proposed public engagement scale from Table 1).

Supportive Behavioral Intentions

Loyalty was measured with the following four items modified from Zeithaml et al.'s (1996) measure of supportive behavioral intentions on a seven-point Likert scale (i.e., 7 = strongly agree): “my first choice to attend performances” (M = 5.53; SD = 1.34); “attend more performances in the next few years (M = 5.59; SD = 1.16); “will continue to attend performances even if its prices increase somewhat” (M = 5.46; SD = 1.26); and “will continue to attend performances even if experiencing a few problems” (M = 5.30; SD = 1.26). The resulting scale led to a Cronbach's alpha of .77.

Positive WOM was measured with the following four items adopted from Brown et al.'s (2005) WOM intentions on a seven-point Likert scale (i.e., 7 = strongly agree): “encourage friends to attend performances” (M = 6.47; SD = .77); “encourage family members to attend performances” (M = 6.42; SD = .82); “recommend the organization to someone who asks my advice” (M = 6.42; SD = .81); and “say positive things about the organization and its performances to other people” (M = 6.46; SD = .78). The resulting scale led to a Cronbach's alpha of .89.

The proposed eight-item scale of supportive behavioral intentions performed well with a Cronbach's alpha of .88. Furthermore, the proposed eight-item scale of supportive behavioral intentions retained two theoretical dimensions clearly in exploratory factor analysis. The minimum standardized factor loading was .51 in the item of my first choice to attend performances; standardized factor loadings range from .51 to .92 (i.e., encourage friends to attend performances).

Trust

Using existing measurement systems for trust from Morgan and Hunt (1994), Ganesan (1994), and Hon and J. Grunig (1999) on a seven-point Likert scale (i.e., 7 = strongly agree), I defined the concept of trust in two dimensions: integrity and competency. Both dimensions have four measurement items. Overall, research participants reported they have a higher level of competence (four-item composite M = 5.61; SD = .90) than integrity (four-item composite M = 5.41; SD = .96).

To measure integrity, defined as the fairness and justice of the organization's management, the following four items were used on a seven-point Likert scale (i.e., 7 = strongly agree): “treats patrons fairly” (M = 5.85; SD = 1.05); “concerned about patrons” (M = 5.10; SD = 1.16); “sound principles guide management of the organization” (M = 5.24; SD = 1.10); and “does not mislead patrons” (M = 5.46; SD = 1.09). The resulting scale led to a Cronbach's alpha of .89. Additionally, to measure competence, defined as the extent to which individual patrons believe the organization's ability to achieve what is promised, the following four items were used on a seven-point Likert scale (i.e., 7 = strongly agree): “confident about service quality” (M = 5.62; SD = 1.00); “the organization has ability to accomplish what it says it will do” (M = 5.50; SD = 1.04); “the organization is known to be successful” (M = 5.65; SD = 1.01); and “competent in fulfilling patrons' expectations” (M = 5.67; SD = 1.01). The resulting scale led to a Cronbach's alpha of .91.

Satisfaction

Modifying Oliver's (1980) measure of overall satisfaction and Hon and J. Grunig's (1999) measure of relational satisfaction, I measured relational satisfaction as a single dimension using three items. To measure satisfaction, the following three items were used on a seven-point Likert scale (i.e., 7 = strongly agree): “generally speaking, pleased with the relationship with the organization” (M = 5.64; SD = 1.01); “satisfied with the organization” (M = 5.85; SD = 1.08); and “delighted with the organization” (M = 5.51; SD = 1.20). The resulting scale led to a Cronbach's alpha of .91.

Data Analysis

To analyze the data using structural equation modeling, a two-step process of latent path modeling (Byrne, 2006) was conducted. Because the data-model fit for the initial measurement model was satisfactory, I went to the structural phase. Second, in the structural phase, I compared the CFA model with the proposed structural model in terms of data-model fits. The bootstrap procedure (1,000 samples) was used to generate a 95% bias-corrected confidence interval in testing the mediation effect of public engagement for effects of trust and satisfaction on supportive behavioral intentions. Statistical results were evaluated according to multiple data-model fit indexes.

RESULTS

Confirmatory Factor Analysis (CFA)

First, I conducted initial CFA by imposing a model where all factors are allowed to covary. Then, I modified the initial CFA model by covarying error variance between affective commitment and empowerment (r = .33; p < .001). By this modification, data-model fits were substantially improved (Δχ ²  = 72.74, Δ df = 1, p < .01). Based on Hu and Bentler's (1999) joint-cutoff criteria, the final CFA model can be retained as a valid model: χ ² (14, N = 1084) = 95.46, p < .001, SRMR = .02, TLI (NNFI) = .95, RMSEA = .07 (90% Confidence Interval: .06, .09), and CFI = .98.

Structural Model Analysis: Testing Hypotheses

Hypothesis 1: Effect of Trust on Engagement

H1 posited trust would lead to public's enhanced engagement with the organization. This hypothesis was supported: B = .48, SE = .06, β = .50, p < .001. This path turned out to have the largest effect size among all relations of variables imposed in this study (see Table 2).

TABLE 2 Standardized Coefficient of Direct Effects in the Hypothesized Structural Equation Model for Latent Variable Analysis (N = 1084)

Independent factor		Dependent factor	H	B	SE	β
Baseline model
Relational trust	→	Supportive behavioral intentions		.22	.05	.24***
Relational satisfaction	→	Supportive behavioral intentions		.44	.04	.59***
Proposed model mediated by public engagement
Relational trust	→	Public engagement	H1	.48	.06	.50***
Relational trust	→	Supportive behavioral intentions	H2	.04	.06	.04
Relational satisfaction	→	Public engagement	H3	.19	.05	.23***
Relational satisfaction	→	Supportive behavioral intentions	H4	.38	.04	.50***
Public engagement	→	Supportive behavioral intentions	H5	.37	.05	.39***
Note. χ^2(14, N = 1084) = 95.46, p < .001, χ ^2/df = 6.82, SRMR = .02, TLI (NNFI) = .95, RMSEA = .07 (90% Confidence Interval: .06, .09), and CFI = .98.

Hypothesis 2: Effect of Trust on Supportive Behavioral Intentions

H2 posited trust would lead to publics' supportive behavioral intentions. The result partially supported this hypothesis. As shown in Table 2, in the baseline model (with mediation of engagement), effect of trust on publics' supportive behavioral intentions was statistically significant: B = .22, SE = .05, β = .24, p < .001. However, this effect became insignificant—to almost null/independent—in the mediated model: B = .04, SE = .06, β = .04, ns, because public's level of engagement strongly mediated the effect of relational trust on publics' supportive behavioral intentions toward the organization.

Hypothesis 3: Effect of Satisfaction on Engagement

H3 posited satisfaction would increase public's engagement with the organization. This hypothesis was supported: B = .19, SE = .05, β = .23, p < .001.

Hypothesis 4: Effect of Satisfaction on Supportive Behavioral Intentions

In addition to H3, H4 posited satisfaction would lead to publics' supportive behavioral intentions. This hypothesis was supported: B = .38, SE = .04, β = .50, p < .001. As in the path from trust to engagement, this path turned out to have a strong effect size. In both the baseline model and mediated model, the effect of satisfaction on supportive behavioral intentions remained statistically significant: β = .59 in the baseline model (p < .001); β = .50 in the mediated model (p < .001).

Hypothesis 5: Effect of Engagement on Supportive Behavioral Intentions

H5 posited public's engagement would lead to supportive behavioral intentions toward the organization. This hypothesis was supported: B = .37, SE = .05, β = .39, p < .001.

Testing Research Question: Mediation Analysis of Engagement

To the question of a mediation of public engagement, two structural models were compared in a hierarchical/nested relation in terms of the χ ² -df test: a model with structural paths from public engagement and another model (i.e., the baseline model) without such paths, to examine if the mediation model is statistically better than the baseline model.

The mediation model turned out to perform substantially better than the baseline. According to Hu and Bentler's (1999) joint-cutoff criteria, the baseline model was not acceptable: χ ² (18, N = 1084) = 611.78, p < .001, χ ² /df = 33.99, SRMR = .25, TLI (NNFI) = .73, RMSEA = .18 (90% Confidence Interval: .25, .18), and CFI = .86. Across all data-model fit criteria, it is clear that the baseline model without mediating paths via public engagement is not tenable as a valid model. However, when mediating paths via public engagement were added into the baseline model, its performance was significantly improved so as to be retained as a valid model: χ ² (14, N = 1084) = 95.46, p < .001, χ ² /df = 6.82, SRMR = .02, TLI (NNFI) = .95, RMSEA = .07 (90% Confidence Interval: .06, .09), and CFI = .98. Thus, although the baseline model is more parsimonious by 4 df (Δ df = 4), the mediation model should be selected as a better model: Δ χ ² (4, N = 1084) = 516.32, p < .001.

Also, the bootstrap procedure (1,000 samples) was used to generate a 95% bias-corrected confidence interval, testing the mediation effect of public engagement for effects of trust and satisfaction on supportive behavioral intentions. This study found a strong mediation of public engagement between relational trust and supportive behavioral intentions: β = .24 (95% Bias-Corrected Confidence Interval: .13, .25), p < .01. The results suggest that the direct link between trust and supportive behavioral intentions (β = .24, p < 001) were fully mediated by public engagement (β = .04, ns). Although the effect was not as much strong, the study also found a mediation of public engagement between relational satisfaction and supportive behavioral intentions: β = .09 (95% Bias-Corrected Confidence Interval: .09, .04), p < .01.

DISCUSSION

Contemporary public relations practice has increasingly emphasized the notion of public engagement with organizations due to the changing nature of publics' active communication behaviors and higher expectations to collaborate with organizations. This study introduced the concept of engagement as the focal concept that can bring behavioral supports such as loyalty and positive WOM from the publics.

First, this study defined the concept of engagement and suggested a sound measure at the micro-level as an affective connector that bridges two key concepts—trust and satisfaction—with the subsequent public's supportive behavioral intentions. Surprisingly lacking in public relations scholarship are affective components of relationships between organizations and publics. The introduction of engagement as a predominantly affective concept that connects the cognitive experiences to loyalty and positive referral communication behaviors demonstrates the importance of fostering intimate relationships with the publics that are not just characterized by cognitive beliefs and assessments of relational quality. Identifying engagement with three core dimensions of affective commitment, positive affectivity, and empowerment, this study suggested reliable ways to measure public's level of engagement with 13 items. The proposed 13-item scale of public engagement had sound reliability and validity and retained three theoretical dimensions clearly in exploratory factor analysis.

Second, this study found significant mediation effects of engagement between relationship variables (i.e., trust and satisfaction) and supportive behavioral intentions (i.e., loyalty and positive WOM support). This finding supports the level of engagement as being critical for linking public's level of trust and satisfaction with an organization to their actual supportive behavioral intentions. This study supports the assertion that publics' evaluations on the level of satisfaction and trust are necessary antecedents for publics to feel affectively motivated to act out their beliefs on organizations to support the organizations (Macey & Schneider, 2008). Further, the results support the introduction of engagement as an important motivator that connects the cognitive evaluations to publics' supportive actions. These findings indicate that publics' supportive behavioral intentions are more likely to occur when their interactions are based on high levels of trust and satisfaction as well as on affective bonding that is marked by feelings of passion, pride, and excitement. For example, in the context of the studied theater organization, highly satisfied patrons were more likely to come back to the theater for more performances and recommend them to their friends than less satisfied patrons of the theater. Also, engaged patrons were more likely to be excited about preperformance activities and had enjoyed interactions with the actors and the director pre- or postperformance than patrons who were less engaged. They were also more likely to be involved in fund-raising activities such as purchasing tickets to these extracurricular programs and auctions to support the theater.

Further, this study supports the importance of feelings of empowerment by the publics to behaviorally engage to support the organization. That is, when individuals feel that their efforts can make a difference in organizational success, they are more likely to show supportive behaviors. Particularly, feelings of empowerment, based on both cognitive and affective appraisals of situations and subsequent effects of their involvement, may provide additional insights into the understanding of individual public's choice and circumstances that lead him/her to become a loyal supporter or ardent protagonist.

Implications

Findings of this study can offer insights on explication and operationalization of engagement, which can be incorporated into different practice contexts. Public relations professionals can utilize the three dimensions of engagement to test which area of engagement is lacking in their current efforts to be connected with their key publics. More importantly, the results of this study indicate that organizational efforts can be most valuable when they are focused on increasing public engagement by fostering positive feelings, affective commitment, and a sense of empowerment via participation. That is, getting satisfactory and reliable services or goods is not sufficient for publics to become loyal with or an advocate for organizations. When trust and satisfaction are present in the interactions between organizations and publics, organizations have to work to develop a sense of a community that is centered on feelings such as affection and pride.

Further, by conceptualizing engagement as a psychological concept from the public's perspective, this study demonstrated the role of engagement in the ongoing OPR scholarship as a critical link between positive relational qualities and behavioral supportive outcomes. Practitioners who work for a member organization, for example, would understand from this study's findings that in order for them to increase member engagement, trust and satisfaction are vital. Also, the study findings suggest that WOM behavior (offline and online) is not to be equated with engagement but rather an outcome of engagement. This is particularly relevant in the ongoing discussion of engagement as more than a simple click to like or retweet.

Limitations and Future Research

This study does not come without a set of limitations. Even though there are sound fits between the data and the hypothesized structural model in this study, it is still possible that the proposed model might have been misspecified to some degree. Because a researcher cannot account for all potential causal elements in a hypothesized model, the scope of this research had to be delimited into relational trust and satisfaction. This limitation is also largely related to the study's reliance on quantitative survey data as the sole and primary method of inquiry. Future studies should incorporate qualitative approach for a richer and deeper exploration of this phenomenon. This study is only exploratory or suggestive in this regard; future studies can search for other relevant precursors of public engagement than what were not examined in this study. There is also the issue of peculiarity of the organization studied. The unique nature of the organization studied and the sample might have influenced the results of this study and limits its general application across different organizational contexts and stakeholders. The organization in this study has mixed characteristics of nonprofit and profit organizations and relies on financial support from donors and loyalty from its patrons for its operation. To examine public engagement in a valid, reliable manner, studying an organization with long-term donors and patrons was essential. Further, although typical art patronage tends to be skewed toward female population, the gender disparity between male and female survey respondents of the study might have impacted the results of the study to an extent. Finally, I acknowledge that public engagement is not inherently a positive affect. In this study, consistent with how engagement is conceptualized in organizational studies, engagement was viewed as the opposite force of disengagement or burnout. For future study, the concept of public engagement can be expanded to include negative aspects of engagement in the conceptualization and operationalization.

Notes

¹Because there were approximately three times more women, I tested possible gender differences in the measurement items of public engagement, using independent-samples t-tests. The results indicated that female participants, in general, reported higher levels of positive affectivity and empowerment than male participants. In the case of affective commitment, male participants reported higher level of engagement than female participants. However, such gender differences were found to be statistically significant only in the item of enthusiastic in positive affectivity, t(962) = − 2.26, p < .05.

²According to the most recent report by the 2002 National Endowment of Art on American Survey of Public Participation in the Arts, a typical theater patron is likely to be 45 years or older, Caucasian, female (1.4 times more than male patrons), and highly educated (roughly about 35% of theater patrons have graduate school education) with an income level of $75,000 or above.

***p < .001.