Where the previous post dealt with the popularity (measured as the median amounts of likes, comments and shares) of Facebook posts made by Norwegian party leaders, the current post presents the popularity of posts characterized by their content. While an exhaustive description of the themes employed for this purpose goes beyond what is presented here, it is my intention here to provide such detail – rather, to give a quick glimpse into one of my ongoing research projects. Here we go:
As several of the error bars visible in this figure appear comparably larger to those reported in the previous one, the spread around the reported medians must be taken into account – especially when it comes to feedback in the form of likes, which much like in the previous figure comes up as most common across all themes. For this metric, then, posts featuring Acknowledgements (Md = 2494) – giving thanks to supporters, especially at the very end of the campaign – and those characterized as primarily focused on Critique (Md = 1881) emerge as yielding the highest amounts of likes, while the other metrics must be considered diminutive in comparison. As such, while showing gratitude towards or criticizing others appear to result in ample amounts of feedback, levels of engagement as a result of other thematic uses must be regarded as minimal in comparison. While an overarching analysis like the one performed here cannot provide a more qualitative assessment of the feedback measured, we can point to tendencies regarding what type of content, posted by politicians, that appears to resonate the most with their respective followers. As such, the findings provides some insight into what seems to “work” in an online setting – an empirical finding that is perhaps not especially conducive to the basic tenets of various normative theories of deliberation.
In a couple of previous posts, I compared the activity popularity of Swedish and Norwegian politicians on Facebook during a time when the latter of these countries underwent a parliamentary election – the idea being to look for tendencies of so-called permanent campaigning. In the present and coming post, I have looked more closely at the Norwegian data and tried to delve into two aspects of Facebook use at the hands of politicians – first, the division of likes, comments and shares between the party leaders; second, that same division across different themes posted about. Lets look at the former of these first. The data here were collected during a one-month period leading up to the elections, held in September of 2013. Please note the comparably small median shares and comments per post – a finding that suggests a need to re-think the so-called viral potential of social media services like these (at least in the context at hand).
The figure gauges the median amounts of likes, comments and shares per posts made by each party leader during the specified period. The error bars visible in the figure indicate the confidence intervals (95 %) for the reported medians, suggesting considerable variations regarding these metrics – especially with regards to the median amount of likes per post. With these distributions in mind, we can nevertheless conclude that especially in terms of likes, the leaders of the three largest parties – Stoltenberg (Md = 1587), Solberg (Md = 970) and Jensen (Md = 3878) – emerge as the most popular. As such, the suggestion made by Vaccari in his study of political web sites during the 2007 French presidential elections that larger parties “usually have stronger ICT infrastructures due to their superior resources” (2008: 6) appears to hold true also in a social media context (see also Gibson and McAllister, 2014). Jensen in particular stands out – especially when it comes to the median amounts of comments and shares. Here, the leader of the right-wing populist Progress Party appears to have enjoyed comparably larger amounts of success in terms of raising discussion and leveraging the viral aspects offered by the sharing functionality.
Gibson, R. K., & McAllister, I. (2014). Normalising or Equalising Party Competition? Assessing the Impact of the Web on Election Campaigning. Published before print in Political Studies
Vaccari, C. (2008). Surfing to the Elysee: The Internet in the 2007 French Elections. French Politics, 6(1), 1-22.
The past few weeks have brought with them good news in terms of publications. Specifically, three papers that I have been involved have been accepted for publication in three different journals. First of all, together with Jakob Svensson, I have written a research review outlining some possible ways forward for scholars interested in the online activities of politicians. Entitled Politicians Online - Identifying Current Research Opportunities, the paper has been accepted for publication in First Monday, an open access journal. While waiting for it to get published on the First Monday web site, the paper can be accessed in its pre-print version here. The abstract reads as follows:
For more than a decade, researchers have shown interest in how politicians make use of the Internet for a variety of purposes. Based on critical assessments of previous online political communication scholarship, this paper identifies a series of overlooked areas of research that should be of interest for researchers concerned with how politicians make use of online technologies. Specifically, three such research opportunities are introduced. First, we suggest that research should attempt to move beyond dichotomization, such as conceiving of the Internet as either bringing about revolutionary changes or having a normalizing effect. Second, while there is a considerable body of knowledge regarding the activity of politicians during election campaigns, relatively little is known about the day-to-day communicative uses of the Internet at the hands of politicians. The third section argues that as political communication research has typically focused on national or international levels of study, scholars within the field should also make efforts to contribute to our knowledge of online practices at the hands of politicians at regional and local levels – something we label as studies at the micro level. In synthesizing the literature available regarding the use of the Internet at the hands of politicians, the paper concludes suggesting routes ahead for interested scholars.
Building on the second suggestion outlined in the abstract above, the second paper (co-authored with Bente Kalsnes) is titled “Of course we are on Facebook” - Use and Non-Use of Social Media among Swedish and Norwegian Politicians and deals with the adoption and continued use of social media by politicians in the two specified countries. The abstract provides a bit more detail:
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 paper 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 paper 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 labeled as ‘underdogs’, as they are more likely to be younger, in opposition and out of the political limelight.
The study has been accepted for publication in European Journal of Communication, and while this is not an open access journal, I am working on making this paper (as well as a previous paper of mine already published in that journal) available by means of the SAGE Choice option. Stay tuned. In the mean time, two blog posts have featured some of the findings from the paper (1, 2), and the accepted version of the paper can be accessed here.
Finally, the third paper to recently be accepted is entitled Everyday Elites, Citizens or Extremists? Assessing the Use and Users of Non-Election Political Hashtags and will be featured in a forthcoming issue of MedieKultur. The abstract is featured below:
As research has indicated that what is sometimes described as traditional forms of political-parliamentary participation are dwindling in most western democracies, the role of the Internet has often been pointed to as harboring the means to hinder these developments. While empiricalstudies on these matters have at best provided mixed results, social media services, like Twitter,has yet again fanned the flames of the most enthusiastic debaters. This paper moves beyond the often-studied context of parliamentary elections and instead offers a structural study of everyday political discussions on Twitter. Specifically, tweets from political contexts in Sweden and Norway are collected and analyzed with a specific focus on the top users and their activities.Results indicate that while thematic Twitter discussion can indeed serve as a potential channel for citizens, the influence of established as well as political extremist actors is also clearly discerned.
The accepted version of the paper is available here. While the focus of this particular paper goes somewhat beyond the parliamentary context that I normally study, I feel that it is important to look also outside of the supposed orderly fashion of government to see how services like Twitter are being used for political expression. The key, I guess, is to study both contexts. But where to find the time?
The annual ICA conference is an important meeting place for communication scholars around the world, and given the results of the review process that were distributed yesterday, I will have the opportunity to once again present some of my ongoing research at this years meeting. This year, the conference is held in Seattle, a city I had the opportunity to visit during the 2011 AOIR conference – a great experience indeed, and I look forward to go back. I will be presenting some of the findings featured in the paper “Everyday Elites, Citizens or Extremists? Assessing the Use and Users of Non-Election Political Hashtags” which, as the title implies, presents a structural analysis of the types of users that take place in hashtagged political discussion on Twitter. Specifically, tweets from political contexts in Sweden and Norway are collected and analyzed with a specific focus on the top users and their activities. Results indicate that while thematic Twitter discussion can indeed serve as a potential channel for ordinary citizens, the influence of established as well as political extremist actors is also clearly discerned. This tendency for in particular right-wing populist actors to utilize the novel medium is especially visible in the Swedish contexts – a finding which is further problematized and discussed in the paper. If you would like to read the draft that was submitted to ICA, get in touch and I will be happy to send you a copy.
Moreover, my abstract entitled “Birds of a Feather Flock Together? Party Leaders on Twitter during the 2013 Norwegian Elections” (co-authored with Øyvind Ihlen) has been accepted for presentation during the Strategic Communication Campaigns in a Contemporary, Digital and Networked Society preconference workshop. All in all, it looks like a promising conference.
Continuing from my last post, dealing with Norwegian politicians and parties on Facebook, the current post presents data regarding the Swedish context. The graph posted here utilizes the same color scheme to make sense of the respective type of politicians visible - Black nodes equals parties, dark gray nodes show party leaders, light grey nodes denote ministers and “celebrity politicians” (e.g. van Zoonen 2005), and white nodes show activity undertaken by members of parliament without specific portfolios or indeed public profiles.
In the previous post, we could see that Norwegian non-parliamentary parties were comparably quite popular in terms of the median amount of Likes and Shares received per post. The figure presented above suggest similar tendencies, although not as stated, for the Swedish case. Consider the node representing the Pirate Party, whose placement in the graph indicates a median of shares per post on par with major, “catch-all” parties like the Social Democrats or the Conservatives. Similarly, the finding that two of the parties in the right-wing coalition currently governing Sweden (the Liberal Party and the Centre Party) are not present in the figure could be an indication of what could be labeled as an ‘ekection year effect’ – seeing as all Norwegian parties (who underwent an election earlier this year) were present as visible in the corresponding figure. With regards to the most popular Facebook Pages, we see another tendency repeated from the Norwegian context. Much like the Progress Party appear to have produced a series of posts yielding high amounts of both Likes and Shares, so do the right-wing populist Sweden Democrats seem to enjoy a similar status in the Swedish context. However, when one considers the change of scale on the vertical axis, representing the median of shares per post – 0-100 shares for Norway, 0-1000 shares for Sweden – the dominance of the Sweden Democrats in this regard is further affirmed. As such, far right parties appear to have succeeded in getting their message across through Facebook in both countries, while this tendency is arguably more affirmed in the Swedish context.
Earlier this week, I was interviewed by Norwegian Public Service Broadcaster NRK (you can see the piece here) regarding some work I have been doing on the how political actors (parties and politicians) make use of their Facebook Pages – and to what degree their activity spreads through the platform at hand. The figure below is referred to in the brief interview and I thought I’d post it here as well. Please click on the image to enlarge it if necessary.
Black nodes represents parties, dark gray nodes denote party leaders, light grey nodes identify ministers and “celebrity politicians” (e.g. van Zoonen 2005), whereas white nodes show activity undertaken by members of parliament without specific portfolios or indeed public profiles. Each actor is identified with name and party abbreviation. The vertical axis represent the median number of Shares per post by the identified actors, whereas the vertical axis shows the median number of Likes per post.
Starting with the Norwegian case, the figure above finds the node representing the official party account for the right-wing populist Progress Party – as well as the node corresponding to their party leader, Siv Jensen – to be among the political actors enjoying the highest medians of likes and shares per Facebook Page Posts. Beyond these and other top actors in this regard (such as PM Jens Stoltenberg), all political party accounts save for two (Socialist Left and Christian Democrats) are positioned above the horizontal dividing line, indicating the apparent popularity of official party accounts. As for the two parties below the aforementioned middle line, these are both small parties in terms of voter share. This suggested relationship between ballot recognition and Facebook Page post popularity is perhaps particularly interesting when considering the case of the Socialist Left Party. From another analysis – not published here – we could tell that while their official party account produced the highest yield of Page posts during the studied period, the figure above shows that their reach in terms of Likes and Shares was comparably limited. As a small Party, albeit with seats in government and a role as incumbents going into the 2013 elections, the Socialist Left Party appears to have had some difficulty in getting their messages across on Facebook.
This finding on the activities of an incumbent but small party on the left side of the Norwegian political spectrum can be contrasted with the spread that other accounts, operated by somewhat similar parties, appear to have enjoyed. Consider the nodes representing The Green and Red Parties in the figure above – both without representation in parliament. As visible here, these parties appear to have hosted comparably popular Facebook Pages, resulting in corresponding nodes placed in the middle of the figure. Taken together, this would seem to indicate that while party size appears to hold explanatory power regarding the online coverage enjoyed by parties, smaller, non-imcumbent parties are indeed able to get their message across on Facebook.
In a couple of days, I’ll post results regarding the popularity of Swedish politicians on Facebook. Stay tuned!
I morse deltog jag i norsk radio – närmare bestämt P2 Kulturnytt, sändningen kl. 08.05 (min del är tillgänglig här som mp3-fil). I inslaget diskuterar jag några av de fynd jag kommit fram till när det gäller hur norska partiledare i Stortinget använt sig av Twitter under pågående valkampanj – mellan 1 augusti ocb 1 september. Nedan återfinns då några av de grafer och figurer som jag refererar till under inslaget. Först: ett stapeldiagram som visar hur många tweets – och vilken typ av tweets – som partiledarna skickat under den månadslånga perioden.
Den partiledare som tycks ha varit mest aktiv på Twitter under den senaste månaden är Audun Lysbakken från Sosialistisk Venstreparti (SV) – ett parti som kan jämföras med svenska Vänsterpartiet. Tätt därefter finner vi så Trine Skei Grande från Venstre (V) – ett parti som namnet till trots kan förstås som ett parti i samma mer socialliberala tradition som svenska Folkpartiet. Om man då försöker hitta gemensamma nämnare hos dessa två partiledare som varit mest aktiva så kan man peka på att de båda leder relativt små partier – som sådana får de eventuellt svårare att få utrymme i de etablerade medierna. Då kan sociala medier som Twitter fungera som en kanal för att få ut sina budskap. Allra minst tweets skickades av Siv Jensen från Fremskrittspartiet (Frp) – ett parti som med svenska ögon sett påminner lite om det högerpopulistiska Ny Demokrati som satt i Sveriges Riksdag på 90-talet. Jensens begränsade närvaro här är inte helt oväntad – man har från partiet förklarat att Facebo0k, och inte Twitter, ska prioriteras när det gäller bruket av sociala medier.
Överlag är det slående hur kommunikativa partiledarna tycks vara – vi kan då fokusera på de grå staplarna som indikerar antalet riktade meddelanden (inleds med “@användarnamn”) – som partiledarna har skickat. Just denna typ av tweets tycks ligga i topp för samtliga politiker här utom Jensen.
Nästa graf är en nätverkskarta som visar vilka det är som poltikerna har skickat sådana @-meddelanden till under den senaste månaden. Klicka på bilden för en större version.
Figuren visar på tydliga kluster kring var och en av partiledarna: exempelvis ser vi tidigare nämnde Lysbakken längst upp i bild. De linjer som går mellan de olika klustren visar på den eventuella kontakt som finns mellan varje politiker och de personer som andra politiker twittrat med – ett slags mått på hur pass ofta man på Twitter anropar någon utanför sitt eget kluster, om man så vill. I mitten finner vi kontot Spdama – som tillhör Liv Signe Navarsete, partiledare för Senterpartiet. Just mittenplaceringen för denna partiledare tyder på kommunikationsmönster som sträcker sig utanför det egna klustret – något som också indikeras av de linjer som utgår från Navarsetes gröna kluster. Längst ner till höger i bild finner vi kontot som tillhör tidigare nämnda Siv Jensen. Avståndet från resten av politikerna, samt det faktum att inga linjer finns dragna från Jensens konto till hennes partiledarkollegor kan ses som symptomatiskt på dels den berøringsangst som finns kring Frp hos övriga norska partier – de är populära bland väljarna, men anses som kontroversiella bland de flesta andra partier. Det kan också relateras till den tidigare nämnda prioriteringen av Facebook för just detta parti.
Here, then, is the follow up post to my previous post on the paper I am currently working on together with Bente Kalsnes, regarding social media adoption and uses by Swedish and Norwegian politicians. In the last post, I presented the adoption rates for Twitter and Facebook Pages for our sample of politicians. Below, a series of luxurious greyscale box-and-whiskers plots provides some insights into the actual uses of these services by those elected to serve in the Swedish and Norwegian Parliaments respectively. Specifically, the scales on the y axis represent the number of tweets (upper row, grey boxes) or Facebook posts (lower row, white boxes) that the politicians had sent per day since they created an account on each service.
To begin with, a note on interpretation. Box-and-whiskers plots provide a visual overview the distribution of specified numeric variables. First, the ‘whiskers’ inform us about the spread of the data – indicating the lowest and highest data points. Any case – here, politician – found outside the whiskers is considered an outlier – here, a politician who stands out in his or her comparably extreme high-frequent use of social media. Second, for the ‘boxes’, the lines visible within them indicates the median of the distribution. Given the nonparametric distribution of the data, the median was determined as a suitable statistic. The areas in-between the whiskers, the box edges and the median line altogether allow us to discern the quartiles of each distribution – which in turn is helpful when assessing the skew of the analyzed data.
With these guidelines for interpretation in place, a few results stand out. First, for Twitter (upper row of the figure, grey boxes), the median values for both countries indicate that politicians tend to send out under one tweet per day (NO = 0.60 tweets, 0.95 tweets in Sweden). Compared to the means of tweets per day, these are considerably higher in both countries (NO = 1.01, SE = 2.85), suggesting a skewed distribution where comparably few politicians account for a large amount of tweets being sent. The same tendency can also be discerned when noting the slight upward skew of the boxes for both countries. Particularly for Sweden, this upward skew, indicating the activity of high-end users, is complemented by a series of outlier values – politicians who, due to their frequent use of the service at hand, appear to be literally ‘off the charts’ when compared to their colleagues. In both countries, these politicians could be described as “mid-level” with regards to their respective roles in parliament – while there are exceptions, particularly in the Swedish case (such as Annie Lööf or Birgitta Ohlsson), these outliers can largely be identified as members of parliament without specified ministerial duties or other, similar tasks usually associated with important work portfolios ascribed to them.
Second, for Facebook data (lower row, white boxes), the difference in scale as detailed on the Y-axis is striking when compared to our results regarding Twitter activity – while the former scale ranges from 0-40 tweets per day, the Page activity of the politicians in our sample can fit comfortably within a scale of 0-4 posts per day. This very basic results provides us with some initial insights into everyday uses of this particular service at the hands of politicians – a mode of communicating that is arguably not characterized by abundance. According to our findings, Norwegian politicians tended to provide a median of 0.28 Facebook Page posts per day (Mean = 0.36), with that same statistic for Sweden amounting to 0.11 posts per day (Mean = 0.26). While the middle boxes are still skewed upwards, they appear more evenly distributed around the median line in comparison with the Twitter data, indicating a comparably limited spread around the reported median for both countries – a claim that seems especially valid for Norway. Furthermore, as the medians and means are relatively closer to each other for the Page activity variable than for its Twitter counterpart, we can conclude that while Twitter activity is more abundant and characterized by upward skews on the scale caused by highly active politicians, Facebook Page activity appears as rather limited – a result that is further corroborated by the fact that the number of outliers for the latter of these scales are rather limited in comparison with the former. In comparison with the “mid-level” politicians found in our results pertaining to Twitter, the outliers in terms of Page activity tend to be top-level politicians – Siv Jensen is the leader of the Progress Party in Norway, while Jonas Sjöstedt leads the Swedish Left Party.
Time for some insights into the research I’m currently working on – over two posts, I’ll be sharing some of the work-in-progress data interpretations undertaken by myself and co-author Bente Kalsnes, focusing on the degree to which government politicians in Sweden and Norway adopt and make continuous use of social media services Twitter and Facebook. Please note that the results presented here have not been subjected to peer review – the headline for this post certainly stands. With that caveat in place, we look first at adoption rates. The figure featured below is based on data from the beginning of may 2013. Furthermore, it is arguably somewhat simplistic, but it does it’s job…
The figure depicts adoption rates – whether an MP in Norway or Sweden had an account on Twitter or Facebook or not – during our time of data collection. The bars indicate the percentages as well as the “raw numbers” (in parenthesis) of MPs with accounts.
While the fact that the Swedish parliament boasts more members than its Norwegian counterpart has influence on the numbers presented in the figure, the difference of scale is adjusted for when discussing these issues in terms of percentages. While Norwegian politicians have apparently adopted Facebook Pages to a slightly higher degree compared to their Swedish colleagues (twenty-four percent of Norwegian politicians use Facebook pages, compared to nineteen percent in Sweden), the differential when looking at Twitter is minimal (fifty-seven percent for Norway, fifty-eight for Sweden).
The results presented here indicate that Twitter adoption rates were slightly higher than the Facebook counterparts. As Twitter use in the two case countries is often reported at rather low levels, the popularity of Facebook among the general populace becomes relatively sizeable. Our results here indicate that politicians are generally more present and more active on the former of these two, suggesting a potential online communicative mismatch between those electing and those elected. While the demographics of social media use can vary considerably in different contexts, use of Twitter in the Scandinavian countries has mostly been associated with elite, urban, media-savvy groups. In sum, then, if politicians wish to engage more clearly with everyday voters, they might be well served to look beyond Twitter for their social media activities.
Of course, merely looking at whether a politician has an account on Twitter or Facebook or not does not take us very far in understanding the social media practices of elected officials – we also need to take their activity into account. This will be dealt in part II, coming up in a couple of days. Stay tuned!
At the end of the 2013 spring semester I was fortunate enough to be presented with not one, but two awards for my PhD thesis. First, the Börje Langefors Award (site in Swedish) for best Swedish thesis in Informatics defended during 2012. Apparently, the jury haven’t gotten round to updating the site with this year’s winner yet, but someone (not me, I swear) has provided Wikipedia with the correct details. The Wikipedia entry also features a nice translation of the motivation for me getting the award, which reads as follows:
The thesis is based on a socially relevant contemporary topic, well-designed and well-defined subject area with contrasting perspectives based on exceptional and interesting empirical material. The thesis is easy to read and well structured with well linked articles. It has a very good international exposure.
Second, I was named the first ever recipient of the FSMK (The Swedish association for media and communication research) Doctoral Dissertation Award. While neither the FSMK web, nor Wikipedia feature information regarding this, I have the diploma to back it up – in fact, both diplomas are pictured above. The motivation for this latter award translates as follows (translation by myself)
A well-written, well-structured and advanced thesis making valuable theoretical and methodological contributions within a topical and important area – especially in the study of Twitter, and utilizing Gidden’s theories. Larsson has contributed with important knowledge to the research area of new and social media in relation to politicians and audiences, thereby making an important contribution to the international field of media and communication studies.
Obviously, I feel very proud and honored to have received recognition from the two disciplines that I consider my two academic homes. Drawing on methods, perspectives and ways of thinking from both traditions have proven extremely fruitful for me so far, and I hope to be able to continue to do so in my future efforts.