Trusting Others: The Credibility of Numbers on Personal Messaging Platforms

• In Event: What to Believe? Facts in a Post-truth World

Abstract

WhatsApp has over two billion active monthly users, Facebook Messenger around 1.3 billion. On these platforms, citizens encounter myriad numbers, from the daily number of coronavirus cases to record numbers of "excess deaths." As some prior research has shown, quantitative information can exert particular power because it can falsely convey mathematical certainty and scientific rigor. But nothing is known about how numbers are appraised on personal messaging, where distinctive social dynamics apply. So, in this study we ask: how do people distinguish between numerical information and numerical disinformation on messaging? Our findings, based on in-depth longitudinal qualitative fieldwork over sixteen months, reveal the importance of two trust-based strategies: trusting peers and trusting public discourse.

As we show, these strategies interact in complex ways due to what we define as the hybrid public-interpersonal nature of personal messaging. Numbers cascade from the public spheres of journalism, scientific expertise, and politics into these private platforms, entering into complex interpersonal dynamics between friends, family, colleagues, and neighbors. Numbers also flow the other way in the form of guesses, inferences, and assumptions made by individuals, which are then shared. In this context, trust plays an important role in determining numbers’ credibility. But this is a far from simple process. Someone may trust information from the New York Times, but when it is shared on WhatsApp by a cousin notorious for sharing false information in the past, the group may be sceptical about its accuracy. Therefore, we argue that trust is not fixed and individualized in these environments—it is practiced, context-driven, and relational.

To examine these processes in detail, we used an exploratory, longitudinal qualitative design, conducting in-depth interviews with a sample of people from the UK that broadly reflect the key demographic features of the population (Wave 1 N=102, Wave 2 N=80, retention 78.4%, April 2021–July 2022). Between the first and second interviews, participants were also invited to upload content from their personal messaging using our customized smartphone app. We conducted emergent interpretive coding of the data.

Key Findings

Contrary to popular narratives and some evidence from prior research, personal messaging users rarely saw numbers as objective scientific “facts.” Instead, they perceived them as uncertain and potentially biased. They actively engaged in determining the credibility of numbers, but this was not usually performed through “DIY research” where people dig into the details using online searches. Instead, participants constantly negotiated the credibility of numbers by drawing on complex trust practices that fell into two main categories.

First, participants talked in detail about trusting peers, by which they meant friends, family members, or work colleagues in their messaging interactions. They used two main tactics: they would assess what they considered to be the expertise of peers, which was often loosely defined in terms of qualifications, or they would consider their peers’ reputation, which was based on longer-term observations of personal characteristics, accomplishments, and demonstrated behavior.

Second, participants devised ways to trust in what they self-defined as public discourse. For some, this involved navigating the mainstream, either by relying on perceptions of loose consensus among trustworthy media and public institutions or actively seeking out different sources. For others, however, public discourse meant “alternative” media sources and positioning fringe news sources as “trustworthy” counter-voices speaking a “truth” they thought was rarely found in mainstream news.

Our study provides important interpretive insights that could be developed in future research and policy on disinformation. It is unrealistic to expect citizens to directly verify all information they encounter on personal messaging platforms; too much of this requires time and technical expertise. Therefore, we argue that efforts could focus on understanding and nurturing the trust-based strategies we have revealed. Trusted peers are clearly vital for people to navigate the sea of information. But there are challenges too. Those in trusted roles in personal messaging interactions need to take their responsibility seriously, and greater attention should be paid to the power of individuals in messaging networks to operate as trusted sources. Equally, those placing trust in peers need good ways to establish trustworthiness. Peer expertise and reputation are starting points, but people could also be encouraged to more actively juxtapose those traits with genuinely trustworthy media and public institutions. This holistic, practices-based approach provides a new way of countering numerical disinformation in the increasingly important, yet difficult to research, world of personal messaging.

Authors

• Brendan Lawson, Loughborough University

• Andrew Chadwick, Loughborough University

• Natalie-Anne Hall, Loughborough University

• Cristian Vaccari, University of Edinburgh