Contact Preference on Social Media: A Novel Approach to Rank Network Positions
Hwai-Chung Ho | Academia Sinica | Taiwan
Key words: network position, Facebook interaction, contact preference index.
The surge of social media over the last decade has given rise to alternative and additional occasions for extensive social interactions in everyday life. On popular social networking sites, in particular, users can interact beyond the constraints of time and space, and have more chances to express their preferences about whom to contact more frequently and intensely. Most previous studies about social media have based their investigations on the number of links in the online social network of interest. Taking advantage of the rich information embedded at both link and contact levels of social interactions on social media, the present study instead examines an ego’s contact intensities and preferences associated with her alters. Our approach is facilitated by the availability of representative survey data and contact records retrieved from the Facebook accounts of 959 senior college students in 26 classes from 22 Taiwanese universities. This particular cohort of students entered college in the autumn of year 2012, and we analyze the online records from June 19, 2013 to December 22, 2015. To better capture the students’ dynamic interactions within the class, we develop a measure to quantify the extent to which students in the same class prefer to contact a particular individual over others. We analyze those classmates who had at least one exchange with the ego on Facebook during the period, through texting or tagging, with a total of 181,544 contacts. The value of the measure represents how popular an ego is among the Facebook friends in her class, thus incorporating the asymmetric nature of interactions between every pair of Facebook friends. We further standardize the measure by the number of active Facebook users in the ego’s class, so that its value ranges from zero to one, which we call a contact preference index (CPI). Such standardization allows us to aggregate the CPIs of all students from their respective class networks to address some important issues about interactions on social media. Preliminary analyses show that the distribution of the CPI value follows the power law, which indicates clear effects of preferential attachment in Facebook interactions. The CPI value also tends to associate closely with certain aspects of ego’s interaction patterns. For example, an ego tends to score a higher CPI value if she interacts with more alters who are popular themselves. We expect that the CPI may help explain network effects on substantive research issues based on survey data.
Qualitative and quantitive analysis of Twitter activity: a network perspective
Lucia Falzon | DST Group & the University of Melbourne | Australia
Twitter represents an idealised platform for individuals to rapidly and widely share information with others. The ease with which data can be sourced through Twitter’s API has resulted in a disproportionate number of ‘big data’ studies aimed at assessing what information is most widely propagated through the ‘Twittersphere’ network and why. The focus on digital data has led researchers to overlook the importance of ‘small data’: traditional methods of data collection, such as survey and experimental studies. Although big data can measure the impact of Twitter messaging in terms of observable retweet/mention/reply behaviour, the richness at the level of the individual could be lost if big data approaches are exclusively used, and therefore lead to inappropriate generalisations and conclusions. In response to this shortcoming, we take a different approach: we synthesise survey and social network methodology and use these two forms of data to explore Twitter behaviour.
A comprehensive survey was designed on the basis of a model of Twitter information sharing informed by prior literature. 173 active and registered Twitter users answered questions relating to their: (1) motivations for information sharing, (2) their imagined audience, (3) strategies used for facilitating information propagation, (3) choice of Tweet content, and (4) motivations for following others. The survey also assessed the role of message reception on the users’ attitudes and consequent information spreading. Participants provided their Twitter handle as part of the survey allowing us to collect data from the Twitter API to construct ego-centred networks of users and their followers; discussion networks; and networks of information flows. For each class of network, we propose analysis questions that focus on the spread and reach of Twitter users as information spreaders and the network positional attributes of users. Preliminary analysis of qualitative data in conjunction with network analysis results provides new insight on Twitter user behaviour.
Tweeting louder - becoming bigger?
Jasmin Fitzpatrick | JGU Mainz | Germany
This article examines the agenda-setting potential of small and large CSOs and political parties using a weighted, additive index. The comparison of political parties and civil society organizations (CSOs) does not have a long tradition. However, comparing these two forms of political organizations opens up a promising array of possibilities for a better understanding of their networks. Both forms of organizations have professionalized their strategy to communicate goals and set the agenda. Agenda-setting has been analyzed by scholars of political communication since the 1970s (McCombs and Shaw 1972; McCombs 2014). The key idea is to influence what topics people think about. The internet has changed this game, because organizations reach a large audience directly without media agencies filtering the content (Bekkers et al. 2011; Besiou et al. 2013; Becker 2014). Therefore, for both, political parties and CSOs, the internet is a major platform for creating a network with the public, mobilizing support and obtaining goals. This opportunity seems especially interesting for small CSOs with strict budget limitations, because they are now capable of communicating with a large number of people at little cost (Bennett 2004; Matschke et al. 2012; Neuman et al. 2014). Essentially, this mirrors the debate about the digital divide (e.g. Norris 2001; DiMaggio et al. 2001; Wilson et al. 2003) on the meso-level. For scholars of CSOs and political parties the new setting reveals a particularly interesting ground for comparative research. Studies have so far focused on the adoption of social media by CSOs (Eimhjellen et al. 2014), the content of tweets (Waters and Jamal 2011) or how to engage stakeholders (Lovejoy et al. 2012). The two leading research questions in this article are: (a) do political parties and CSOs differ in their agenda-setting-potential and (b) do small organizations differ from their larger competitors in their agenda-setting-potential?
Both systems, in Austria and in Germany, have recently undergone change that resulted in the successful establishment of small parties. We also find a wide array of small and large CSOs as part of a vivid civil society. This makes Austria and Germany very interesting test cases. In both countries, Twitter is widely used by professionals and politicians. Therefore, Twitter is a perfect tool for professional communication and agenda setting for CSOs and political parties. The results indicate that especially small CSOs struggle to adopt Twitter as a communication tool, while small parties find it more easy to employ the microblogging service. If they manage to adopt Twitter, small organizations can push their topics with similar success on the agenda.
This article contributes to the debate about the impact of networks and social media on political communication and provides a new measurement for agenda-setting. The findings are relevant to scholarly debate and to practitioners. The agenda-setting-index was developed as part of the PhD-project of the author (forthcoming) especially for CSOs, yet, it is employable to other political actors in other contexts as well.