What Social Data Says About the Upcoming Snap General Election: The Studies, Trackers and Insights
Ahead of the snap UK general election on the 12th of December, we review the latest social data analysis initiatives and research conducted by analysts and agencies in The Social Intelligence Lab ecosystem. There’s everything from misinformation to candidate perceptions to finding out the issues relevant to Britain, and the use of paid social media.
Ipsos with Social Media Britain: The Views on Issues Relevant to Britain
The social intelligence team at Ipsos has been monitoring social media conversations about key political and societal topics. The research does not cover every topic that British social media users discuss, but it does give a good overview of the engagement around important societal topics. Have a flick through of the presentation below to see more specific insights.
Ipsos MORI – Social Britain September 2019 from Ipsos MORI
The study highlights the issues Britons are talking about, with Brexit and Europe being the most discussed. Followed by crime, the environment, healthcare, and the economy. Interestingly, these results are also consistent with the findings of the Ipsos MORI Issues Index.
Read more about the Ipsos Social Media Britain study here.
Storyful with Misinformation Detection
The team at Storyful analyze conversation and content across digital platforms, from mainstream to fringe networks. To counter the consequences of disinformation, they verify news and trends for news, media, and brands. For journalists reporting the news around the election, they recently shared their five fundamental questions they deploy to evaluate and verify content on social media.
The Who: finding the original uploader of a particular video or piece of content to understand its origins and the authenticity of the person sharing the video.
The What: what does the image or video suggest is taking place and action the content inspires the viewer to pursue?
The Where: geolocation to verify if the video is really where it purports to be from.
The Why: to understand the motivations behind sharing that particular piece of content.
Think that this election has been free of misinformation? Nope. Reddit recently announced the results of its internal investigation into a potential Russian disinformation campaign that was behind the leak of contentious US-UK trade documents on the platform. Read the full article here.
Pulsar, 89up and CrowdTangle with the Biggest Election Issues
The Guardian recently published research from social media analytics firm, CrowdTangle, and social media listening companies Pulsar and 89up. Data from public Facebook groups with the most election-related activity was gathered using CrowdTangle and then Pulsar and 89up analyzed which topics have received the most attention in online articles, posts, and tweets.
Here’s a selection of what they found…
Biggest Shares: with the purpose of producing and distributing pro-Labour memes, the Dank Meme Stash is one of the most active political groups on Facebook. The recent meme juxtaposing the claim that young people aren’t very productive with several pictures of MPs being unproductive of sleeping in parliament caught a lot of attention and shares online.
Biggest Conversation Topics: like the findings of the Ipsos study, they found that Brixt was the most popular topic. Then followed by the NHS, economy, racism, EU and the environment.
Bakamo with UK Elections Social Media Analysis
Bakamo has deployed their Open Polling which taps into social media users’ discussions with the aim of analyzing the narratives framing the election. They believe that the numbers reflect the intensity of political motivation and the intent to influence others to change the expected outcomes of the elections.
Here’s what they found…
Relative Insight with Newington Communications with Perceptions of Jo Swinson
Relative Insight are specialists in language analysis, they worked with Newington Communications to examine the online conversations about the Liberal Democrat Leader, Jo Swinson.
The research analyzed a snapshot of one typical day on Twitter during the campaign to compare the way in which Swinson is talked about in comparison with the language used to describe Boris Johnson and Jeremy Corbyn. There are few details on what the snapshot comprised of, but here’s what they found…
Arrogant: Swinson is three times more likely to be described using language such as arrogant, egotistical or patronizing than Johnson or Corbyn.
Deluded: Swinson is two and a half times more likely to be described of misleading or lying than Johnson or Corbyn.
If, like me, your shouting at the screen about all the studies showing that women receive harsher criticism than men in politics and leadership roles, and are described differently to men, the authors do state that this is a limitation of the study. Perhaps having other women in the study would have helped to understand this gender perception/description phenomenon.
Politico and NYU with British Political Groups Spending on Facebook Ads
To better understand how political parties are targeting voters online, Politico and a group of researchers from NYU analyzed all paid-for partisan ads brought by British groups on Facebook from the 7th of November onwards to find out who was buying what in order to target which type of voter. Here’s what they found…
Paid-for partisan ads bought by British groups on Facebook account for around 80% of all political messaging.
The researchers note that as Facebook has increased its transparency efforts around political ads, the research shows that partisan groups are aggressively turning to so-called organic content through pages to reach voters with messaging that often flies below the radar of existing digital transparency tools.
Third-party groups have been outspending traditional practices since the beginning of the election period.
Read the full article and results here.
Final Thoughts
This snapshot of our partners’ analysis highlights that there are many questions you can ask of social data, and various ways to segment the data. The research by Ipsos and Pulsar/89up shows the tracking of various popular conversations but they have chosen to segment the data in a different way. For example, Ipsos has Brexit and Europe together whereas Pulsar/89up has them in separate segments. The challenge for analysts is to decide on what is the most meaningful segment.