Sport communication research: A social network analysis

G Model

SMR-382; No. of Pages 14 Sport Management Review xxx (2016) xxx–xxx

Contents lists available at ScienceDirect

Sport Management Review journal homepage: www.elsevier.com/locate/smr

Sport communication research: A social network analysis Marion E. Hambrick[1_TD$IF]* University of Louisville, United States

A R T I C L E I N F O

A B S T R A C T

Article history: Received 30 October 2015 Received in revised form 27 July 2016 Accepted 18 August 2016 Available online xxx

Sport communication research has experienced exponential growth since the 1980s. As one of the four primary sport management functions, sport communication has formed a synergistic relationship with sport management. Researchers have documented this relationship and the continued role of communication within sport. The current study explored the evolution of sport communication research through social network analysis (SNA). This methodological approach offers a visual display of research collaborations and helps identify areas for growth—among researchers, academic institutions, and topics—in an effort to expand research productivity and diffusion. From January 1980 to June 2015, 1255 sport communication researchers shared 2537 collaborations and authored 1283 publications. Their studies most frequently examined topics such as gender, mass media, and sport consumption. The number of researchers, publications, collaborations, and researchers per publication increased over time. A select group of researchers hailed from a smaller number of universities and emerged as key contributors to the field. The findings underscore the importance of prominent researchers, academic institutions, and collaborations in the production of sport communication research. The study also outlines the benefits of using SNA to investigate a field’s development and growth opportunities. ß 2016 Sport Management Association of Australia and New Zealand. Published by Elsevier Ltd. All rights reserved.

Keywords: Sport communication Research Collaboration Social network analysis

1. Introduction Sport communication research has grown exponentially since the 1980s. This expansion has included the introduction of new sport communication textbooks, journals, associations, and conferences, and these outlets have given researchers an opportunity to disseminate ideas and advance the field (Abeza, O’Reilly, & Nadeau, 2014). Sport communication is defined as ‘‘a process by which people in sport, in a sport setting, or through a sport endeavour, share symbols as they create meaning through interaction’’ (Pedersen, Miloch, & Laucella, 2007, p. 196). Communication within sport has evolved to share a synergistic relationship with sport management. The Commission on Sport Management Accreditation (COSMA, 2016) lists sport communication as one of the four primary functions of sport management in conjunction with sport operations, sport marketing, and sport finance and economics. Pedersen (2013) asserted sport communication will continue to have a tremendous influence on sport—with the industry as a whole as well as its individual components of people, places, and events. He further argued, ‘‘sport cannot exist without communication’’ (p. 57). Indeed, textbook authors and researchers have highlighted this persistent connection between sport communication and sport management. The sport management textbook Contemporary Sport Management outlined sport communication’s

* Correspondence to: Department of Health & Sport Sciences, University of Louisville, 2314 South Floyd Street, Louisville, KY 40292, United States. Fax: +1 502 852 6683. E-mail address: [email protected] http://dx.doi.org/10.1016/j.smr.2016.08.002 1441-3523/ß 2016 Sport Management Association of Australia and New Zealand. Published by Elsevier Ltd. All rights reserved.

Please cite this article in press as: Hambrick, M.E., Sport communication research: A social network analysis. Sport Management Review (2016), http://dx.doi.org/10.1016/j.smr.2016.08.002

G Model

SMR-382; No. of Pages 14 2

M.E. Hambrick / Sport Management Review xxx (2016) xxx–xxx

standing within the broader field of sport management in various contexts at the interpersonal, group, and organisational levels (Stoldt, Dittmore, & Pedersen, 2014). The sport communication textbook Strategic Sport Communication also discussed this relationship with the varying forms of sport communication used in sport management: personal and organisational communication, mass media, and communication services and support (Pedersen et al., 2007). From a research perspective, Pedersen (2013) forecasted related research would continue to emerge as scholars examine the ongoing interplay between sport communication and sport management. Sport management journals have featured this research, which addresses both theoretical and practical implications related to the utilisation of sport communication within sport management (Abeza, O’Reilly, Seguin, & Nzindukiyimana, 2015; Filo, Lock, & Karg, 2015; Fink, 2015). This research has included examinations of mass media and social media in the sport industry as well as media portrayals of gender, race and ethnicity, and nationality in sporting events. The current study builds upon the extant work and explores this extensive research activity in greater detail. As a field advances, researchers may investigate and document its development and evolution (Moody, 2004). Sport studies have included content analyses of publications from a field’s flagship journal or other representative journals in order to identify salient works, trends, and potential gaps for exploration (Abeza et al., 2015; Filo et al., 2015; Fink, 2015). Abeza and colleagues (2014) studied sport communication research activity in their review of research published in the International Journal of Sport Communication. Sport management journals also have included this research (Abeza et al., 2015; Filo et al., 2015). Sport Management Review featured Fink’s (2015) research regarding media coverage of women’s sports and female athletes. Her study concluded that while some progress has occurred with this coverage, more research should address and help to mitigate issues with hegemonic masculinity, negative stereotypes, and minimal media exposure for female athletes and women’s sports. In their research reviews published in the Journal of Sport Management and Sport Management Review, respectively, Abeza and colleagues (2015) and Filo and colleagues (2015) concentrated their focus on social media research within sport management. The two reviews concluded by emphasising the need to incorporate more theoretical frameworks and to include more research using diverse methodologies in order to advance social media research in sport management. With their focus on sport communication within sport management, the studies underscored the interaction between the two and the benefits of continued exploration. Thus, the current study sought to extend the previous sport communication research and explore its growth in greater detail by examining the three sport communication journals—Communication & Sport, International Journal of Sport Communication, and Journal of Sports Media—in conjunction with sport management journals such as the International Journal of Sport Management and Marketing, Journal of Sport Management, and Sport Management Review and other academic outlets disseminating this research. This study used social network analysis to better understand the researchers and their scholarly activity and to document the field’s development (Love & Andrew, 2012; Quatman & Chelladurai, 2008). Social network analysis (SNA) provides tools to study social networks, which contain network members (individuals, groups, and/or organisations) and the relationships shared among them. The network members and their relationships coalesce to form a social network, where members can disseminate information, knowledge, and other resources (Scott, 2013; Wasserman & Faust, 1994). Studies have used SNA to investigate research relationships and trends, as a field’s scholars and their collaborative activities help to define the field and further its development (Crane, 1969; Crawford, 1971; Moody, 2004). Only two studies have used this approach with sport research. Quatman and Chelladurai (2008) examined sport management studies and identified substantial productivity attributed to a small group of influential researchers. Similarly, Love and Andrew (2012) investigated the relationship between sport management and sport sociology research. They also documented significant growth as well as a smaller number of researchers who served to bridge the two areas. The current study sought to extend the existing literature and address this research gap by conducting a similar investigation focused on the sport communication field and by using SNA to outline the field’s evolution and opportunities for future research. 2. An investigation of research collaborations Studies have examined research collaborations to better understand how researchers produce scholarly work (Endersby, 1996; Moody, 2004). ‘‘Co-authorship is the most formal manifestation of intellectual collaboration in scientific research’’ (Acedo, Barroso, Casanueva, & Gala´n, 2006, p. 959), and studies have documented these collaborations in greater detail (Crane, 1969; Crawford, 1971; Moody, 2004). An array of research collaborations exist, including academic mentor relationships resulting in dissertations and theses as well as academic research partnerships to create conference presentations and textbooks (Crane, 1969; Crawford, 1971; Katz & Martin, 1997; Laband & Tollison, 2000). Most studies to date have concentrated on documenting journal article collaborations (Scott, 2013). Seminal studies by Price (1965) and Merton (1968) laid the foundation for the current research on scholarly activity. Two primary propositions emerged from their work: (a) Price’s (1965) ‘‘invisible colleges,’’ where a smaller group of influential researchers benefit from collaborations outside of or in advance of the formal academic publication process, and (b) Merton’s (1968) Matthew effect, or ‘‘the principle of cumulative advantage that operates in many systems of social stratification to produce the same result: the rich get richer at a rate that makes the poor become relatively poorer’’ (p. 62). This occurs when popular researchers accrue numerous publications and citations, leading to additional funding and collaborations. Subsequent studies examining the evolution of research have provided support for these two propositions. Researchers have investigated collaborations found within the fields of management (Acedo et al., 2006); operations, marketing, and

Please cite this article in press as: Hambrick, M.E., Sport communication research: A social network analysis. Sport Management Review (2016), http://dx.doi.org/10.1016/j.smr.2016.08.002

G Model

SMR-382; No. of Pages 14 M.E. Hambrick / Sport Management Review xxx (2016) xxx–xxx

3

human resource management (Martins, Martins, Csillag, & Pereira, 2012); sociology (Moody, 2004); tourism and hospitality (Racherla & Hu, 2010; Ye, Li, & Law, 2013); and other social sciences (Endersby, 1996). Their studies have documented the presence of more publications and authors over time (Endersby, 1996; Ye et al., 2013). The number of collaborations also has increased (Acedo et al., 2006; Martins et al., 2012; Moody, 2004; Racherla & Hu, 2010; Ye et al., 2013). With this research activity often comes a smaller, select group of authors (Acedo et al., 2006; Crane, 1969; Martins et al., 2012). These individuals have an extended tenure within the field (Moody, 2004), produce more research (Acedo et al., 2006; Moody, 2004; Racherla & Hu, 2010; Ye et al., 2013), and collaborate more frequently with one another (Acedo et al., 2006; Newman, 2001; Racherla & Hu, 2010). They also connect different fields or focal areas (Martins et al., 2012) as well as geographic locations and academic institutions (Racherla & Hu, 2010; Ye et al., 2013). They may come from esteemed academic institutions and hold leadership positions within academic journals and professional associations (Acedo et al., 2006). Benefits frequently accrue to the authors, and these include opportunities to create new research (Crane, 1969). Studies examining various research fields can identify important contributors and relationships within a given area. They can outline where interdisciplinary efforts occur in contrast to research silos and fragmentation. Additionally, SNA can help determine how a field has evolved and where potential expansion exists in the future. 2.1. The study of sport-related research collaborations Yet despite the popularity of this examination in the social sciences, only two sport studies have used SNA to document research collaborations and trends. Their works explored sport management (Quatman & Chelladurai, 2008) and the link between sport management and sport sociology (Love & Andrew, 2012). Quatman and Chelladurai (2008) first used SNA to examine sport management as a field. The authors studied this research activity from pre-1985 to 2007 and documented exponential growth. This expansion included increases in the number of researchers, publications, collaborations, and researchers collaborating on a single publication. A core group of scholars produced a significant portion of the resulting research; others also benefited from this emphasis on collaboration. The findings revealed continued scholarship among existing research partners and the inclusion of new researchers. Love and Andrew (2012) continued the work of Quatman and Chelladurai (2008) in their study of sport management and sport sociology research. They also documented considerable scholarly output. Collaboration increases occurred within sport management and to a lesser extent within sport sociology. Sport management demonstrated greater cohesiveness, as a large group of researchers connected and collaborated. This activity facilitated the diffusion of knowledge within the field. Additionally, a smaller group exhibited proficiency in both fields and helped to bridge the two. Love and Andrew applauded the bridging efforts but cautioned that the removal of three to five scholars in particular could cause the link between the sport management and sport sociology fields to dissolve. Thus, their study identified strategies for researchers in both areas to sustain these relationships. The above studies (Love & Andrew, 2012; Quatman & Chelladurai, 2008) extend the previous work examining research collaborations. They also support the need for additional investigations within sport research. 2.2. Research collaborations and social network analysis One similarity of these studies (Acedo et al., 2006; Love & Andrew, 2012; Moody, 2004; Quatman & Chelladurai, 2008; Ye et al., 2013) is their use of SNA to better understand research collaborations and the resulting outcomes (Racherla & Hu, 2010). SNA provides an opportunity to investigate research collaborations—their development and the participating researchers—in more detail. This methodological approach proposes a social network perspective, which places emphasis on a social network’s members and their shared relationships. Applying SNA to a research field, the social network includes the researchers as the social network members and their research collaborations as the shared relationships. This combination of researchers and collaborations forms the field’s social network, which can be studied to assess its research activities and growth potential (Scott, 2013). Researchers have used SNA to dissect research network structures and their properties (Baraba´si et al., 2002), and some similar findings have emerged. Baraba´si and Albert (1999) detailed attributes of these social networks, including the presence of scale-free networks and preferential attachment. The social networks are often characterised as scale-free, meaning a small number of members make the largest contributions and possess the most resources within the network (Baraba´si & Albert, 1999). The departure of productive network members would have a notable effect on the network’s overall structure. These members benefit from favoured positions within the network and accrue resources through preferential attachment, whereby other network members gravitate towards and seek to develop relationships with them. Network members have a limited capacity to make connections and want to maximise their efforts. Connecting with those who possess the most resources (e.g., grant funding, graduate assistants) represents an efficient use of time. Preferential attachment gives members with the most resources additional benefits as others seek to partner with them (Baraba´si & Albert, 1999). These network attributes provide support for the hypothesised invisible college (Price, 1965) and Matthew effect (Merton, 1968) occurring within a field. In examining the fields as social networks, studies have documented sizeable expansion, frequently achieved through an increased number of publications over time (Crane, 1969; Crawford, 1971; Endersby, 1996; Love & Andrew, 2012; Martins

Please cite this article in press as: Hambrick, M.E., Sport communication research: A social network analysis. Sport Management Review (2016), http://dx.doi.org/10.1016/j.smr.2016.08.002

G Model

SMR-382; No. of Pages 14 4

M.E. Hambrick / Sport Management Review xxx (2016) xxx–xxx

et al., 2012; Moody, 2004; Quatman & Chelladurai, 2008; Racherla & Hu, 2010). This could occur through greater productivity by existing researchers, an influx of new researcher contributions, or a combination of the two. Sport communication studies also have identified increases in research productivity and the number of researchers (Abeza et al., 2014; Pedersen, 2015). The current study used a social network perspective to document the sport communication field’s development. Previous studies also have noted increases in the number of researchers collaborating in order to produce academic publications (Endersby, 1996; Katz & Martin, 1997; Laband & Tollison, 2000; Racherla & Hu, 2010; Ye et al., 2013). The studies have outlined factors leading to increased collaborations such as more interdisciplinary research, greater proximity to potential researchers, and better communication methods (Katz & Martin, 1997; Laband & Tollison, 2000); increases in expertise, funding, and division of labour (Endersby, 1996; Katz & Martin, 1997; Laband & Tollison, 2000); and research training needs (e.g., advisors mentoring younger researchers; Endersby, 1996; Katz & Martin, 1997). While the current study did not pursue specific explanations for research collaborations, it did explore whether a growing number of collaborations arose within the field. This study also investigated whether a larger number of research collaborations on a single study occurred. Studies have reported certain scholars often take part in more collaborative works with other researchers, including their advisees and peers (Acedo et al., 2006; Crane, 1969; Katz & Martin, 1997; Moody, 2004; Price, 1965; Racherla & Hu, 2010). These collaborations can increase productivity by helping researchers partner with prolific researchers and train the next generation of scholars. Some scholars also can serve as academic gatekeepers, dictating what gets published and how new researchers are trained (Crane, 1969; Crawford, 1971; Merton, 1968; Price, 1965; Quatman & Chelladurai, 2008). The current study used SNA to explore whether the sport communication field contained an influential group of individuals who played a substantial role in the field’s expansion. The above studies provide evidence that using SNA to examine research collaborations can offer insights regarding a field’s formation and development. This approach also can help in identifying the most significant researchers within the field. The purpose of this study was to assess the sport communication field and its research productivity, operationalised through academic publications, over time. The following research questions were examined. RQ1. How did the sport communication field evolve based on the (a) number of researchers, (b) number of publications, (c) number of collaborations, (d) number of researchers collaborating on a single publication, and (e) research area? RQ2. To what extent did prominent researchers emerge within the sport communication field, and what was their effect on the field? 3. Method The study employed SNA to address the research purpose and questions. 3.1. Sample and data collection Studying research fields as social networks requires (a) clear specification of the social network’s boundaries (i.e., who should be included versus excluded) and (b) inclusion of the identified network members within the study’s sample. SNA demands the inclusion of every viable network member (or most of them) in the dataset to provide a complete view of the social network and to use the accompanying network variables and analysis. Exclusion of one or more notable members and their respective relationships represents a threat to validity, because the failure to include missing members could skew the network’s portrayal and the study’s findings (Borgatti, Everett, & Johnson, 2013; Scott, 2013). To increase validity, previous studies of research collaborations have used one of three approaches with their respective samples: (a) The sample contained researchers who have published in a certain subset of journals, typically the most highimpact journals or those most germane to the research in question. (b) The study used snowball sampling to first identify the most influential scholars within the field through citation counts and/or suggestions from knowledgeable researchers. The sample then incorporated additional citations and researchers from the initial citations and suggestions. (c) The study combined the two approaches (Quatman & Chelladurai, 2008; Scott, 2013). The current study used a combined approach and centred on researchers publishing peer-reviewed sport communication research articles in related sport management, communication, and other academic journals. These publications included empirical and conceptual articles, case studies, and reviews, but the sample excluded introductions to special issues, book reviews, and other commentaries. This decision followed previous studies adopting a similar approach (Acedo et al., 2006; Endersby, 1996; Laband & Tollison, 2000; Love & Andrew, 2012; Martins et al., 2012; Racherla & Hu, 2010). First, complete citations from Communication & Sport, International Journal of Sport Communication, and Journal of Sports Media were collected, making the assumption that individuals who published in one or more of the journals possessed at least a minimal interest in sport communication research. The citations for the researchers were gathered through ProQuest, EBSCO, and Google Scholar. Citation data were input into Excel, and the details included the researcher(s) for each publication, publication title, publication year, and journal title. This list of researchers was examined, and citations for newly included researchers were obtained through the aforementioned sources. Citations for the researchers were added to the existing list, and the process of identifying new researchers and adding their citations was followed for third time. Next, the resulting data were sorted by journal titles to identify other outlets for sport communication research beyond the three journals cited above. They included sport management journals such as Journal of Sport Management and Sport

Please cite this article in press as: Hambrick, M.E., Sport communication research: A social network analysis. Sport Management Review (2016), http://dx.doi.org/10.1016/j.smr.2016.08.002

G Model

SMR-382; No. of Pages 14 M.E. Hambrick / Sport Management Review xxx (2016) xxx–xxx

5

Management Review as well as communication journals such as Communication Studies, Journalism & Mass Communication Quarterly, and Mass Communication & Society. A search of potential sport communication articles published in these and other journals was conducted using the above databases and each journal’s online search function. A reading of each article’s title, keywords, and abstract helped to determine whether the article should be part of the dataset (Racherla & Hu, 2010). Those which contained some topical connection to sport communication research were included (e.g., discussed mass media or social media in sports; investigated sport communication professionals; assessed the roles and portrayal of gender, race, or nationality within communication channels). The citations were added to the list, and the search for newly identified researchers and journals was completed again. The process of adding citations from the emerging researchers and journals was conducted an additional three times. The final searches most commonly uncovered researchers and journals with only one sport communication citation, and a determination was made to conclude the search process after these last searches. Previous studies have noted a small percentage of researchers within a field produce a large portion of the work (Acedo et al., 2006; Quatman & Chelladurai, 2008; Racherla & Hu, 2010) in comparison to those who may publish once or twice before exiting the field (Crane, 1969; Moody, 2004; Racherla & Hu, 2010). Citations identified in the latter rounds included researchers or journals receiving a single reference. Thus, data saturation occurred when the final searches yielded few new researchers and journals beyond the existing ones collected in the earlier rounds. The data collection process ended in June 2015, and the sample contained sport communication research published through the end of that month. The resulting data were cleaned to ensure data consistency and accuracy. With SNA of research collaborations, Borgatti and colleagues (2013) and Scott (2013) stressed the importance of researcher name accuracy, particularly with instances where two researchers share the same name; researchers initiate name changes during their publishing careers; or researchers use one combination of initials, nicknames, or names in some citations but a different combination in other citations. The data were sorted via Excel and checked to ensure names were reported consistently and no citation duplications occurred. Next, details regarding where each researcher obtained his or her doctoral degree and the academic institution of current employment, where applicable, were obtained. This information would be used to assess the potential effects of doctoral advising relationships (Quatman & Chelladurai, 2008) and institutional affiliations (Love & Andrew, 2012; Racherla & Hu, 2010) on collaborative activities. The outlined process yielded 1255 researchers who authored 1283 sport communication publications. To capture the research collaborations resulting in these publications, a separate case was created in Excel for each relationship. For example, a publication with three researchers would equate to three cases in order to depict the relationships between Researchers A and B, Researchers A and C, and Researchers B and C. Solo publications resulted in one case listed as a relationship between Researcher A and A. A total of 2537 collaborations were included in the analysis. Each sport communication researcher equated to a node, and each collaboration equated to a line (Hansen, Shneiderman, & Smith, 2011). This combination of nodes and lines, or researchers and their collaborations, reflected the sport communication field’s social network. Finally, the articles included in the sample were categorised into research areas in order to document trends in research foci over time. The categories were determined by first reading the title, keywords, and abstract for each article (Racherla & Hu, 2010). This led to 245 different codes, which were further analysed until a smaller set of 20 categories emerged. Additional readings were conducted, leading to a more parsimonious ten research areas for analysis: (a) Crisis – scandals and image repair; (b) Gender – portrayal and discussion of females and males in the media; (c) Industry – sport communication professionals and related issues; (d) Marketing – use of communications to sell products; (e) Media – role of mass media within sports; (f) Nationalism – portrayal and discussion of nationality or national identity in the media; (g) Olympics and Paralympics – broader media discussion of these mega-events beyond individual athletes or countries; (h) Race – portrayal and discussion of race and ethnic identity in the media; (i) Sport Consumption – use of media to facilitate fanship and other sport consumer behaviours; and (j) Other – portrayal and discussion of sexuality and ability in the media, organisational communication, and studies examining sport communication research. 3.2. Data analysis The study used this SNA data to examine two research questions, and the following provides an overview of how each research question was addressed. RQ1. How did the sport communication field evolve based on the (a) number of researchers, (b) number of publications, (c) number of collaborations, (d) number of researchers collaborating on a single publication, and (e) research area? First, frequencies, means, and standard deviations were examined to determine the number of publications, researchers, collaborations, and researchers collaborating on a single publication. Frequencies also were used to report potential shifts in research areas over time. Next, sociograms, or visual representations of the social networks, were created and explored. NodeXL, an SNA software tool, allows researchers to create these representations (Hansen et al., 2011). The sociograms present the combination of nodes and lines and portray the network members and their shared relationships, respectively. In this case, nodes represented the researchers, and lines connecting researchers represented their collaborations. Sociograms can display a single point in time, and a series of sociograms for a social network can reveal the development and evolution of the network. The current study used a sociogram series to assess the field’s social network over four periods: Period 1 (1980–1989), Period

Please cite this article in press as: Hambrick, M.E., Sport communication research: A social network analysis. Sport Management Review (2016), http://dx.doi.org/10.1016/j.smr.2016.08.002

G Model

SMR-382; No. of Pages 14 M.E. Hambrick / Sport Management Review xxx (2016) xxx–xxx

6

2 (1990–1999), Period 3 (2000–2009), and Period 4 (2010–June 2015). Previous studies have presented time periods ranging from four-year (Ye et al., 2013) to five-year periods (Martins et al., 2012; Racherla & Hu, 2010; Quatman & Chelladurai, 2008). Examining the sport communication social network in its entirety and over four periods would detail how the researchers interacted and produced research during the 36-year timeframe. In conjunction with sociograms, SNA software tools such as NodeXL offer quantitative data about the social networks (Hansen et al., 2011). Density quantifies the existing number of relationships within a social network and compares this to the number of possible relationships within the network. Density can range from 0.000, where no relationships exist among network members, to 1.000, where all network members share relationships. In larger networks, density values are typically smaller, as network members have a reduced capacity to establish and maintain relationships with numerous network members. Lower density values indicate a more limited ability to diffuse information and other resources across the network, because network members share fewer connections (Wasserman & Faust, 1994). RQ2. To what extent did prominent researchers emerge within the sport communication field, and what was their effect on the field? SNA statistics provide information about the social network as a whole; they also offer details about individual members within the social network. One SNA statistic type, centrality, indicates a network member’s position (Lusher, Robins, & Kremer, 2010). This information provides insights regarding which network members assume key locations within the network and serve to increase the network’s resource exchanges. The current study examined the following centrality values: (a) degree centrality, (b) betweenness centrality, and (c) eigenvector centrality. Degree centrality captures ‘‘a node’s connection to other nodes’’ (Tremayne, 2014, p. 113), where network members with the highest degree centrality values are deemed the most popular members within the network. Betweenness centrality indicates ‘‘the likelihood a particular node is a bridge between two other nodes. These nodes are important conduits of information’’ (p. 113). Network members with high betweenness centrality values serve as connectors within the network, linking members to other members. Lastly, eigenvector centrality ‘‘considers not only the volume of ties but proximity to other well-connected nodes’’ (p. 113). The lowest eigenvector centrality values are reserved for members who have the best connections within the network. The current study used the three centrality values to identify key network members. Additionally, researcher institutional affiliations would help to determine whether certain universities contributed to the field’s development. 4. Results SNA data collection and analysis were used to address the two research questions. The results of this analysis are discussed in greater detail below. RQ1. How did the sport communication field evolve based on the (a) number of researchers, (b) number of publications, (c) number of collaborations, (d) number of researchers collaborating on a single publication, and (e) research area? From January 1980 to June 2015, 1255 sport communication researchers authored 1283 publications, and the number of researchers increased over this period (Table 1). Of the 1255 researchers, 30 published in Period 1 from 1980 to 1989, 129 published in Period 2 from 1990 to 1999, 442 published in Period 3 from 2000 to 2009, and 843 published in Period 4 from 2010 to June 2015. (The summation of the numbers exceeds the 1255 reported, because some researchers published in more than one period.) The number of publications also increased over time. The researchers authored 25 publications (2% of the total publications) in Period 1, 100 (8%) in Period 2, 416 (32%) in Period 3, and 742 (58%) in Period 4. Their participation included an increasing number of collaborations across the periods. A total of 2023 collaborations occurred within the research network with 9 collaborations (0.4% of the total collaborations) in Period 1, 140 (7%) in Period 2, 556 (27%) in Period 3, and 1318 (65%) in Period 4. The number of solo authorships also increased over the 36 years. The researchers had 514 solo authorships with 16 (3% of the total solo authorships) in Period 1, 44 (9%) in Period 2, 189 (37%) in Period 3, and 265 (52%) in Period 4. The number of researchers collaborating on a single paper increased across the periods. The average number of researchers on one publication equalled 2.005 (SD = 1.095). Examining the collaborations by period Table 1 Social network analysis metrics for the sport communication research network. Network metrics

Researchers Publications Collaborations Solo authorships Total Average researchers/publication Network density Average degree centrality Average betweenness centrality Average eigenvector centrality

Period 1

Period 2

Period 3

Period 4

All

1980–1989

1990–1999

2000–2009

2010–2015

1980–2015

30 25 9 16 25 1.360 2.07% 1.600 0.033 0.033

129 100 140 44 184 1.890 1.45% 2.419 1.124 0.008

442 416 556 189 745 1.882 0.45% 2.525 4.776 0.002

843 742 1318 265 1583 2.111 0.29% 2.859 86.734 0.001

1255 1283 2023 514 2537 2.005 0.19% 2.924 209.331 0.001

Please cite this article in press as: Hambrick, M.E., Sport communication research: A social network analysis. Sport Management Review (2016), http://dx.doi.org/10.1016/j.smr.2016.08.002

G Model

SMR-382; No. of Pages 14 M.E. Hambrick / Sport Management Review xxx (2016) xxx–xxx

7

Table 2 Number of articles by research area and period. Research area

Period 1

Period 2

Period 3

Period 4

All

Crisis Gender Industry Marketing Media Nationalism Olympics/Paralympics Other Race Sport consumption Total

3 3 6 0 5 2 4 0 0 2 25

12 32 4 1 14 14 3 6 9 5 100

42 114 47 21 74 32 15 12 29 30 416

57 163 55 60 95 59 40 47 42 124 742

114 312 112 82 188 107 62 65 80 161 1283

revealed an average of 1.360 (SD = 0.480) in Period 1, 1.890 (SD = 1.057) in Period 2, 1.882 (SD = 1.013) in Period 3, and 2.111 (SD = 1.142) in Period 4. Of the ten research areas, Gender proved most popular (Table 2). The researchers wrote 312 articles related to Gender (24% of the total), followed by 188 (15%) related to Media and 161 (13%) related to Sport Consumption. Researchers in Period 1 focused on Industry with six articles (24% of the period) and Media with five articles (20%). In the next periods, researchers shifted to Gender with 32 (32% of the period) in Period 2, 114 (27%) in Period 2, and 163 (22%) in Period 4. Media articles remained popular in Periods 2 and 3 with 14 (14%) and 74 (18%), respectively. Sport consumption articles increased in Period 4 with 124 (17%). An increase in the number of researchers corresponded with a decrease in network density. The density for the field’s social network was 0.19%, where less than 1% of the possible network relationships (i.e., collaborations) existed among the researchers. The density value was highest at 2.07% in Period 1, which had the smallest number of researchers. This value continued to decrease to 1.45% in Period 2, 0.45% in Period 3, and 0.29% in Period 4, respectively. The sociograms provided a visual depiction of the field’s social network overall and for the four periods. The sociogram for January 1980 to June 2015 revealed many researchers were located centrally in the network with a smaller group forming the network’s nucleus (Fig. 1). Other researchers were dispersed throughout the network. Instances of solo authorship were evidenced by circular loops within the sociogram, where a researcher linked back to himself or herself, and most of these loops were found in the network’s outer region. The Period 1 sociogram contained the smallest number of researchers and relationships, but demonstrated the highest [(Fig._1)TD$IG]density as the researchers collaborated with one another (Fig. 2). A small portion of researchers were located at the network’s

Fig. 1. Sociogram of the sport communication research network, January 1980–June 2015.

Please cite this article in press as: Hambrick, M.E., Sport communication research: A social network analysis. Sport Management Review (2016), http://dx.doi.org/10.1016/j.smr.2016.08.002

G Model

SMR-382; No. of Pages 14

[(Fig._2)TD$IG]

8

M.E. Hambrick / Sport Management Review xxx (2016) xxx–xxx

Fig. 2. Sociogram of the sport communication research network, Period 1 1980–1989.

centre, which featured circular loops indicating solo authorships. The network also included researchers more widely scattered along the periphery. This sparse network structure reflected the field’s initial development. The Period 2 sociogram increased in size in comparison to the number of researchers and publications found in the first period (Fig. 3). This network contrasted with the prior period’s network, which featured fewer researchers and relationships. The Period 2 sociogram instead revealed a collection of researchers spread throughout the network. The network’s centre continued to reflect a concentration of researchers who engaged primarily in solo work. The Period 3 sociogram again increased in size over the previous period (Fig. 4). The number of researchers, publications, and relationships grew considerably. The network’s centre became more populated, and groups of researchers coalesced there. A larger number of solo authorships also emerged within the network. Researchers still persisted along the periphery, but more researchers help to enlarge the network. Finally, the Period 4 sociogram most closely resembled the overall network with the largest number of researchers and relationships (Fig. 5). More researchers helped to fill the network’s centre, and others emerged along the periphery. Solo authors remained at the network’s edges. These researchers became less central than in previous periods, representing a smaller portion of the overall productivity within the network. RQ2. To what extent did prominent researchers emerge within the sport communication field, and what was their effect on the field? Moving from the network to the members within it, centrality values were examined for each member. Degree centrality quantifies the number of connections members have with other members. Betweenness centrality indicates whether members hold influential positions within the network, linking members to other members. Eigenvector centrality assesses whether members connect to other prominent members. The average degree centrality for the January1980–June 2015 social network equalled 2.924 (SD = 3.281). This value increased over time with 1.600 (SD = 0.611) in Period 1, 2.419 (SD = 1.790) in Period 2, 2.525 (SD = 2.006) in Period 3, and 2.859 (SD = 2.996) in Period 4 (Table 3). The researchers generated more collaborations from one period to the next. Billings (50) emerged as the top researcher with the most frequent collaborative activity, followed by Pedersen (36), Hardin (35), and Clavio (32), respectively. The lead researchers shifted by period with Theberge (3) and Trujillo (3) as the most central in Period 1, Eastman (13) in Period 2, Pedersen (20) in Period 3, and Billings (33) in Period 4, respectively. The average betweenness centrality for the research network equalled 209.331 (SD = 1440.794). This value also increased over time with 0.033 (SD = 0.180) in Period 1, 1.124 (SD = 6.769) in Period 2, 4.776 (SD = 26.674) in Period 3, and 86.734 (SD = 505.530) in Period 4. In addition to sharing more collaborations, the researchers helped establish links between other researchers. Billings (33,108) assumed the top position in the betweenness centrality rankings, followed by Hardin (17,052), Kian (11,872), and Vincent (11,473). Shifts also occurred over time, where Salwen (1) claimed the top betweenness centrality position in Period 1, Eastman (63) in Period 2, Pedersen (338) in Period 3, and Billings (8521) in Period 4, respectively. The average eigenvector centrality for the network equalled 0.001 (SD = 0.003). This value decreased over time with 0.033 (SD = 0.088) in Period 1, 0.008 (SD = 0.026) in Period 2, 0.002 (SD = 0.011) in Period 3, and 0.001 (SD = 0.004) in Period 4. The researchers generated collaborations, but some collaborated more frequently with other beneficial researchers. Billings (0.035) again assumed the top position overall, followed by Clavio (0.032), Pedersen (0.031), and Vincent (0.023). Similar shifts also occurred across the periods. Theberge (0.309) and Trujillo (0.309) held the top positions in Period 1, followed by Eastman (0.134) in Period 2, Pedersen (0.133) in Period 3, and Clavio (0.048) in Period 4, respectively.

Please cite this article in press as: Hambrick, M.E., Sport communication research: A social network analysis. Sport Management Review (2016), http://dx.doi.org/10.1016/j.smr.2016.08.002

G Model

[(Fig._3)TD$IG]

SMR-382; No. of Pages 14 M.E. Hambrick / Sport Management Review xxx (2016) xxx–xxx

Fig. 3. Sociogram of the sport communication research network, Period 2 1990–1999.

[(Fig._4)TD$IG]

Fig. 4. Sociogram of the sport communication research network, Period 3 2000–2009.

Please cite this article in press as: Hambrick, M.E., Sport communication research: A social network analysis. Sport Management Review (2016), http://dx.doi.org/10.1016/j.smr.2016.08.002

9

G Model

SMR-382; No. of Pages 14

[(Fig._5)TD$IG]

M.E. Hambrick / Sport Management Review xxx (2016) xxx–xxx

10

Fig. 5. Sociogram of the sport communication research network, Period 4 2010–June 2015.

Table 3 Top network members by centrality type, January 1980–June 2015. Researcher

Degree centrality

Researcher

Betweenness centrality

Researcher

Eigenvector centrality

Billings, AC Pedersen, PM Hardin, MC Clavio, GE Vincent, J Geurin, AN Kian, EM O’Reilly, NJ Sanderson, J McGannon, KR

50 36 35 32 25 22 21 21 19 18

Billings, AC Hardin, MC Kian, EM Vincent, J Rowe, D Hutchins, B Clavio, GE Pedersen, PM Geurin, AN Falcous, M

33,108 17,052 11,872 11,473 11,418 9830 9793 8687 7112 7066

Billings, AC Clavio, GE Pedersen, PM Vincent, J Burch, LM Frederick, EL Geurin, AN Kian, EM Zimmerman, MH Brown, NA

0.035 0.032 0.031 0.023 0.018 0.018 0.018 0.017 0.016 0.016

Previous research has examined the presence of smaller groups of influential researchers connected by their geographic locations, institutional affiliations, and research areas (Acedo et al., 2006; Racherla & Hu, 2010; Ye et al., 2013). In conjunction with the centrality data, SNA software provides the means to explore the network’s centre and its most productive members. This can be accomplished by examining k-cores, or groups of members who most closely link to and share multiple relationships with one another (Borgatti et al., 2013; Scott, 2013). Love and Andrew (2012) documented a 4-core in their study of sport management and sport sociology researchers. This smaller group contained 100 members who shared at least four relationships with one another and represented the most influential scholars within the two fields.

Please cite this article in press as: Hambrick, M.E., Sport communication research: A social network analysis. Sport Management Review (2016), http://dx.doi.org/10.1016/j.smr.2016.08.002

G Model

SMR-382; No. of Pages 14

[(Fig._6)TD$IG]

M.E. Hambrick / Sport Management Review xxx (2016) xxx–xxx

11

Fig. 6. Sociogram of key researchers and their institutional affiliations. (a) Rows: Researcher name, current institutional affiliation, doctoral degree.

In the sport communication network, a 13-core emerged, which contained 27 closely connected researchers (Fig. 6). The members collaborated with others in this smaller group to produce either 13 or more co-authored publications, solo authorships, or a combination of the two. They represented 2.15% of the network and shared 341 relationships, or 13.44% of the network’s relationships. The density of this group equalled 48.60%, which contrasted with the overall research network’s density of 0.19%. This higher value indicated a relatively connected group, and its members leveraged more of the network’s possible relationships. Many of these members shared institutional affiliations, where they were employed at the time of the study or received their doctoral degree. The three most frequently appearing universities within this smaller group were Florida State University, Indiana University, and the University of Alabama. The 27 researchers benefitted from relationships with current and former colleagues as well as mentorships with doctoral advisors and other proximate faculty members. 5. Discussion This study explored the sport communication field’s social network, using SNA to document the field’s research productivity. The findings revealed considerable growth over the 36 years, as more researchers, publications, and collaborations emerged. The theoretical and practical implications of the results are discussed below. 5.1. Theoretical implications The findings link to previous work using SNA to assess the evolution of research fields (Endersby, 1996; Martins et al., 2012; Moody, 2004). Within sport communication, growth occurred with the number of researchers, publications, collaborations, and researchers collaborating on a single publication. In contrast to these increases in collaborations, the field experienced a decrease in the rate of solo authorships. More researchers opted to work together, and the number of researchers on a single publication increased. The results echo previous findings using SNA to document research fields (Racherla & Hu, 2010; Ye et al., 2013). The findings also highlight previous research indicating the presence of notable scholars within a field (Baraba´si & Albert, 1999; Crane, 1969; Crawford, 1971; Love & Andrew, 2012; Quatman & Chelladurai, 2008). A smaller, closely-knit group emerged within the social network. This group of 27 researchers had a higher density value, and shared institutional affiliations at a select number of universities. The individuals collaborated frequently amongst themselves and with other researchers. They obtained research advantages by working with colleagues, doctoral advisors/advisees and faculty

Please cite this article in press as: Hambrick, M.E., Sport communication research: A social network analysis. Sport Management Review (2016), http://dx.doi.org/10.1016/j.smr.2016.08.002

G Model

SMR-382; No. of Pages 14 12

M.E. Hambrick / Sport Management Review xxx (2016) xxx–xxx

members, and other doctoral students. This network characteristic follows the findings of previous research, which also documented research collaborations based upon institutional affiliations and geographic locations (Acedo et al., 2006; Love & Andrew, 2012; Quatman & Chelladurai, 2008; Racherla & Hu, 2010; Ye et al., 2013). Influential researchers included Billings, Hardin, and Pedersen, who assumed key positions and connected others through their collaborative activities. These scholars assumed pertinent roles, collaborated with one another, and partnered with other researchers to bolster their individual levels of productivity and advance the field as a whole. Similar findings have been documented in previous research (Love & Andrew, 2012). Such individuals are frequently characterised by having multiple connections, a range of research collaborations, and significant productivity (Acedo et al., 2006; Katz & Martin, 1997; Moody, 2004; Racherla & Hu, 2010). They help to illustrate the Matthew effect (Merton, 1968) as well as the scale-free and preferential attachment attributes of larger networks (Baraba´si & Albert, 1999). Others likely gravitated towards these sport communication scholars in efforts to learn more about the research process, collaborate, and increase their own productivity. Changes did occur among these connected researchers as they passed the torch of scholarship from one generation to the next (Endersby, 1996; Katz & Martin, 1997; Quatman & Chelladurai, 2008). These shifts coincided with increases in research activity overall, suggesting knowledge transfer and continuity. Preferential attachment existed (Baraba´si & Albert, 1999), as the leading researchers benefited from their ongoing collaborative efforts. Yet, the influx of others who published only one or two times also indicated an open environment where numerous contributors could produce and disseminate research. The commonality of findings in the current study and previous research indicates that scholars can make some basic assumptions about these fields, leading to a framework for how fields emerge and evolve. In the earliest stages, a small number of researchers initiate studies in a particular area. They first may work in isolation, but when the number of studies grows larger, they begin to work together. Their collaborations continue to increase, and this shared approach and division of labour help to increase productivity within the field. Their published research leads to the generation of additional works and attracts researchers with similar interests. The researchers also may have doctoral students or other research prote´ge´s to whom they transfer knowledge and ideas, leading to new generations of researchers. Research productivity begins to occur, and the number of researchers, publications, and collaborations increases over time (Acedo et al., 2006; Crane, 1969; Love & Andrew, 2012; Quatman & Chelladurai, 2008). This emphasis on growth suggests these fields continue to expand as they benefit from circumstances outlined in the previous research: greater divisions of labour and costs, better communication, more expertise among scholars, and mentoring of young researchers (Endersby, 1996; Katz & Martin, 1997; Laband & Tollison, 2000). Additionally, the research network in this study revealed the prominence of certain academic institutions. These universities and their academic programs could continue developing and disseminating research. Likewise, researchers at other universities could partner with these individuals in order to extend the group’s position. They also could strengthen the relationships at their own or other institutions, possibly prompting the creation of new network groups. Trends in research focal areas suggest further prospects for expansion. For example, studies on gender remain an important area within sport communication research. Fink (2015) asserted more research should focus on the limited and problematic representations of female athletes and women’s sports. Researchers could continue this exploration, building upon the previous works. Investigating other popular research areas such as sport consumption and marketing, which offer additional connections to sport management research, could provide more opportunities. Underrepresented areas identified in this study included sexuality, ability, and organisational communication. Researchers could produce studies investigating these topics in order to spur more sport communication research contributions within the broader sport management scholarly activity. 5.2. Practical implications Sport communication journals—Communication & Sport, International Journal of Sport Communication, and Journal of Sports Media—have created a home for sport communication research, where researchers can submit their work on a variety of topics within the field. An overlap between sport communication and sport management research exists, as evidenced by growing emphasis on sport consumption within sport communication research. The findings suggest an opportunity for more sport communication studies to inform sport management research and the sport industry. They also highlight the need for continued research in efforts to expand the field’s development. Beyond the three aforementioned journals, communication and sport management journals have embraced this research, facilitating additional dissemination of the research and expansion of the field overall. Popular communication outlets included Mass Communication & Society (39 sport communication studies published during the examined time period), Journal of Broadcasting & Electronic Media (34), and Journalism & Mass Communication Quarterly (25). Sport journals such as the Journal of Sport & Social Issues (122), International Review for the Sociology of Sport (66), and Sport in Society (71) featured a larger selection of articles. Sport communication studies also were published in journals specific to sport management such as the International Journal of Sport Management and Marketing (19), Journal of Sport Management (19), and Sport Management Review (15). Researchers can continue targeting a variety of sport management and communication journals to publish their works. This diffusion of scholarship may attract additional researchers and continue the field’s growth. The strong presence of notable scholars advanced productivity through individual and collaborative efforts. These individuals opened the field to others and increased the production of sport communication research. The number of

Please cite this article in press as: Hambrick, M.E., Sport communication research: A social network analysis. Sport Management Review (2016), http://dx.doi.org/10.1016/j.smr.2016.08.002

G Model

SMR-382; No. of Pages 14 M.E. Hambrick / Sport Management Review xxx (2016) xxx–xxx

13

researchers grew from 30 in the first period to 843 in the fourth period. More researchers worked together, and the number of studies increased. Continuity was also evidenced, where senior scholars held important positions but connected with others. Prolific researchers can continue to collaborate as well as mentor their academic prote´ge´s and other researchers. In turn, younger scholars can develop new relationships to pursue their own lines of research as they establish a presence and make contributions. Those interested in collaborative work should review the current publications, researchers, and focal areas to identify potential research partners and salient topics in order to continue their efforts. Fostering this growth and encouraging the submission of substantive research building upon previous work will help to advance the field. Previous studies have noted the potential gatekeeping role of influential scholars, who can dictate what gets published and when (Moody, 2004; Newman, 2001; Quatman & Chelladurai, 2008). Encouraging the entry of new sport communication researchers as well as researchers from other fields will help to populate this research and integrate different perspectives, research approaches, and ideas. This can occur when researchers make an active effort to expand their research collaborations to include a variety of current and new research partners. 6. Limitations and future research The study has several limitations. First, the study focused only on peer-reviewed articles in scholarly journals as opposed to other collaborative outcomes (e.g., conference proceedings, textbooks) and may have limited the number of members included in the sport communication research network. Data incorporating other research outcomes may have presented a different network. Additionally, data on prominent researchers in the network revealed connections based upon institutional affiliations. A greater exploration of these potential connections and other relationship types may have resulted in a different network. Next, the use of SNA placed greater emphasis on research collaborations. A more balanced weighting on authorship types may have created an alternative research network. Finally, the study examined sport communication research in isolation. Including other research may have resulted in a different research network. Future studies could investigate other research outcomes and relationships. Studies also could examine what factors contribute to collaboration, whether formal or informal. This data could be captured through surveys, semi-structured interviews, and/or focus groups with researchers to investigate their sentiments regarding research collaborations (Tribe, 2010). This research could explore what processes researchers used to create collaborations. Studies also could document how collaborations were created and what factors proved important in the process. The information could be used to understand research collaboration patterns and habits as well as what variables facilitate or hinder these activities. Similarly, future studies could investigate why some individuals pursue solo authorships versus working with others and what promotes or limits this type of research. Finally, an opportunity exists to expand the data collection and investigate relationships between sport communication and sport management research or between sport communication and the broader communication field. These efforts would help in documenting how sport communication researchers fit within other contexts. 7. Conclusion The current study examined the sport communication field over 36 years from January 1980 to June 2015. SNA helped map the sport communication researchers and their research activity, operationalised through academic publications. The results revealed a total of 1255 researchers within the field. Their solo authorship and collaboration activities yielded 2537 relationships and 1283 publications. The number of researchers, publications, collaborations, and researchers per publication increased over time. Shifts in research areas also occurred. Studies focused on gender, mass media, and sport consumption proved most popular, and were published in sport communication, sport management, and communication journals. A smaller group of individuals emerged as key contributors to the field within each respective period. These researchers were characterised by their collaborations, institutional affiliations, and productivity. They collaborated with a variety of other researchers, who also worked with one another and contributed to the network’s overall growth. The findings mirrored previous studies examining research networks, suggesting academic research networks operate within a similar fashion despite potential differences in research areas and foci[2_TD$IF]. Acknowledgements The author would like to thank Dr. Alison Doherty, the two anonymous reviewers, and University of Louisville graduate assistant Alicia Cintron for their thorough readings and detailed comments and suggestions, which helped to substantially improve this manuscript. References Abeza, G., O’Reilly, N., & Nadeau, J. (2014). Sport communication: A multidimensional assessment of the field’s development. International Journal of Sport Communication, 7, 289–316. Abeza, G., O’Reilly, N., Seguin, B., & Nzindukiyimana, O. (2015). Social media scholarship in sport management research: A critical review. Journal of Sport Management. http://dx.doi.org/10.1123/jsm.2014-0296

Please cite this article in press as: Hambrick, M.E., Sport communication research: A social network analysis. Sport Management Review (2016), http://dx.doi.org/10.1016/j.smr.2016.08.002

G Model

SMR-382; No. of Pages 14 14

M.E. Hambrick / Sport Management Review xxx (2016) xxx–xxx

Acedo, F. J., Barroso, C., Casanueva, C., & Gala´n, J. L. (2006). Co-authorship in management and organisational studies: An empirical and network analysis. Journal of Management Studies, 43, 957–983. Baraba´si, A. L., & Albert, R. (1999). Emergence of scaling in random networks. Science, 286, 509–512. Baraba´si, A. L., Jeong, H., Ne´da, Z., Ravasz, E., Schubert, A., & Vicsek, T. (2002). Evolution of the social networks of scientific collaborations. Physica A, 311, 590–614. Borgatti, S. P., Everett, M. G., & Johnson, J. C. (2013). Analysing social networks. Thousand Oaks, CA: Sage. COSMA (2016). Accreditation principles and self-study preparation May 2016. Retrieved from http://www.cosmaweb.org/accreditation-manuals.html Crane, D. (1969). Social structure in a group of scientists: A test of the ‘‘invisible college’’ hypothesis. American Sociological Review, 34, 335–352. Crawford, S. (1971). Information communication among scientists in sleep research. Journal of the American Society for Information Science, 22, 301–310. Endersby, J. W. (1996). Collaborative research in the social sciences: Multiple authorship and publication credit. Social Science Quarterly, 77, 375–392. Filo, K., Lock, D., & Karg, A. (2015). Sport and social media research: A review. Sport Management Review, 18, 166–181. Fink, J. S. (2015). Female athletes, women’s sport, and the sport media complex: Have we really ‘‘come a long way, baby’’? Sport Management Review, 18, 331–342. Hansen, D., Shneiderman, B., & Smith, M. A. (2011). Analysing social media networks with NodeXL: Insights from a connected world. Burlington, MA: Morgan Kaufmann. Katz, J. S., & Martin, B. R. (1997). What is research collaboration? Research Policy, 26, 1–18. Laband, D. N., & Tollison, R. D. (2000). Intellectual collaboration. Journal of Political Economy, 108, 632–662. Love, A., & Andrew, D. P. S. (2012). The intersection of sport management and sociology of sport research: A social network perspective. Sport Management Review, 15, 244–256. Lusher, D., Robins, G., & Kremer, P. (2010). The application of social network analysis to team sports. Measurement in Physical Education and Exercise Science, 14(4), 211–224. Martins, M. E., Martins, G. S., Csillag, J. M., & Pereira, S. C. F. (2012). Service’s scientific community: A social network analysis (1995–2010). Journal of Service Management, 23, 455–469. Merton, R. K. (1968). The Matthew effect in science. Science, 159, 56–63. Moody, J. (2004). The structure of a social science collaboration network: Disciplinary cohesion from 1963 to 1999. American Sociological Review, 69, 213–238. Newman, M. E. (2001). The structure of scientific collaboration networks. Proceedings of the National Academy of Sciences of the United States of America, 98, 404–409. Pedersen, P. M. (2013). Reflections on communication and sport: On strategic communication and management. Communication & Sport, 1, 55–67. Pedersen, P. M. (2015). Routledge handbook of sport communication. London: Routledge. Pedersen, P. M., Miloch, K. S., & Laucella, P. C. (2007). Strategic sport communication. Champaign, IL: Human Kinetics. Price, D. J. D. (1965). The scientific foundations of science policy. Nature, 233–238. Quatman, C., & Chelladurai, P. (2008). The social construction of knowledge in the field of sport management: A social network perspective. Journal of Sport Management, 22, 651–676. Racherla, P., & Hu, C. (2010). A social network perspective of tourism research collaborations. Annals of Tourism Research, 27, 1012–1034. Scott, J. (2013). Social network analysis (3rd ed.). Thousand Oaks, CA: Sage. Stoldt, G. C., Dittmore, S. W., & Pedersen, P. M. (2014). Communication in the sport industry. In P. M. Pedersen & L. Thibault (Eds.), Contemporary sport management (5th ed.). Champaign, IL: Human Kinetics. Tremayne, M. (2014). Anatomy of protest in the digital era: A network analysis of Twitter and Occupy Wall Street. Social Movement Studies: Journal of Social Cultural, and Political Protest, 13, 110–126. Tribe, J. (2010). Tribes, territories, and networks in the tourism academy. Annals of Tourism Research, 37, 7–33. Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and applications. Cambridge: Cambridge University Press. Ye, Q., Li, T., & Law, R. (2013). A coauthorship network analysis of tourism and hospitality research collaboration. Journal of Hospitality & Tourism Research, 37, 51–76.

Please cite this article in press as: Hambrick, M.E., Sport communication research: A social network analysis. Sport Management Review (2016), http://dx.doi.org/10.1016/j.smr.2016.08.002