Development and Validation of a Facebook Relational Maintenance Measure

This manuscript details the construction of a measure of Facebook relational maintenance behaviors. The first study generated an item pool by drawing from previous qualitative investigations, and adapting an established relational maintenance scale. Participants were then invited to evaluate these items in order to establish face validity. During study two, participants were asked how often they used the behaviors represented in these items to maintain a specific friendship. Exploratory factor analysis was conducted to determine the underlying structure of these items, and three latent factors emerged, social contact, response-seeking, and relational assurances. This factor structure was then assessed using confirmatory factor analysis during phase three. Study three participants were also asked to complete measures of friendship quality, Facebook intensity, and online social communication. The relationship of the three factors of Facebook relational maintenance to friendship quality, Facebook intensity, and online social communication suggests convergent and discriminant validity for the Facebook relational maintenance measure. FACEBOOK RELATIONAL MAINTENANCE MEASURE 3 Development and Validation of a Facebook Relational Maintenance Measure People enact behaviors to maintain the existence and quality of their relationships (Dindia, 2003). Relational maintenance behaviors can be strategic or routine in nature, and are prevalent in varied relational contexts (Canary, Stafford, Hause, & Wallace, 1993). Early work on the measurement of relational maintenance focused primarily on face-to-face behaviors (e.g., Stafford & Canary, 1991); yet, the proliferation of computer-mediated communication such as social network sites (SNS), has provided new fora for relational maintenance (Walther & Ramirez, 2009). SNS serve important relational functions for users worldwide, and relational maintenance is considered to be one of the most important reasons for using SNS (Bryant, Marmo, & Ramirez, 2011; Debatin, Lovejoy, Horn, & Hughes, 2009; Ellison, Steinfield, & Lampe, 2011; Walther & Ramirez, 2009). Facebook continues to outperform other SNS in terms of number of users (Pew Research Center, 2013) making it an ideal place to examine relational processes. While traditional relational maintenance scales offer a starting point for considering maintenance behaviors that can be communicated via SNS, the unique affordances of SNS provide new maintenance opportunities that are worthy of consideration. Unfortunately, the measurement of SNS relational maintenance is inconsistent across studies and lacks empirical validation. Hence, the present study seeks to design and validate a Facebook Relational Maintenance Measure (FRMM) using items gleaned from traditional relational maintenance scales (e.g., Stafford, 2011) and SNS relational maintenance research (e.g., Bryant and Marmo, 2010; Cowden, 2012). The FRMM helps to unify relational and SNS measures of relational maintenance, and provides a validated scale that can be applied by researchers of Facebook and other SNS. Relational Maintenance via Facebook FACEBOOK RELATIONAL MAINTENANCE MEASURE 4 SNS allow users to display their identity, articulate social connections, and communicate with others in their network (boyd & Ellison, 2007). The use of SNS is pervasive among American adults and teens. More than 73% of online adults utilize SNS (Duggan & Smith, 2013), with 48% using SNS on an average day (Brenner, 2012). Ninety-two percent of these SNS users have a Facebook account (Hampton, Goulet, Rainie, & Purcell, 2011). Additionally, 94% of teens have a Facebook profile, with 81% saying Facebook is the profile they use most often (Pew Research Center’s Internet & American Life Teen-Parent Survey, 2012). Facebook surpassed 1.3 billion active users as of April 2013 (Facebook.com), and trails Google as the world’s second most trafficked website (Alexa.com). Approximately 618 million people visit Facebook on a daily basis (Facebook.com), which suggests that the site is a habitual part of users’ daily lives. Given Facebook’s steadfast usage and focus on interpersonal relationships, it is a well-suited site for relational maintenance research. Relational maintenance consists of strategic and routine behaviors that help people keep their relationships in desirable states (Canary & Stafford, 1994). Early research by Stafford and Canary (1991) developed a five-item typology called the Relational Maintenance Strategies Measure (RMSM). Maintenance strategies were inductively developed from married and dating participants’ response to an open-ended question, “What do you do to maintain a satisfactory relationship?” Each strategy consists of specific behaviors that people engage in to maintain their relationships. Within the RMSM, positivity involves efforts toward pleasant and cheerful communication; openness involves acts of self-disclosure; assurances include behaviors that indicate a relationship will persist; social network behaviors integrate a partner with other social ties; task sharing involves performing routine tasks and chores together. Later, Stafford (2011) FACEBOOK RELATIONAL MAINTENANCE MEASURE 5 refined the RMSM creating the Relational Maintenance Behavior Measure (RMBM), which included the additional strategies of understanding (e.g., showing sympathy and forgiveness) and relational talk (e.g., communication about the relationship). The RMSM and additional behaviors have guided relational maintenance research; however, the extent to which these typologies remain valid within the SNS context remains understudied. Walther and Ramirez (2009) labeled relational maintenance as “the greatest utility of social networking systems” (p. 302), with other scholars echoing the sentiment that relational maintenance is the primary function of SNS such as Facebook (Bryant et al., 2011; Debatin et al., 2009; Ellison et al., 2011). Facebook is a powerful relational maintenance tool because it is convenient, overcomes spatial distances, and reduces communication costs (Cowden, 2012; Dwyer, 2007). Moreover, asynchronous Facebook communication provides extra time for users to create and edit messages that maximize their self-presentational goals (O’Sullivan, 2000; Walther & Boyd, 2002). As a result of these features, Facebook users can maintain relationships with a large number of people with relative ease. Measuring Maintenance on Facebook. There is a well-documented need to understand Facebook relational maintenance behaviors (Bryant et al., 2011; Tong & Walther, 2011; Walther & Ramirez, 2009), but disagreement exists regarding the best approach. One approach has applied traditional relational maintenance measures to online environments (Houser, Fleuriet, & Estrada, 2012; Ledbetter, 2010). We argue that this represents a primarily deductive approach because these studies operate under the assumption that general maintenance behaviors can be accurately extended to specific online contexts. For example, Ledbetter (2010) adapted Stafford and Canary’s (1991) RMSM to study instant messaging, and concluded that, “the RMSM is a statistically sound tool for future online communication research” (p. 30). Houser and colleagues FACEBOOK RELATIONAL MAINTENANCE MEASURE 6 (2012) took a similar approach by utilizing the RMSM to study multiple online environments, including SNS. It, therefore, appears that the RMSM, and updated versions such as Stafford’s (2011) RMBM, might be successfully applied to study Facebook relational maintenance. The deductive approach of applying existing relational maintenance measures has provided useful insight, but is limited because offline communicative actions are not necessarily replicated in online behaviors (Tong & Walther, 2011). Certain offline relational maintenance behaviors, such as sharing tasks, might not easily translate to online social environments, such as SNS. For example, in Ledbetter’s (2010) study, task sharing was found to occur less frequently via instant messaging than face-to-face. Conversely, the openness strategy has been found to be more critical for online relational maintenance than offline (Rabby, 2007). Moreover, SNS such as Facebook might provoke the creation of entirely new behaviors that qualify as relational maintenance. Indeed, many SNS users report that browsing a friend’s profile makes them feel closer to that friend, even if one-on-one communication does not occur (Bryant & Marmo, 2010). This behavior might qualify as relational maintenance, despite lacking the active communication that one would expect from traditional offline conceptualizations (e.g., Canary et al., 1993). Utilizing a purely deductive approach to the development of a Facebook relational maintenance measure might fail to capture new forms of maintenance behaviors that SNS afford. A second approach to the study of SNS is evident in scholars’ attempts to inductively develop a measure of Facebook relational maintenance (Bryant & Marmo, 2010; Marmo & Bryant, 2010; Cowden, 2012). Cowden (2012), for example, conducted qualitative interviews to thematically analyze Facebook behaviors. Likewise, Bryant and Marmo (2010) carried out focus groups in which college students discussed their Facebook behaviors. The resulting relational maintenance scale included SNS-specific behaviors such as “poking” someone, utilizing FACEBOOK RELATIONAL MAINTENANCE MEASURE 7 Facebook applications and games, and passively keeping surveillance over a friend’s updates by scrolling via the newsfeed. The authors, however, noted that their inductively generated list of Facebook relational maintenance behaviors closely resembled existing offline typologies. For example, posting a funny Facebook wall comment might qualify as a positivity behavior and sharing one’s thoughts in a status update is a form of openness. Hence, Bryant and Marmo (2010) and Marmo and Bryant (2010) ultimately utilized Canary et al.’s (1993) relational maintenance typology to categorize their Facebook specific behaviors. In sum, several relational maintenance measures exist, but have problems precluding their direct application to the SNS context. The first issue deals with the aforementioned lack of consistency in how Facebook relational maintenance measures are developed. Both deductive and inductive approaches are useful, but their divergent results make it difficult to align the existing body of SNS relational maintenance research. The present study will address this issue by deductively applying traditional relational maintenance items from Stafford’s (2011) RMBM to the Facebook context, while also including items inductively developed for the Facebook context by Bryant and Marmo (2010) and Cowden (2012). Doing so should help provide a robust set of potential SNS relational maintenance behaviors for analysis. The second problem with existing research lies in the lack of an empirically validated SNS relational maintenance measure. For example, Dainton (2013) utilized Marmo and Bryant’s (2010) openness and positivity items, with a newly developed set of Facebook assurance items. While the items held a three-factor structure during CFA, Dainton did not include any additional RMBM strategies, leaving only a partial understanding of the larger scope of potential SNS relational maintenance. Marmo and Bryant (2010) present a more comprehensive set of six Facebook relational maintenance strategies, but their items failed to hold a six-factor structure FACEBOOK RELATIONAL MAINTENANCE MEASURE 8 when McEwan (2013) put them through quantitative factor analysis tests. In light of these conflicting results, the present study aims to determine if there is a unique, replicable factor structure underlying SNS relational maintenance behaviors. For research to effectively move forward in this area, and to provide consistency across studies for the purpose of replication and comparison, it is necessary to not only understand the relational maintenance strategies being utilized on SNS such as Facebook, but also to establish the psychometric properties of items. A valid Facebook relational maintenance measure must not only possess a sound factor structure, but should also demonstrate convergent and discriminant validity in line with previous relational maintenance research. Therefore, the present study seeks to empirically validate a Facebook relational maintenance measure that might also serve as a model or springboard for investigators studying relational maintenance on other SNS and socially oriented websites.