Bodong Chen

Crisscross Landscapes

Notes: Malinick, T. E., Tindall, D. B., & Diani, M. (2013). Network centrality and social movement media coverage



Citekey: @Malinick2013-ub

Malinick, T. E., Tindall, D. B., & Diani, M. (2013). Network centrality and social movement media coverage: A two-mode network analytic approach. Social Networks, 35(2), 148–158.





abstract This article examines the relationship between structural location (namely, degree centrality) and news media coverage. Our central hypothesis is that the network centrality of social movement actors is positively associated with the prevalence of actors being cited in the print news media. This paper uses two-mode data from a communication network of environmentalists in British Columbia, and examines the relationship between their structural location and the frequency by which they are cited in newsprint media with regard to particular frames (about forest conservation, environmental protest, and related issues). We asked a sample of social movement participants about their ties to a target list of relatively high profile actors (environmental activists). We turned the resulting network matrix into a bipartite graph that examined the relationships amongst the target actors vis a vis the respondents. Next we calculated point in-degree for the target actors. For the target actors we also have data from a representative sample of 957 print news articles about forestry and conservation of old growth forests in British Columbia. We compare the effects of network centrality of the target actor versus several attributes of the target actors (gender, level of radicalism, leadership status) on the amount of media coverage that each of the target actors receives. We find that network centrality is associated with media coverage controlling for actor attributes. We discuss theoretical implications of this research. Finally, we also discuss the methodological pros and cons of using a “target name roster” to construct two-mode data on social movement activists. (p. 148)

The relationship between social movements and media has long been viewed as a complex and dynamic interaction. (p. 148)

In this paper we contribute to this discussion from a distinctive angle, asking whether media coverage of social movements is affected by the properties of the activists mobilizing within them. (p. 148)

On the other hand, we introduce an explicit relational dimension to analysis, by exploring the correlation between media coverage and the position that activists occupy within movement networks. Building on earlier contributions (Diani, 2003), we ask in particular whether media attention to specific activists is positively related to their centrality in movement networks. (p. 148)

2.2. Attributes versus relational properties (p. 149)

Four specific source characteristics are considered in this study. Three of them refer to standard individual attributes, namely: (1) whether or not the source is radical; (2) the source’s gender; and (3) whether or not the source holds a formal leadership position in some movement organization. The fourth property is, in contrast, a relational property, as it consists of the source’s centrality in the movement’s communication network. (p. 149)

  1. Literature (p. 149)

2.1. Mediaandsocialmovements’impact (p. 149)

2.3. Radicalism (p. 149)

some believe that media selects content based on the dual objectives of maximizing profit and market share (Croteau and Hoynes, 2003; McAdam et al., 1996), and most typically these objectives are met by covering issues in a contentious, exciting, or sensationalized way. (p. 149)

A major problem is, of course, what factors may affect the relationship between media and movements, and more specifically, the former’s attention to the latter’s stances and actions. (p. 149)

Some have pointed at the role of frames (Goffman, 1974), namely, interpretive packages that condense and summarize the facts of the world ‘out there’ in a way that eases our understanding of them (Benford, 1993a,b; Gamson, 1992; Snow et al., 1986; Snow and Benford, 1988). Collective action frames may be intentionally and strategically formulated by social movements – and analogously by their opponents – to garner general public support for their cause and to attract and mobilize potential constituents (Gamson and Modigliani, 1989; Noy, 2009; Scheufele, 1999; Snow et al., 1986; Snow and Benford, 1988). Frames are also simultaneously used by media to simplify presentation of complicated issues in the course of reporting the news (Anderson, 1997; Dispensa and Brulle, 2003; Gamson et al., 1992; Ryan, 1991; Scheufele, 1999; Tuchman, 1978). (p. 149)

Others, however, suggest that radical views and behaviours are too far outside the norm and actually serve to confuse matters (Fitzgerald and Rodgers, 2000; Rootes, 2007; Snow and Benford, 1992). (p. 149)

Other observers have stressed the asymmetrical dependency between social movements and media, with the former much more dependent on the latter than the other way around (Entman and Rojecki, 1993; Gitlin, 1980; McCarthy et al., 1996; Baylor, 1996; Klandermans and Goslinga, 1996; Gamson and Modigliani, 1989; Carroll and Hackett, 2006). (p. 149)

2.4. Gender (p. 149)

Carolyn Merchant (1997) estimates that globally, roughly 80 percent of grassroots environmental activists are female. (p. 149)

However, Merchant (1997) also notes that a distinct minority of these women are actually formal leaders. (p. 149)

2.7. Analyticstrategyandhypotheses (p. 150)

First, correlational analyses are utilized to explore particular bivariate relationships. Second, multiple regression analyses are used to explore hypotheses regarding the relationships between three of the independent variables (social network centrality, leadership, gender and radicalism) and the dependent variable. Finally, interviews with print-news media-workers who were key players in covering the British Columbia forestry conflict are used to supplement findings, adding further insight to the interpretation of the results. (p. 150)

2.5. Formalleadershiproles (p. 150)

leaders strive to garner general public support for the movement’s cause and promote mobilization of constituents (Snow and Benford, 1988). (p. 150)

H1. Formal leaders will have higher communication-network centrality scores than non-leaders. (p. 150)

2.6. Socialnetworkcentrality (p. 150)

H2. Formal leadership will be positively associated with media coverage. (p. 150)

‘Leaders’ are defined in terms of factors such as influence, authority, and status, rather than formal roles (Diani, 1995, 2003; Nepstad and Bob, 2006). (p. 150)

H3. Network centrality will be positively associated with media coverage. (p. 150)

As Freeman (1979:219) states: “With respect to communication, a point [actor] with relatively high degree [centrality] is somehow ‘in the thick of things’.” (p. 150)

H4. Males will be more likely to be leaders than females (p. 150)

Actors’ centrality in a movement’s communication network would then grant them control over the flow and transmission of information. (p. 150)

H5. Women will have relatively higher centrality scores than males. (p. 151)

  1. Thedata (p. 151)

In this paper we combine evidence from two initially unrelated data sets. The first consists of relational (and some attribute) data for 34 individuals active in the environmental movement in the Vancouver and Victoria, British Columbia regions in the late 1980s and early 1990s.3 Their media coverage is measured by drawing upon a database containing records of print-news media citations relating to the forestry conflict in western British Columbia between 1986 and 1992. In addition, qualitative data from faceto-face interviews held with key news-workers operating in the field during the 1980s and 1990s are also used to supplement the findings and discussion. (p. 151)

H6. Radicals will be less covered than moderates. (p. 151)

  1. Thestudy (p. 151)

3.1. Theforestryconflictandtheenvironmentalmovementin British Columbia (p. 151)

4.1. Socialnetworkdata (p. 151)

The social network data were collected for prior studies of the underlying structure and functioning of environmental movement groups active in the British Columbia forestry conflict (see Tindall, 1994, 2002; Tindall et al., 2003). (p. 151)

a survey was conducted of 28 activists who were part of the environmental movement. Amongst other questions, they were asked to indicate whether they had ties to a predetermined list of individuals active in the forestry conflict. (p. 151)

We refer to the people who were interviewed as the “respondents”. (p. 151)

The questionnaire allowed the 28 respondents (egos) to indicate ties to a wide range of actors (targets) who participated in the forestry debates, including members of environmental groups, individual activists, representatives of business, politicians, union leaders, and First Nation representatives. (p. 151)

The resulting dataset consisted of a two-mode (28 egos×34 targets), single-relation sociomatrix, where each cell xij equalled 1 if ego i had a tie to target j; otherwise the cell value was 0. (p. 152)

their seminal contributions: diagnostic frames, involving not only identifying the problem, but also attribution of cause, blame, or culpability for the situation to someone or some group; prognostic frames, consisting of statements meant to explore and define possible solutions to the problem and strategies meant to resolve them; and motivational frames, consisting of what Benford and Snow (2000:617) quite simply described as a moral “call to arms.” (p. 152)

4.2. Citationdata (p. 152)

As an indicator of media coverage we adopted the number of times that the 34 environmental movement activists (‘targets’) we just mentioned were cited in the press. (p. 152)

each of the 19 keywords was later classified as illustrations of one of the core framing tasks identified by Snow and Benford (1988) in (p. 152)

4.3. Datamanagementanddatafilecreation (p. 153)

The 957 news articles contained in the citation database account for a total of 22,004 total citations. (p. 153)

210 of these citations were from environmental social movement organization members that were listed as targets in the social network instrument (p. 153)

  1. Results (p. 153)

from Table 3 we see that the leadership measure though somewhat sizable (r = .29) is not significantly associated with the centrality score, thus failing to support H1. (p. 153)

All of this information was compiled in a single SPSS data file (p. 153)

with regard to H2, we see from Table 3 that the formal leadership indicator is not statistically significantly related to citations (p. 153)

From Table 3 we also see support for H3 when we note that the network centrality score is very significantly, positively associated with citations. Thus, those who were more central to the movement’s communication network actually got more attention from the press (p. 153)

the correlation between gender and formal leadership is not significant (p. 154)

gender is not associated with social network centrality, thus failing to support H5 (p. 154)

the relationship between radicalism and gender is significant, but with a negative correlation, suggesting that, all else being equal, women were more likely to be radical than men. (p. 154)

Table 4 shows the results for an individual quasi-Poisson generalized linear model,13 where media coverage (number of total citations) is evaluated as function of network centrality, plus the four actor attributes. (p. 154)

So far, we found that formal leaders were not the most central actors in the movement’s communication network, nor were they significantly associated with rates of citation, while the most central actors were actually cited the most. (p. 154)

Thus, there is evidence to suggest that leadership and communication network centrality in the BC wilderness preservation movement are indeed different phenomena.12 (p. 154)

the fact that activists who were more central in the movement communication network were also more cited in the press is consistent with studies like Diani’s (2003). (p. 155)

On the other hand, the fact that moderates get stronger coverage than radicals might be at least partially at odds with the fact that highly contentious, possibly violent protest events have much better chances of being reported than more moderate ones (Koopmans and Rucht, 2002, 246–251). (p. 155)

Reporters indicate that covering drama admittedly has some initial appeal, and is often necessary given the scale and scope of certain protest events such as the events surrounding Clayoquot Sound. More important, however, these reporters also state that when covering ongoing conflicts, at some point a shift must be made towards reporting on the underlying issues. (p. 155)

In the present study we adopt an approach that overcomes a number of these challenges. Because the target individuals on the network instrument were identified through public sources, this avoids the IRB problem (p. 155)

  1. Discussionandconclusion (p. 155)