Citizens' Expectations on Brexit Outcomes: 'Fact' Transmission and Persuasive Power in a Digital World

Lead Research Organisation: University of Edinburgh
Department Name: Sch of Social and Political Science

Abstract

This collaborative project between the Neuropolitics Research Lab (NRlabs), at the University of Edinburgh and Full Fact, the UK's independent fact-checking organization, employs neuroscientific, psychological and behavioural insights to help us to understand what makes Brexit-related claims spread on digital platforms. Using cutting edge scientific techniques in big data analysis this project offers new insights into how citizens' expectations on Brexit and its consequences are shaped in an increasingly digital world. It will inform organisations on how to communicate what is often dry and complex information related to Brexit in a credible, trustworthy and memorable way using digital communications. These insights will be essential for the strategic management, implementation and public communication of the Article 50 process for the UK's withdrawal from the EU.

The question of what constitutes a fact (or an alternative fact) has perhaps never been more salient in public debate. The thirst for 'facts' during the Brexit referendum campaign was a key feature of public debate as was the question of whose facts count. The role of experts in the delivery of factual information came under close scrutiny and became a substantive feature of campaign dialogues. The question of trust and authority in information transmission has been under serious challenge. Citizens' expectations of Brexit and its consequences are, at least in part, shaped by their evaluation of the facts - but how do they decide what is a trustworthy fact? What factors lead them to imbue some sources of information with greater authority than others and under what circumstances do they choose to engage with, share or champion certain 'facts'? How does the context in which 'facts' are disseminated shape the expectations of the citizens on Brexit?

Digital technology and online communication platforms such as Twitter and Facebook, play an increasingly important role in the public communication of both information and misinformation. To date, however, we have little information on how 'facts' transmitted in these digital platforms are internalized by recipients and on how this information impacts on citizens' expectations. We investigate how membership of a specific social media bubble impacts on the evaluation of the information received; how the status of the sender or even the content of the communication (whether it contains an image or a web link) matters; and how the nature of the information received, confirmatory or challenging of previous knowledge, impacts on fact transmission to different publics.

This project builds on the extensive engagement of two research teams on Brexit-related research and with the UK in a Changing Europe team. Both teams are engaged at the highest level in stakeholder engagement and the project is built on a co-production model, ensuring that the issues addressed are of direct interest to those most likely to utilise the insights developed directly in their daily work. The project is designed in close collaboration with stakeholders to ensure that it can adapt swiftly to maintain relevance in the fast-moving Brexit environment. The project has access to a unique social media data-base of over 40 million tweets that NRlabs has collected on the Brexit debate since August 2015; the cutting edge skills and facilities for conducting experimental research at NRlabs; and ensures daily policy relevance through Full Fact's engagement, nationally and internationally, in the fact checking environment. The contribution of this project addresses the very heart of the mission of the UK in a Changing Europe programme - to be the authoritative source for independent research on UK-EU relations, underpinned by scientific excellence and generating and communicating innovative research with real world impact.

Planned Impact

Fact-checkers and Public Engagement Organisations:
This research will feed directly into Full Fact's work and has direct application to the UK in a Changing Europe programme's work, guaranteeing direct research impact. Understanding how a piece of information comes to be perceived as a fact, how this varies between groups and according to the characteristics of the communication, the status of the communicator and the context in which communication takes place is essential for the effective communication of factual information on Article 50 and the Brexit process more generally.

Policy-makers dealing with Brexit:
Harnessing insights from political science and public policy experts and working with psychologists, information scientists and cognitive neuroscientists, NRlabs has developed close links with government officials, as well as public and private users interested in gaining new insights into the psychology behind decision processes. In addition to the expected impact of our Brexit-related research findings (such as citations in the international press and the House of Lords report on Brexit; contributions to House of Common's inquiries on Brexit; Brexit MOOC; European Futures web platform; Laura Cram's role as the Scottish Housing Federation's 'Brexpert' and regular monthly contributions to their in-house publication; regular expert panel appearances; school events; public exhibitions; and providing overnight EU referendum analysis for STV), one of NRlabs major impacts has been our role in encouraging innovative patterns of joint working between government officials, policy practitioners and academics at a fundamental level thus helping to develop the very practice of making policies. This innovative and intensive model of co-production was noted as an important outcome of a major impact event on Brexit, that we organized 12-13 Dec, just prior to the release of the Scottish Government's Brexit plan. A second Scottish Government workshop to run in April 2017, will employ this same co-production model, examining citizens' expectations of Brexit outcomes.

Influencers of public debate on Brexit and beyond:
Full Fact has worked with many of the major influencers of public debate, from the ESRC, the Institute for Fiscal Studies, the Office for National Statistics, the UK Statistics Authority, Wikimedia UK, the House of Commons Library and Parliamentary Office for Science and Technology at the information-providers end of the scale, through to national media such as the BBC, ITV News, Sky News and LBC, and membership organisations including 38 Degrees and the Federation of Small Businesses. Full Fact also has well developed relationships with the International Fact-Checking Network, and strong links with fact-checkers in Argentina, South Africa and the USA, where there is huge demand for this work. The unique nature of this collaboration brings significant added value to existing work on big data and the digital environment in the Brexit policy space and has potential for extension to other policy debates and electoral campaigns. Full Fact is working to predict and preempt information gaps before the 2020 election so that the public debate can happen on a solid footing. Our regular sand-pits are designed to ensure involvement of these stakeholders in the digital data analysis and in the conception, design and execution of the pilot experiments throughout the project.

Polling Agencies:
NRlabs tackles political and policy-relevant issues in a transdisciplinary environment using a neuropolitical perspective. The unique insights generated into citizen expectations, through the combination of psycho-social, behavioural and big data analysis, provide a powerful compliment to existing opinion poll data. Polling experts, including current UK and EU fellow, John Curtice, IPSOS Mori and Yougov will be invited to participate in our regular sand-pits to help develop and pilot our experimental interventions.

Publications

10 25 50

publication icon
Moore A (2021) Trust in information, political identity and the brain: an interdisciplinary fMRI study. in Philosophical transactions of the Royal Society of London. Series B, Biological sciences

 
Description Our aim in this project was to get 'under the hood' of citizens' expectations in relation to ongoing Brexit developments and to shed light on: (i) how expectations about the Brexit issue and its outcomes are influenced by the presentation of 'facts' and 'counterfacts' on digital platforms; and (ii) how the most effective and accurate digital communication to citizens of accurate information in relation to the Brexit process and its potential outcomes can be achieved. We achieved this through a combination of big data Twitter analysis, online experiments, lab-based behavioural and physiological studies and fMRI brain imaging. Our Twitter analysis revealed the power of images and source (including celebrity and verified status) in ensuring the resonance of a message in the digital sphere. We demonstrated a clear lack of overlap between and within Leave and Remain communication networks in the Twittersphere underlining the heavy reliance on selected sources within the Brexit camps. The ongoing resonance of the £350 million promise for the NHS, in Twitter dialogue, despite its immediate statistical invalidation demonstrated the power of repetition over the power of correction. At the experimental level, we explored the issue of trust and the role of different information sources at increasingly granular levels. Our work was novel in that we attempted to combine research strategies/findings from work on (a) fake news, misinformation, and motivated cognition/reasoning with (b) the psychology of explanation (i.e. what makes a good explanation). From the former we leveraged the (mis)match between claims, rebuttals, and pre-existing political beliefs/ideologies to examine the impact on the likelihood of belief change as a result of conflicting information, and from the latter we employed differing types of explanations for why claims were upheld/rebutted in corrections. We also explored the impact of information regarding the source of information provided in corrections. Overall, it was difficult to change beliefs (to cohere with rebuttals) when measured over very short time intervals. In fact, there was some evidence for backfire effects - participants increasing belief in (a) rebutted/disproved claims when confronted with a correction that clashed with their political/ideological preference and (b) decreasing belief in claims, supported by corrective evidence, where both clashed with participants' political/ideological preferences. These findings are limited by the aforementioned short time scales - beliefs are resilient to challenge, typically, and so change may not be reliably measurable on timescales used here. Similarly, we could not address the asymmetry between offering evidence-based corrections on a single occasion versus having rebutted/debunked claims repeated large numbers of times in the media by politicians/authority figures before and during the course of an experiment. Examining the mechanism underpinning trust in information sources, we find that information source matters. Our physiological, behavioural and fMRI's study confirmed that exposure to the logo of selected information sources affected responses to the Tweet messages at a behavioural, emotional, physiological and neural level. Emotional self-report and physiological measures did not, however, directly align - this is an emerging discussion in the emotion and cognition literature and is worth pursuing further.
Exploitation Route The project will feed into and stimulate further theoretical and methodological innovation in the relation to the mis/disinformation ecology as well as the wider literatures on emotion, cognition and trust. The team has worked throughout with government at both Scottish and UK-levels, and with EU officials, developing better understanding, and measures of digital influence, and of how our political nature impacts on public and policy-maker behaviour and political communication. Our collaborative stakeholder event on political persuasion was hosted in the offices of Twitter and attended by Twitter's senior public policy manager for the UK. We have published extensively on political influence on political social media activity and our work has been referenced in public inquiries in the US Senate and UK government select committee reports. We are currently collaborating with major corporate actors on developing an extension of this project, exploring the use of data and digital influence for the public good. This will feed into future collaborative research and impact agendas as well as informing programmes of executive education.
Sectors Digital/Communication/Information Technologies (including Software),Government, Democracy and Justice,Security and Diplomacy

 
Description We have been consulted by both the FCO to the presence of Russian IRA troll activity in our Twitter data set and by the Home Office in relation the process of disinformation in digital media. In November 2017 it became clear that our Brexit-related Twitter data set contained evidence of activity from accounts identified by Twitter to US congress as fake Russian 'troll' accounts, set up to disrupt the US election. This issue involved: (I) complex legal issues (in terms of our user agreement with Twitter); (2) ethical issues, in terms of our commitment to data privacy and our commitment to preserve evidence of public interest; (3) diplomacy as this was an issue of international concern that attracted the attention of the Russian and French embassies as well as the Foreign and Commonwealth Office, who all contacted us about these findings; and, (4) public and media relations - our findings were in almost all national papers and were published on the front page of the Guardian, in the New York Times, Wall Street Journal, Washington Post, the Economist, the Times, the Independent, the Huffington Post, VICE, Buzzfeed and featured on ITV, BBC and Channel 4. We have met with Stewart McDonald, SNP spokesman for defence, have been asked to brief the FCO on data methods and will be running a cross-party briefing on cyber-defence in early 2018.
First Year Of Impact 2017
Sector Government, Democracy and Justice
Impact Types Policy & public services

 
Description Full Fact Partnership 
Organisation Full Fact
Country United Kingdom 
Sector Charity/Non Profit 
PI Contribution Partners on Brexit Priority grant
Collaborator Contribution Application stage, experiment design consultation , co-production of presentation, impact
Impact Twitter-hosted presentation; Home Office consultation; First Draft disinformation conference; Social Research Network presentation
Start Year 2017
 
Description Consultation with Home Office 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Policymakers/politicians
Results and Impact Consultation
Year(s) Of Engagement Activity 2018
 
Description Disinformation - First Draft 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact Presentation
Year(s) Of Engagement Activity 2017
 
Description Presentation in House of Commons 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Policymakers/politicians
Results and Impact Presentation
Year(s) Of Engagement Activity 2017
 
Description Sandpit hosted by Twitter 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact Sandpit event
Year(s) Of Engagement Activity 2018