Issues in Interdisciplinarity 2020-21/Evidence in the Propagation of Digital Fake News

Introduction
Fake news is the collection of shocking, surprising and necessarily false stories that come across as news, often digital. The propagation of fake news has in recent years subverted political and social institutions, posing a threat to democracy and human unity. This case-study will observe the issue through an interdisciplinary lens and look at evidence research undertaken in psychology, sociology and economics. By comparing the different disciplinary perspectives on evidence in the propagation of digital fake news, it becomes clear that the most effective solutions to combat this worldwide threat will come from combining disciplinary approaches.

Evidence in Psychology
Psychology is the study of human behaviour and consists of biological, cognitive and social aspects. Research in this discipline focuses on the effects of one or more variables on one or more outcomes. Psychologists aim to look at how certain variables lead to a change in behaviour.

To do this psychologists most commonly conduct studies that can either be qualitative (case studies, interviews, etc.) or quantitative (quasi-experiments, true experiments). The results obtained are then used as evidence to support models, such as the idea of reconstructive memory and the misinformation effect, which is when information after the event interferes with memory of the original event. To provide evidence as to why people believe fake news, one study presented participants with six news reports including two fake ones. Interviewees were asked if they remember the events of the six stories. Almost 50% of the participants said they remembered the false stories, and even provided details about the fake event. This study suggests the extent to which people can believe fake news because they are recalled as real memories. It also shows that people are more likely to create a false memory if the event is aligned with pre-existing beliefs. Studies have also shown that people are more susceptible to fake news that is phrased in a simplistic way (e.g. immigrants are responsible for unemployment). Furthermore, if one lacks knowledge of the topic, researchers have shown that a motivational bias may play a larger role.

Evidence in Sociology
Sociology has been defined as 'the study of the development, structure, and functioning of human society'. Sociological analysis covers a wide range of social problems, changes, states of order and instability, and at different scales, from the family unit to society. It embraces critical theory as well as the scientific method of production of evidence.

In sociology, empirical evidence is in the form of either quantitative or qualitative data. Quantified data on social phenomena is conveyed through surveys, statistics and censuses. A large number of participants responding to closed questions allows for a large-scale analysis, representative of society. However, selection bias – such as an over representation of men or women – is a perennial issue in such studies. By contrast, qualitative data relies on unstructured, open-ended interviews. When analysing data, one can distinguish between independent and dependent variables.

Utilizing these methods, the Pew Research Center conducted a survey in 2016 sampling 1,002 adults 18 and older looking at factors that could increase the sharing of fake news. Goyannes and Lavin identified the dependent variable as the probability of sharing fake news and the independent variables as demographic (age, gender, political orientation, income) and situational (how often people access fake news, unconscious sharing of fake news, who is most responsible for preventing fake news). Using regression models to analyze the correlation between variables, they concluded that older, Republican men with a lower income are most prone to sharing fake news. Answers as to why results like these arise may be better approached with qualitative methods.

Evidence in Economics
Economics is the study of how 'individuals, businesses [and] governments...make choices about how to allocate resources' based on the assumption that all individuals act rationally. To gather evidence in the form of empirical results, the discipline applies statistical methods to economic theory and models, in a branch of economics known as 'econometrics'. Qualitative evidence is infrequently used within the discipline.

One of the main assumptions within economics is that firms want to maximise profit. Accordingly, many digital technology companies - from 'Big Tech' to online blogs - use advertising revenue models, since their largest source of revenue is from advertising. Quantitative evidence is collected on both ends of the advertising model, from the platform and from the advertiser, in order to optimise key performance indicators - that is, to maximise engagement and to achieve the most effective customer response to an advert. Similarly, beneficiaries of fake news include both the people that have adverts on websites containing fake news and the creators of the fake news. Fake news generates more attention than most news stories due to the strong reactions that it can produce, such as 'surprise, fear and disgust'. Hence, according to the assumptions of economics, there is an incentive for websites to spread fake news in order to direct traffic, and therefore revenue, towards their website or publishing platform. For example, the 'cost per click' advertising model illustrates how, if a fake news post is 'clicked' on, the publisher or owner of the website will receive revenue, demonstrating how the advertising ecosystem benefits from the propagation of fake news.

Conflicts of Evidence: informing policy-making
Each of these established disciplines undertakes very different approaches and methods with regards to producing evidence about the same issue. Although all three are labelled a social science, the outcomes of their individual evidence methods about this issue do not necessarily correlate.

There are clear tensions between the research methods undertaken to produce the evidence. Extrapolating results from psychology - i.e. humans do not always act rationally - contradicts a key assumption in economics; sociology because it assumes neither adds another dimension. A practical example of this is the limited impact of the enforcement by social media companies of 'flagging' posts identified as fake news. 'Flagging' assumes what economics presupposes, that the individual, as a rational actor, would naturally dismiss the post when confronted with a cautionary message. By contrast, the study of biases in psychology and social factors in sociology challenge this assumption by suggesting individuals and groups are far less independent than assumed.

Furthermore, the value of collaboration across disciplines is seen in the importance of scale in affecting each discipline's research progress. A sociology study may provide evidence about demographic behaviour on a smaller scale that cannot be extrapolated to apply to society as a whole, which is the default parameter of econometrics. The need for collaboration is also seen when evidence used and produced in psychology is often applied in economics to assess larger model assumptions from a more specific perspective.

Policy makers have recently begun to adopt an interdisciplinary perspective, as illustrated by the 2019 UK Government White Papers reassessing state regulation of social media. These suggest a framework led by a varied board of academic leaders: practical proof that the interdisciplinary approach towards solving this worldwide threat to public discourse is the most effective solution.