Digital life has changed dramatically and the observation of it needs to keep up.
There has never been more interest and anxiety about the effects of media, whether about mental and physical health, education and relationships or politics and democracy. Already this year, there are new World Health Organization guidelines about children’s screen time; Congressional investigations about social media, political bias and voting; new state laws restricting smartphone use in schools; Presidential candidates discussing the breakup of large technology companies; and a number of advocacy groups championing technology policies they hope will reverse what many believe are existential threats to society. The proponents of new policies are eager for data, but it has been a challenge for scientists to accommodate.
We believe the reason for inadequacies in research, and lost opportunities in policy, is that no one knows what people actually see and do on their screens. The cost of inaction is considerable. Social policy in numerous areas, and the research on which it is based, is currently incomplete, absent, overzealous, irrelevant or wrong.
The major problem that scientists and policy makers face is that media experiences now defy easy characterization. The range of media content has become too broad (media now include relationships, work, entertainment, money, news, transportation, and even home lighting and irrigation), consumption patterns too fragmented (the average task on personal screens lasts only seconds), information diets too idiosyncratic (one hour of screen time is radically different from person to person), experiences too interactive (media experiences are created by users, not just passively consumed), and devices too mobile (we carry our screens with us everywhere). Quite simply, researchers need a better microscope, and better theories about how to use it, to see what is now mostly invisible.
The major opportunity is to use interdisciplinary science to develop that microscope and reveal the breadth of life experiences now reflected on our screens. We propose to launch the Stanford Human Screenome Project, a technology platform, analysis process and data repository that facilitates precise mapping of media use via detailed moment-by-moment capture and machine analysis of all the actual content, actions and sequences that appear on personal screens – defining what we call the screenome. In the same way that genomics reshaped understanding, prevention and treatment of different diseases, the screenome will inform and reshape understanding of a broad array of social problems. Social challenges are often siloed to focus on what is unique about each one. But with respect to the role that technology plays in each, a role often cited as a primary cause of the problems, the same screenome will be broadly applicable.
We have already begun to demonstrate diverse applications of the screenome. In politics, we found that the screenome shows close links between personal messaging and interpretations of news (Muise, et al., 2017). In medicine, we showed that the screenome contains the presence of drug and disease-related signals for diabetes (Gijsen, et al., 2019). A comparison of screenomes gathered in the US, China and Myanmar showed that the number of smartphone sessions differed substantially across countries while the structure of individual sessions was quite similar (Muise, et al., 2019). And in comparisons of adolescent screenomes we showed extremely quick switching among highly varied content and idiosyncratic preoccupation with specialized content, patterns that have substantial implications for health, development and well-being (Ram, et al., 2019).
Each problem domain is unique. But each of the examples used the exact same data, tools and analytics to define pieces of the screenome that were most relevant to each challenge. Consequently, we believe that the value of change for the Human Screenome Project will be the sum across rather than within the social challenges. And the sum – whether calculated by number of people affected, monetary costs or lost opportunities to individuals and society – is substantial. Policy advocates are clear that the problems are perilous (e.g., in the cases of psychological wellbeing, altered attention spans or tobacco advertising) or even existential (e.g., in the cases of media threats to democracy, formulation of climate change opinions or adolescent brain development).
There is a major gap in scientific discovery, however, that jeopardizes the value of knowledge about media effects. Research and policy is based primarily on simple survey measures of screen time that inaccurately represent a complex digital world. New systems are needed that can record the details of digital life and facilitate exploration of the complete record of screen use. The full potential of the screenome to solve social problems will be realized when we can simplify the process to (1) identify portions of the screenome relevant to each problem, (2) describe complex embedded patterns, and (3) deliver compelling summaries of the data to change agents via data dashboards for individuals, researchers, and policy makers.