Digital Media for Behavior Change: Review of an Emerging Field of Study

"[N]ew technologies and opportunities to apply digital media technology for behavior change (e.g., how social media...can be used to promote healthy behaviors) are growing rapidly. Social and behavioral scientists must harness the potential of these technologies and develop this emerging field of study."

Researchers are increasingly studying the impacts of digital media on human behaviour - both positive and negative. The field of digital media for behaviour change is characterised by its use of digital device/human interactions: as an intervention strategy, as a methodology for data collection and research, and as the study of environmental influences (i.e., infodemiology) that may affect behaviour and moderate the effects of behaviour change interventions. Focusing primarily on the first two of these approaches, this paper provides case studies to illustrate how they have been applied and studied in recent research aimed at identifying effective strategies to bring about positive, population-level health behaviour change. The paper also raises questions to be addressed in future digital health research for behaviour change.

As an intervention strategy, digital media for behaviour change uses features of digital devices to communicate and create environmental cues and incentives (i.e., following behavioural economics) that encourage the adoption of new behaviours and/or the maintenance of existing ones. These approaches may take place on social media, mobile phone apps, chatbots, text, and social messaging (e.g., WhatsApp), as well as many specific modalities of information that can be communicated within them, such as video, memes, website links, and graphic images. Digital interventions have the flexibility to be based on tailored, individual communication and on group-level communication, such as in a Facebook group; peer-to-peer interventions are also possible. Digital platforms such as social media are inherently interactive, provide the opportunity for participant co-creation, and have the potential to engage participants and populations in the context of their social networks, thus building a sense of identification and connection with the intervention (e.g., through "gamification" features). Intervention strategies like these have the potential to create a sense of widespread support for and adoption of specific behaviours. Theoretically, the effect of such social media campaigns may be to promote a social norm in support of specific health behaviours, such as COVID-19 vaccination, healthy eating and physical activity, or avoidance of nicotine consumption.

As a data collection and research strategy, digital media may be used to conduct outcome and impact evaluations, or for formative research. For instance,

• Participants could be recruited into a research study using Facebook advertising (e.g., to join a research study), the data would then be then collected using chatbot technology, the same participants would be exposed to an intervention using remarketing technology, and subsequent data would be collected. Such an approach creates the potential for social-media-based randomised experimental studies to evaluate online behaviour change interventions. The Facebook Messenger (DM) app is one way to deliver surveys, which may be followed by randomisation to study condition to receive a social media and/or other treatment aimed at promoting behaviour change and/or other outcomes (e.g., changes in social norms or intentions).
• To help design campaigns for behaviour change, a formative study applied a behavioural lens to understand drivers of COVID-19 vaccination uptake among healthcare workers (HCWs) in Nigeria. Findings from an online survey of Nigerian HCWs informed a large-scale COVID-19 vaccination campaign that began in Nigeria in late 2021. The purpose of this effort is to use social media delivered by trusted organisations and influencers to promote HCW vaccination and broader population vaccination, in part through the role modeling effects of increased rates of vaccinated HCW. One advantage of using digital media for research is that it is relatively low cost. Large-scale data collection may be conducted with relatively low marginal costs of initial survey programming and data management. Incentives may be provided through such modes as electronic gift cards or mobile phone use credits, the latter being highly valuable in many low- and middle-income countries (LMICs).

The paper offers use cases across several content areas, including healthy weight management, tobacco control, and vaccination uptake, to describe and illustrate the methods and potential impact of this emerging field of study. One of the cases focuses on an impact evaluation of use of Nigerian social media to promote COVID-19 vaccination. Evans and colleagues are evaluating the performance of a series of social media behaviour change campaigns to reduce vaccine hesitancy amongst Nigerian healthcare providers (HCPs). The impact evaluation aims to determine which campaigns and strategies are effective in reducing vaccine hesitancy among HCPs and at what level of cost-effectiveness. It examines the mediating effects of social media engagement metrics and changes in norms and related beliefs about the efficacy and safety of vaccines on COVID-19 vaccine hesitancy, intentions, and uptake. The analysis uses structural equation modeling (SEM) and multi-level modeling (MLM) techniques to test the campaign's theory of change (ToC) for mediation based on changes in attitude, beliefs, and norms about vaccination, and for moderation of environmental, personal characteristics of HCPs, and other social ecological factors on the intervention's effectiveness. The project seeks to demonstrate not only multiple evaluation strategies but also how social media recruitment and data collection (via the Virtual Lab LLC tool) may be used to support social media interventions, especially as they focus on vaccination promotion. The materials developed in the project include surveys, interview guides, and social media content and metrics on dosage/exposure to the vaccination campaign at the population level. Researchers are compiling a library of these materials to be shared publicly, which they hope will increase the evidence for what works in digital health interventions worldwide.

In the paper's discussion section, the researchers reflect on the state of the art of the field of research on digital media for behaviour change. While there are many studies of "digital listening" projects and of large social media datasets (infodemiology), rigorous intervention studies on the effectiveness of digital media in actually changing behaviour, specifically in public health, remain relatively sparse. They suggest some specific research directions, such as more studies that involve LMIC populations, that include longitudinal follow-up to assess the long-term effect of social media behavioural interventions, and that examine the effectiveness of ToCs in social media interventions. In addition, future experimental research should rigorously examine the effects of variable levels of engagement with, and frequency and intensity of exposure to, multiple forms of digital media for behaviour change.

In conclusion, "the emerging field of digital media for behavior change...represents an important, relatively new domain....While the evidence base in this field, as illustrated in this paper, is relatively small, the field is growing. Future research and programs should expand the domains of subject matter addressed by digital media for behavior change. More rigorous, controlled, and externally valid studies are needed[,] and the field will need to stay current with rapidly emerging new digital technologies."