What drives the spread? A Discrete Choice Experiment on the role of heuristic thinking in the proliferation of propaganda content on social media

Social Networking Platforms (SNPs) have revolutionized the very conception of communication by enabling their users to perform an unprecedentedly active role in the production and dissemination of information, thus profoundly affecting traditional gatekeeping. This paradigmatic change has affected political communication as well, providing political actors with access to online environments characterized by virtually any entry barriers and very low constrains in terms of content creations. As extensively documented in the literature, this – together with scalability, anonymity, and automation – has facilitated the proliferation of propaganda content on these platforms. However, the existence of this kind of content on SNPs does not prove per se the effectiveness of this communication practice. Assessing how individuals process, evaluate, and act upon these political messages is hence crucial for determining the actual impact this manipulation attempt has on the beliefs and behaviours of its recipients. As such, this study aims at disentangling the functioning of online propaganda by identifying and assessing the factors that increase the likelihood that social media users will consider online political messages valid and shareable, thus enhancing their circulation on SNPs. Drawing from the Dual-process Theory of Cognition, this research investigates how heuristic information processing – a mechanism that eases the cognitive load of decision-making but increases the chances of producing systematically biased judgements – affects users’ evaluations regarding content validity and shareability. The underlying hypotheses are that heuristic processing is 1) activated by the very design of propaganda content and 2) its effect on users’ evaluations is moderated by individual-level characteristics – i.e., cognitive and socioeconomic traits. To test this, an online between-subject Discrete Choice Experiment (DCE) involving 1001 participants from Italy has been designed, preregistered, and implemented. By means of this experimental design, it was possible to simultaneously assess the impact of six different informational cues – source, endorsement, popularity, emotional salience, moral valence, and stereotyping – on the likelihood that political messages would be considered valid and shareable on SNPs, and to evaluate the moderating effect both individual characteristics and debiasing treatments have on such likelihood. Results demonstrate the role played by both informational cues and individual-level characteristics on users’ evaluations and behaviour, thus providing useful insights about online propaganda impact on individual decision-making, showing how small nudges are effective at propagating the diffusion of polarising and misinformative content. Moreover, the deployment of this experimental technique represents a novel methodological approach to the study of political communication effects, since it allows for the assessment of preferences in reproduced real-world settings, where diverse individuals (i.e., with different cognitive and socio-economic traits) have to decide between different options (i.e., social media/political messages) characterized by a set of different features (i.e., different information cues that correspond to different persuasion strategies adopted by political communicators and propagandists).