How Much is Too Much? Leveraging Existing and Emerging Large-Scale Social Data to Build Robust Evidence-Based Policy for Children in the Digital Age

在过去的十年里，英国儿童和青少年花在数字技术上的时间增加了一倍多，关于过度使用屏幕可能产生的影响的辩论在科学和政策领域以及公共媒体上都很突出。英国下议院、DCMS、科学技术特别委员会、皇家儿科和儿童健康学会以及英国首席医疗官最近调查了互联网、智能手机、数字游戏和社交媒体是否会影响儿童和青少年的健康。然而，这些倡议都承认这一领域缺乏科学知识——这阻碍了英国政策制定者制定有效建议、法规和干预措施的努力。该项目旨在首次使用现有的ESRC数据集来生成该领域制定政策所需的科学。我们的目标是为政策制定者、家长、教师和全科医生提供必要的证据，以了解数字技术在英国儿童生活中的作用，并强调潜在的风险和恢复力因素，这些因素可能是未来干预的重点。我们将使用ESRC数据资产，先进的统计方法和强大的开放科学方法来回答三个紧迫的研究问题：1。哪些风险因素和恢复力因素使青少年容易经历数字技术使用对其心理健康的影响？2. 数字技术使用与心理健康之间的定向联系是什么？确定的风险因素是否在其中发挥中介作用？3. 联系风险因素、数字技术使用和心理健康的因果途径是什么，可以为未来的干预提供信息？由于数据质量差和方法上的缺陷限制了对技术效应的研究，这些问题的答案目前是难以捉摸的。我们将利用我们在大规模社交数据集方面的丰富经验来研究数字技术的一般影响和更具体的技术影响（例如社交媒体和游戏）。我们将使用机器学习、网络建模和先进的纵向方法来确定潜在的风险和弹性因素（例如社会支持、经济剥夺），这些因素会改变儿童对数字技术的反应，并有助于指导未来的技术政策。这将创建不同的儿童档案，我们可以利用这些资料来调查数字技术的长期使用和影响，确定哪些可能的技术影响（例如社会隔离）目前尚未得到证实和过度炒作，哪些（例如睡眠质量差）值得仔细研究。我们将彻底记录我们的方法，并公布我们发现背后的每个详细分析。这种透明度在当前的科学和政策辩论中几乎完全缺失。免费提供我们的代码将使学术界、慈善部门和政府的其他研究团队能够直接建立在我们的工作基础上，促进一种迄今为止一直缺失的新型增量知识转移，并为包括慈善机构在内的所有部门提供一个模板，使其能够更经常地使用ESRC数据。至关重要的是，这项工作将告知我们的团队如何继续为英国卫生部正在制定的指导草案、英国互联网安全委员会正在进行的审议、联合国儿童基金会、经合组织、ParentZone等社会企业以及包括Barnardo在内的儿童慈善机构提供建议，我们已经与这些机构建立了牢固的关系。我们的工作还将为理解虚拟现实和人工智能等新技术的影响提供一个视角，这些新技术将很快通过ESRC数据进行测量。虽然虚拟现实和人工智能仍然处于大多数学术和政策辩论的雷达之下，但它们将在未来几年越来越多地进入话语。

The amount of time British children and adolescents spend on digital technologies has more than doubled in the past decade, and debates about the possible impact of excessive screen use are prominent in the scientific and policy arenas, as well as in the public press. The UK House of Commons, DCMS, and Science and Technology Select Committees, the Royal Society for Paediatrics and Child Health, and the UK's Chief Medical Officer have recently investigated whether the Internet, smartphones, digital games, and social media influence the health of children and young people. These initiatives, however, have all acknowledged the lack of scientific knowledge in this area-which stymies the efforts of UK policy makers to craft effective advice, regulations, and interventions. This project aims-for the first time-to use existing ESRC datasets to generate the science required to ground policy in this area. We aim to provide policy-makers, parents, teachers, and GPs with the evidence required to understand the role digital technologies play in the lives of British children, and to highlight potential risk and resilience factors that could be the focus of future interventions. We will use ESRC data assets, advanced statistical approaches, and robust open science methodologies to answer three pressing research questions:1. What risk and resilience factors predispose adolescents to experiencing an effect of digital technology use on their psychological well-being? 2. What are the directional links between digital technology use and psychological well-being, and do the risk factors identified play a mediating role in this? 3. What are the causal pathways linking risk factors, digital technology use and psychological well-being that can inform future intervention? Answers to these questions are currently elusive, due to the poor data quality and methodological shortcomings that restrain research on technology effects. We will leverage our extensive experience working with large-scale social datasets to examine the general effects of digital technologies and more technology-specific effects (e.g. social media and gaming). We will use machine learning, network modelling, and advanced longitudinal approaches to pinpoint potential risk and resilience factors (e.g. social support, economic deprivation) that alter children's reactions to digital technologies, and which could help guide future technology policy. This will create different profiles of children that we can use to investigate the uses and effects of digital technologies over the longer-term-determining which possible technology effects (e.g. social isolation) are currently unevidenced and over-hyped, and which (e.g. poor sleep) deserve a closer look.We will thoroughly document our methodology and publish each detailed analysis behind our findings. This transparency is almost entirely missing from the current scientific and policy debate. Making our code freely available will enable other research teams in academia, the charity sector and government to build directly on our work, facilitating a new kind of incremental knowledge transfer which has hitherto been missing, and providing a template for empowering all sectors, including charities, to make more regular use of ESRC data. Crucially, this work will inform how our team continues to advise draft guidance under development at the UK Ministry of Health, ongoing deliberations at the UK Council of Internet Safety, UNICEF, OECD, social enterprises like the ParentZone, and children's charities including Barnardo's-with whom we already have a strong relationship. Our work will also provide a lens to understand the effects of novel technologies such as VR and AI, which will soon be measured by ESRC data. While VR and AI are still under the radar of most academic and policy debate, they will increasingly enter the discourse in coming years.