Review of "Analysing Social Media Data and Web Networks"

Cover of "Analysing Social Media Data and Web Networks"

 

I was asked by the editors of Information Polity to review Analysing Social Media Data and Web Networks, an edited volume that deals with primarily methodological issues of online research. Below, you will find the preprint version – this is also available over at Academia.edu as well as on ResearchGate.

— — —

Review of Analysing Social Media Data and Web Networks
– Cantijoch, Gibson and Ward (eds.)

Anders Olof Larsson
Department of Media and Communication
University of Oslo

The digital realm offers a multitude of opportunities for research. However, given the ever-changing nature of online environments, research focused on assessing such “moving targets” (McMillan, 2000) need to “freeze the flow” (Karlsson and Strömbäck, 2010) or make the data deluge available online suitable for scientific analysis in some other way. The volume at hand, Analysing Social Media Data and Web Networks is edited by Marta Cantijoch, Rachel Gibson and Stephen Ward and offers a series of useful and often practical insights for those of us who take special interest in analysis of online media. Specifically, the book features ten chapters that all provide insights into (primarily) methodological issues, presented by some of the most well known authors in what could perhaps be described as the field of online political communication (and beyond). In this review of the book, I have arranged my comments around five main issues that permeate throughout the text. In so doing, I’ll provide examples from individual chapters featured in the title, as well as from other sources. I’ve chosen to label the five issues dealt with as follows: The ever-changing nature of online services; Commercialization of data access; Socio-demographic perspectives; Ethical issues and Comparing with what is to come.

First, digital methods are fickle. They need to be fashioned so as to be able to adapt to and catch the aforementioned online flows. Indeed, researchers have dealt with what I like to call the ever-changing nature of online services for some twenty years, painstakingly learning from previous mistakes and developing more efficient ways of data gathering from online sources. Often, the tools used for such endeavors are constructed and maintained by individual scholars and their respective research groups, making it somewhat difficult for the community at large to judge the merits of any particular tool in comparison to some other variety. For example, while I am certain that the services introduced by Thelwall and Hussain et. al. in their chapters are of the utmost quality, the very fact that more and more purpose-built tools are launched could lead to difficulty in performing cumulative, comparative research as researchers select their tool of choice from an ever-increasing array of instruments. We should, of course, always strive to improve our tools, but the lament of the editors regarding the apparent lack of theoretical cohesion would appear to ring true also for these issues: “the field has deviated from [systematic theoretical inquiry] in a rather chaotic fashion, which makes cross-country and longitudinal comparison extremely difficult” (Cantijoch, Gibson, and Ward, 2014: 16-17). A similar statement could arguably made with regards to the methodological development of online research, broadly defined.

Second, such tools for collecting are made subject to almost constantly updated rules of the social media platforms they allow us to study. Such changes often appear to be related to what is understood in this review as ongoing processes of commercialization of data access. We can, for example, point to relatively recent delimitations of free access to a variety of public application programming interfaces (APIs) as hosted by Twitter (e.g. Burgess and Bruns, 2012), or the delimitations of functionalities imposed by Facebook on the freely available Netvizz data extraction service (Rieder, 2015). Indeed, issues like these are touched upon in the chapter penned by Jungerr and Jürgens, but it would have been nice if the authors or editors had touched upon what could be labeled as critical interpretations of these developments. With such a view in mind, the chapter by Graham and Wright correctly suggests that “people’s online data is often commercially valuable” (Cantijoch, et al., 2014: 204) – but what does such value entail for academic conduct? Arguably, the current developments are detrimental for scientists who, often with scarce funding, seek to perform research detailing services like these. As such, there is a clear risk that the increased commercialization of data access will contribute to a further widening of the already existing chasms between “data-rich” and “data-poor” scholars (e.g. Larsson, 2015).

My third point considers socio-demographic perspectives of the users whose digital trace data often end up in our work sheets, research notebooks and eventually (or hopefully, perhaps) published works. Specifically, regardless of how data are collected, we must assess who the producers of these data are – at least in some overarching, structural sense. Here, many of the included authors do a good job at acknowledging the biases that societal divisions like these unequivocally place on the data we gather from online sources. Increased knowledge about such stratifications should help end the sometimes heard happy-go-lucky type argument that data, because it is so plentiful (or even “big”, if you will), would be representative of the public opinion. Certain groups of citizens will always be overrepresented for certain forms of media use – a difficult obstacle to overcome for scholars, but an obstacle to be acknowledged clearly, nonetheless (e.g. Hargittai and Litt, 2012). Of course, such over- or underrepresented groups could be expected to vary across countries and contexts – something that further underlines the necessity of and challenges with comparative research across the strata of your choice.

Fourth, ethical issues are, or at least should be, at the very heart of scholarship. Such choices and prioritizations seemingly become especially poignant in the online context, where data emanating from a variety of user profiles and interactions can be collected and systematized with relative ease. The openness of online platforms like Twitter or YouTube is sometimes discussed as providing a carte blanche for various forms of data collection. Therefore, it is refreshing to see such arguably simplistic approaches to methodology questioned in Thelwall’s chapter, where it is suggested that “[d]espite this openness, there is of course a need for researchers to exercise discretion when personally identifying individuals in the course of their research” (Cantijoch, et al., 2014: 76). Related to such identification of individuals is the topic or theme dealt with in the tweets, Facebook posts or YouTube videos examined. While it might be technically true that “the majority of this data is open for all to examine” (Vargo, Guo, McCombs, and Shaw, 2014: 296), special consideration should be taken when the content deals with what could be understood as sensitive topics, such as sexual preferences or political orientation (Ess, 2013; Moe and Larsson, 2012). A recent overview by Zimmer and Proferes suggests that at least for research into Twitter, reflection on ethical issues are seldom seen (Zimmer and Proferes, 2014). This reviewer would be surprised if the situation was different for scholarship detailing other, similar services. One way to approach ethical issues has been to focus on content that has been actively put forward by users in such a way as to indicate their willingness to be seen in a specific thematic context. On Twitter, for example, this has been done by focusing on so-called hashtags – thematic keywords included by the users themselves to show thematic coherence. Such an approach is favored by a series of authors contributing to the volume at hand, like the previously mentioned chapter by Jungherr and Jürgens as well as the section penned by García-Albacete and Theocharis. Indeed, this way of approaching research could be seen as relatively unproblematic from an ethical point of view. However, the issue of what lies beyond the hashtag – in other words, what contents of relevance we are missing out on by delimiting our searches in this supposedly ethically sound way – remains unanswered.

My fifth point, comparing with what is to come, relates back to the first one. I mentioned at the beginning of this review that the methods discussed here could be seen as in constant flux, given the almost continuous changes taking place within the technical infrastructures we wish to study. For this final point, I’d like to stress the fact that not only do these infrastructures change – they will undoubtedly become out-of-date at some point, replaced by some new variety. Indeed, the services we study today will most likely not be around tomorrow, and it would have been fruitful to see the authors and editors reflect to a higher degree on such issues of cross-platform comparability in the volume. For example, how do we secure longitudinal insights, comparing possible future online platforms with those in fashion today if we construct our data collection tools and phrase our research questions based on the affordances of those services currently available?

Finally, while studies assessing the use of various social media platforms are all the rage, it is good to see that Analysing Social Media Data and Web Networks also features a series of chapters dealing with analyses of web sites, particularly those provided by Rosalund Southern and Benjamin N. Lee. Indeed, while it might be tempting to study comparably new services like social media platforms, the important role of web pages within political campaigning should be acknowledged with a suitable amount of attention from researchers. In conclusion, while the focus of the book is placed on issues primarily of concern to the broader field of political communication, such a thematic delimitation should not keep potential readers with mainly methodological interests at bay – the rich perspectives offered here are sure to be of use also to those coming to the study of online methods from some other disciplinary starting point.

References

Burgess, J., & Bruns, A. (2012). Twitter Archives and the Challenges of “Big Social Data” for Media and Communication Research. M/C Journal, 15(5).

Cantijoch, M., Gibson, R., & Ward, S. (2014). Analysing Social Media Data and Web Networks. London: Palgrave Macmillan. Ess, C. (2013). Digital Media Ethics (Second ed.). Cambridge: Polity Press.

Hargittai, E., & Litt, E. (2012). Becoming a Tweep. Information, Communication & Society, 15(5), 680-702.

Karlsson, M., & Strömbäck, J. (2010). FREEZING THE FLOW OF ONLINE NEWS — Exploring approaches to the study of the liquidity of online news. Journalism Studies, 11(1), 2 – 19.

Larsson, A. O. (2015). Studying Big Data – ethical and methodological considerations. In H. Fossheim & H. Ingierd (Eds.), Internet research ethics. Oslo: Cappelen Damm Akademisk.

McMillan, S. J. (2000). The Microscope and the Moving Target: The Challenge of Applying Content Analysis to the World Wide Web. Journalism & Mass Communication Quarterly, 77(1), 80-98.

Moe, H., & Larsson, A. O. (2012). Methodological and Ethical Challenges Associated with Large-scale Analyses of Online Political Communication. Nordicom Review, 33(1), 117-124.

Rieder, B. (2015). the end of Netvizz (?). Retrieved from http://thepoliticsofsystems.net/2015/01/the-end-of-netvizz/

Vargo, C. J., Guo, L., McCombs, M., & Shaw, D. L. (2014). Network Issue Agendas on Twitter During the 2012 U.S. Presidential Election. Journal of Communication, 64(2), 296-316.

Zimmer, M., & Proferes, N. (2014). A Topology of Twitter Research: Disciplines, Methods, and Ethics. Aslib Proceedings.

Internet Research Ethics

Internet Research Ethics - book cover

I am very happy to see the Internet Research Ethics books in print – as well as freely available in a variety of digital formats. I am even more happy to have contributed to this excellent volume (if I may say so myself), which details some of the current challenges with regards to – you guessed it – ethics that researchers interested in the Internet frequently have to deal with. In my chapter, I approach these issues based on my own experience with “big data”-type research, discussing methodological challenges in tandem with those of the ethical variety. To be a bit more specific, I argue for what could be labeled as a hashtag-based approach to data collection on Twitter. Briefly put, by focusing only on tweets that contain specific thematic keywords, we can be quite certain that the senders involved intended for their tweets to be visible in a certain context – of interest to researchers. Of course, while such an approach might be considered as ethically sound, it also means that any twitter activity of relevance not including the hashtag under study would not be included. This is of course problematic, especially in an international perspective, where ethical recommendations are often more relaxed – essentially meaning that researchers in many countries have broader opportunities for procuring full samples of social media traffic.

Editorial introduction in The Journal of Media Innovations

JMI
The latest issue of The Journal of Media Innovations is now live, hot off the digital press. This is the largest issue published by the Journal so far – with seven full papers, two research briefs and two book reviews, covering a wide array of issues all related to the overarching topic of innovative practices related to the media (industry), there’s bound to be something for interested researcher and practitioners here. Moreover, the issue features an editorial introduction penned by yours truly. Essentially, then, this introduction features summaries of the featured articles and briefs. I also briefly discuss the common themes and topics raised. As such, it could be a suitable starting point for those of you who might be looking for an overview of what’s going on in the current issue. Of course, chances are you’ll jump straight to one of the articles instead. Anyway, on behalf of myself and the rest of the editorial team at The Journal of Media Innovations, we hope you enjoy the featured articles. And hey, it’s free. Check it out, and consider to submit your own work!

… On a personal note, a big thank you to Charles Ess for allowing me to play a large part in the culling together of this issue. Also, thanks to Anders Fagerjord for putting together the nice article template. Hope you like the looks of the Journal as well!

Review of "Media, Society, World: Social Theory and Digital Media Practice" by Nick Couldry

I was asked to review Nick Couldry’s Media, Society, World: Social Theory and Digital Media Practice for European Journal of Communication. Below you will find the pre-print version (also available at academia.edu and ResearchGate).

— — —

The broader field of media and communication studies incorporates a vast array of theoretical perspectives that have been and still are applied to further our understanding of what the media are, how they function and what results (or even ‘effects’ – if there ever was a value-laden word within the field, this is it) our everyday interactions with and through media could be expected to have. Undergraduates are – or at least I was – brought up learning the early history of the field as an almost steady march from “All-powerful media”, understood theoretically through metaphors of ‘hypodermic needles’ or magic bullets’, to latter periods where such power was vested in the recipients of media content – power to the people, if you will. Theories like Uses and Gratifications, for example, suggested that the preferences and goal orientations of individual media consumers would lead them to specific media content, rather than consuming any content that came into their path (e.g. Blumler and Katz, 1974).

In his book Media, Society, World: Social Theory and Digital Media Practice, Nick Couldry embarks and expands on a somewhat similar approach. Specifically, Couldry makes the case for what he labels as a ‘practice approach’ to media and communication studies. He notes that while the aforementioned Uses and Gratifications perspective largely focused on individual, personal uses of specific media content in specific media channels, these diverse media channels or outlets have today become interlocked to the point where the only suitable starting point for an endeavor such as this would be ‘the media environment’, not specific media considered in isolation from each other. Similarly, while individual use is still of the utmost relevance for Couldry, he suggests an approach more akin to a macro level rather than the micro variety often focused on in combination with the Uses and Gratifications approach. Essentially, then, the practice approach, or media practice theory, suggests a focus on questions like “what are people doing that is related to the media?” (Couldry, 2012:35). Couldry does an excellent job at contextualizing what could perhaps be understood as a re-conceptualization of audience studies by suggesting that practice theory be understood as socially oriented media theory – one of four overarching perspectives of media and communication research (the other four being the political economy of media, medium theory and textual analysis). These four are outlined in a helpful graph included at the beginning of the book, which provides a suitable point of entry into Couldry’s thinking and suggestions about the larger areas of the field.

The suggested practice approach is arguably the key contribution of Couldry’s ambitious and meticulously researched text. Throughout the book, the author draws extensively on scholarship from a series of different academic traditions – most notably sociology and media and communication studies – to build his case. The concept of media practice wishes to uncover the ways in which people integrate media into their everyday lives in almost tacit, implicit ways – in order to meet our many diverse human needs. For Couldry, the media are (almost) everywhere, a line of thought that – with some obvious variation – has been touched upon recently also by others (e.g. Deuze, 2012). However, Couldry also makes his case by providing a series of theories more applicable with regards to empirical research, thereby making the book an even more interesting read. Given the authors interest in the humdrum of everyday life in relation to media use, I was nevertheless somewhat surprised to find very little reference to the work of scholars related to the Birmingham school – such as Stuart Hall, Richard Hoggart or Raymond Williams. Given their interest in the role of media institutions within complex issues like these, their perspectives could have strengthened Couldry’s case even further. Moreover, given my own personal research interests, I particularly enjoyed what I perceived as a tastefully critical stance from Couldry with regards to the consequences often thought to follow from whatever media format or channel is the ‘newest’ one at the present time. As one of many examples dealt with in the text, the author shows how despite much of the hype surrounding new opportunities for media consumption, the influence of television will most likely remain large for the foreseeable future, while the machine itself – the television – might converge beyond recognition.

While I agree that the practice approach championed in Media, Society, World: Social Theory and Digital Media Practice seems like a suitable way forward for research into the uses of media, I would argue that the need for more technically informed research is also a clear and present one. This is in some way touched upon by Couldry in the aforementioned model he uses to contextualize his theorizing, but we might want to take the actions made available by new media – their affordances, if you will (e.g. Gibson, 1977; Leonardi, 2011) – into clearer account. While novel technologies often carry with them the potential to change user practices or even overarching structures of use or “produsage” (Bruns, 2008) in relation to the media, empirical research from a number of different contexts has proven that most uses are of the more traditional variety. With this in mind, while Couldry’s admittedly elegant focus on the overarching level seems suitable, I worry that such an approach might miss out on some elusive detail that emerges in relation to whatever media format, service or platform is ‘all the rage’ this month. Such a more artifact-oriented view, detailing what is technically possible or not, is perhaps better understood as the topic of another book.

References

Blumler J.G. & Katz, E. (1974). The uses of mass communications: Current perspectives on gratifications research. Beverly Hills, CA: Sage.

Bruns, Axel (2008). Blogs, Wikipedia, Second Life, and Beyond. From Production to Produsage. Peter Lang.

Couldry, N. (2012). Media, Society, World: Social Theory and Digital Media Practice. Cambridge, United Kingdom: Polity.

Deuze, M. (2012). Media Life. Cambridge, United Kingdom: Polity.

Gibson, J. J. (1977). The theory of affordances. In R. Shaw & J. Bransford (Eds.), Perceiving, acting, and knowing: Toward an ecological psychology (pp. 67–82). Hillsdale, N.J.: Lawrence Erlbaum.

Leonardi, P. M. (2011). When flexible routines meet flexible technologies: Affordance, constraint, and the imbrication of human and material agencies. MIS Quarterly, 35(1), 147–167

Most-Read Article at European Journal of Communication

Larsson_Kalsnes-EJOC

I’m happy to report that the article “‘Of course we are on Facebook’: Use and non-use of social media among Swedish and Norwegian politicians” was apparently the most-read article from the European Journal of Communication during January 2015. The article, which I co-authored with Bente Kalsnes, deals with adoption and use rates of social media by Swedish and Norwegian politicians. The abstract is available below:

While plenty of research has provided important insights into the uses of the Internet by politicians during elections, a relatively scarce amount of work has looked into these uses outside of such parliamentary events. This article seeks to remedy this lack of research by presenting a study on the ‘routine’ uses of two of the currently most popular social media services – Facebook and Twitter. Focusing on politicians elected to the national parliaments of Norway and Sweden, the article employs novel methodologies for data collection and statistical analyses in order to provide an overarching, structural view of the day-to-day social media practices of Scandinavian politicians. Findings indicate that use levels are rather low for both services – the median amount of tweets sent and messages posted on Facebook is close to one per day. Further analyses reveal that the most active politicians could be labelled as ‘underdogs’, as they are more likely to be younger, in opposition and out of the political limelight.

By means of funding provided by my department, we’ve been able to make the paper available for everyone – also for those interested who might happen to reside outside of paywalls, the proprietors of which make Rupert Murdoch look like a socialist.

Going Viral? Comparing Parties on Social Media During the 2014 Swedish Election

convergence

The first of what I hope will be a series of publications dealing with various social media activities during the 2014 Swedish elections has been accepted for publication in Convergence: The International Journal of Research into New Media TechnologiesThe paper, titled “Going Viral? Comparing Parties on Social Media During the 2014 Swedish Election”, makes a case for comparing the oft-studied Twitter platform with the more seldom studied Facebook variety in this regard. You can find the accepted pre-print version on academia.edu or ResearchGate. Moreover, thanks to my department, the paper will be made available open access through the SAGE Choice program. The abstract is inserted below:

While research has provided useful insights into political party use of Twitter, comparably few efforts have focused on the arguably more popular Facebook service. This paper takes both platforms into account, detailing similar functionalities and providing analyses of the social media activities undertaken by Swedish political parties during the 2014 elections. Moreover, the types of feedback received by the parties on these platforms are gauged. Findings suggest that while sizeable parties are not necessarily the most ardent at using social media, they do receive the most attention. In essence, then, party size matters – however, the role of the right-wing populist Sweden Democrats is clearly felt throughout, suggesting the apparent prowess of controversial parties in the online context.

Employing such a comparative approach brings with it a series of challenges – how to present activities undertaken by the parties in a way that makes it easy (or at least relatively easy) to assess both services? Employing my previously discussed suggestions on how to compare Twitter and Facebook, I used the model previously suggested in a more empirical sense. Using the Tableau application, I graphed the types of feedback received by the parties in two graph, one for each service. As an example, the graph depicting Facebook activity is included below, together with an excerpt from the paper itself. The remainder of the analysis is of course available in the paper itself.

Larsson_Convergence_Figure_3

Figure Three. Average feedback received per post on Facebook. Horizontal axis indicates M of Comments/post; Vertical line indicates M of Shares/Post; Node size and label indicate M of Likes/Post.

The results presented in Figure Three suggest a linear tendency among the represented parties – meaning here that as the average statistic for the redistributive type of feedback for Facebook (shares) increase, so does also that same statistic for interactive feedback (comments). Given the sizes of the nodes and their corresponding labels, this tendency of increasing averages as we move diagonally from the downward left corner to the upward right can be discerned also here. Indeed, correlation analyses using Spearman’s Rho proved correlations between all three involved variables to be significant (p =<.000 for all correlations) and comparably strong with Rho varying between .779 and .871 for all correlations. With this in mind, it would appear that while the Social Democrats – marked as ‘S’ in the Figure – only amassed the fourth highest amount of ‘Facebook fans’ on the service at hand (see Table One), these fans appear to have put in quite an effort to make the specified account visible. This seems especially true in terms of Shares (M = 547, Std. dev. = 442) and Likes (M = 6162, Std. dev. = 5128), where high standard deviations nevertheless suggest a considerable spread around the reported means. Similar claims appear as valid for the average number of comments received (M = 368, Std. dev = 323). Focusing on the content provided through the account, the most Liked (N of Likes = 28810) post offered by the Social Democrats is coincidentally also the most commented (N of Comments = 2361). This post, penned late election night, features party leader Stefan Löfven giving thanks to party supporters and staffers after the party had been declared victorious . The most shared content from the Social Democratic account is posted on August 27th – the very same day that advance voting possibilities opened for the upcoming election. Consequently, this post urges supporters to vote in advance and to share the post itself in order to spread the message about this possibility . This tendency of popularity of posts that encourage supporters to vote and to share this encouragement using the redistributive functionality of Facebook is visible also for other parties, such as the Conservatives (M) and the Environmental Party (Mp), although not for the other stand-out party as visible in Figure Three – the right-wing populist Sweden Democrats. Here, the most shared as well as the most commented post is offered on September 4th and features sneak premiere of an election commercial to be broadcast on Swedish commercial television the following day . Their most successful post in terms of Likes (N of Likes = 9648) was made available on August 20th and deals with the controversy started when the party wanted to place their admittedly polarizing political advertisements on Stockholm public transport buses .

Zürich

Zurich

I have just returned home from a week spent in Zürich as a participant in the New Perspectives on Populist Political Communication workshop and launch event of the COST Early Stage Researchers Think Tank, neatly organized by Sven Engesser and Nayla Fawzi. This was a chance for me to relate more clearly to a topic I’ve been approaching over the last couple of years. In my more general studies of political actors online, the results almost always indicated a stand-out role in some regard for actors that could, by the rationale of your choice, be singled out as populist. Consider, for example, the role of the Norwegian Progress Party during the 2013 elections, or the enormous online popularity enjoyed by the Sweden Democrats. As such, the Zürich meeting gave me a chance to re-focus some of my empirical material to focus more specifically on these types of actors – an enjoyable exercise indeed. These were productive and enjoyable days, and I hope that the group convened in Zürich will have the opportunity to meet again.

New publication - European Journal of Communication

EJOC

Recently accepted for publication in European Journal of Communication, “Birds Of A Feather Flock Together? Party Leaders On Twitter During The 2013 Norwegian Elections” will probably be my last research contribution delving into the 2013 Norwegian elections. But then again, a couple of ideas are brewing in the back of my head, so you never know. Anyway, the paper at hand, co-authored with the illustrious Øyvind Ihlen, studies the Twitter use Norwegian party leaders during said election. While previous research has largely found Twitter to be somewhat of an ‘elite medium’, our findings suggest changes in this regard. Indeed, the party leaders make use of Twitter’s @reply functionality to higher degrees than would have been expected from previous scholarship. Nevertheless, this type of communicative behaviour is mostly undertaken in unique clusters of users featuring little overlap. As we suggest towards the end of the paper,

In one sense then, it could be argued that the findings strengthen the idea of the existence of echo chambers, of actors seeking together to have their views reinforced with the assistance of like-minded actors. The researched party leaders have their own clusters of users they choose to communicate with, including some citizens. Thus, Twitter probably functions to maintain good relationships with, literarily, followers. As, pointed out by Davis (2010), online politics of this type may encourage a trend where “tightly linked, cross-referencing and self-regarding” networks alienate others (p. 113). Viewed differently, however, the Twitter exchanges could also be seen as serving as inter-group honing of political arguments before taking part in the wider political discussion. Twitter is, after all, just one of several public spheres where politics is discussed.

Thanks to my department, the paper will be made available open access through the SAGE Choice program. Until then, the preprint version can be accessed at academia.edu or ResearchGate.

Retweeting during the #val2014 Swedish Elections

While the previous post compared the overall uses of #val2010 and #val2014, two hashtags dealing with the Swedish elections of 2010 and 2014 respectively, the current post deals with the latter of the two elections. Specifically, I will offer an initial analysis of what users enjoyed the most popularity under the selected hashtag. While popularity on Twitter can certainly be defined and operationalized in a number of ways – number of followers, for instance – I would argue that the degree to which an account is retweeted in a consistent manner is a complimenting, if not even better measure. With this in mind, the network graph below, produced in Gephi, gives us an overview of who enjoyed plenty of retweets under the heading of the specified hashtag – and who appears to have been especially active in sending those retweets.

RTs-tweet_from-retweeted_by_force atlas-degree range 100

The graph (please click it to enlarge it) consist of a series of nodes, each representing an account. The larger the node, and the label identifying it, the more that particular user was retweeted. The darker red the color of the node, the more active that particular user had been at sending retweets. Moreover, using Gephi’s Force Atlas algorithm, the graph has been delimited to show only the top users in these ways. As such, many of the nodes (accounts) that contributed to the sizes of the other nodes have been omitted so as to only focus on those highly active users. The table below identifies some of the very top users in terms of receiving retweets for #val2014, and compares this distribution with that of the #val2010 hashtag. The latter of these two election hashtag was studied in this paper (pdf).

#val2010
Top Retweeted
N  Comments #val2014
Top Retweeted
N Comments
piratpartiet 1228 Pirate Party niklassvenson 9338 Journalist
Pihlblad 521 Journalist socialdemokrat 6302 Social democratic party
emanuelkarlsten 392 Journalist RebeccaWUvell 3643 Liberal debater, PR consultant
mickep2 335 Journalist Expressen 2426 Tabloid
SDopping 324 Journalist svtdinrost 2366 PSB election feature
danielswedin 318 Journalist jonssonjessica 2174 Anonymous
nikkelin 310 ITConsultant MXCartoons 1474 Anonymous, Immigration critic
Falkvinge 309 Politician, Pirate party ozznujen 1419 Comedian
annatroberg 307 Politician, Pirate party Pihlblad 1418 Journalist
UlfBjereld 305 Professor piratpartiet 1350 Pirate Party

The table shows how the dominance of journalists and Pirate party associates in 2010 appears to have been broken for the period leading up to the 2014 elections. Indeed, while the Pirate party account is featured also among the 2014 roster of high-end users, and while two journalists did indeed make their mark in this regard (niklassvensson and Pihlblad), the distribution for 2014 appears as more varied than for 2010. Much like for top @reply receivers discussed previously, we see media organizations (such as Expressen), rather than a multitude of journalists, emerge as successful in gaining retweets. Moreover, the presence of a well-known comedian – ozznujen – reflects another tendency identified by similar research, where celebrities gain leverage in online political discussions (see Larsson and Moe, forthcoming). Finally, two anonymous users succeeded in getting their tweets redistributed to high degrees in 2014. The person or persons behind the MXCartoons account provide a statement basically saying that the tweets will feature “common sense over political correctness”, which is instated by largely discussing and criticizing Swedish immigration policy. Perhaps more interestingly, the influence of the user jonssonjessica appears to have emanated from one tweet only. As in the case of the 2013 norwegian election, one such well-formulated and likewise timed tweet can gain attention – if ever briefly. Much like for the Norwegian case, this particular tweet was sent on election night – lamenting the fact that the incumbent liberal-conservative government appeared to have lost the popular vote. As such, while the comparably large node representing jonssonjessica might suggest an influential user, the placement of the node, relatively isolated from the others, suggest that this was largely a ‘one-off’.

Comparing Swedish Elections on Twitter - First Impressions

As part of my collaboration with Hallvard Moe, we are now looking at data on Twitter use from the recently held 2014 Swedish elections. We are thus in a position to compare various aspects of Twitter use from two elections – 2010 and 2014. The table below features some of the overarching characteristics of the major hashtags employed for each election – #val2010 and #val2014.

Type of tweet #val2010

N (%)

#val2014

N (%)

Change

N

Undirected 60 088 (60.2) 89 747 (36.5) + 29 659
Retweet 32 780 (32.8) 146 487 (59.6) + 113 707
@replies 6 964 (7) 9 411 (3.8) - 2 447
TOTAL 99 832 (100) 245 645 (100) + 145 813

Looking first at the row providing the “TOTAL” amounts of tweets for both elections, we can see that the traffic more than doubled in 2014 when compared to 2010 – from 99 832 tweets during the month leading up to the former election to 245 645 tweets during the same time period four years later. This growth was perhaps to be expected, given the overall increase of media coverage devoted to Twitter, as well as the more modest increase – but still an increase – in actual Twitter use. A change that was less expected was the shift regarding the most common form of tweets sent. In 2010, just over sixty per cent of tweets were undirected – tweets sent without any intended recipient.  In 2014, retweets – redistributed messages originally sent by some other user – make up for about the same percentage level, effectively becoming the most common type of tweet sent. For @replies, signalling discussions on the platform at hand, we see an increase in numbers of such messages sent from the 2010 to the 2014 period. Looking at percentages, however, these numbers indicate a distribution diminished by almost half, decreasing from seven per cent in 2010 to just under four per cent in 2014.

So what does this all mean? One initial, tentative interpretation goes as follows. The dominance of undirected messages during the 2010 election period would seem to indicate a ‘megaphone’-like application of the service at hand. In other words, Twitter was here used primarily in order to ‘get the message out’ rather than for interaction or dialogue. The prevalence of retweets in the 2014 sample suggests that such practices might have changed. Perhaps due to the increased use levels of social media by political actors – a tendency which is sometimes discussed in relation to the overarching idea of political professionalization processess – the results presented above could be an indicator that such actors (and those similar to them) are successful in assessing the viral qualities of the platform. Plenty of retweets means plenty of spread for your message. Indeed, the patterns discussed here do not reveal which users enjoyed the most retweets. Initial data analysis regarding is coming up in a couple of days.