STRIKE will identify and scope the data infrastructure needed to build a national picture of social interactions in the UK, building on recent developments in the online collection of time-use data, applying methodological innovation to collect information on online and face-to-face social contacts and the places and events where interactions occur.

The team

Objectives

The overarching aim of the project is to triangulate the feasibility, usability and acceptability of social connectivity data across partners, stakeholders and participants to specify a workable, useful, scalable and cost-e|ective data collection infrastructure.

Objective 1: Determine data user needs and gaps in social connectivity data

Deliverables:

• Summary review of existing social connectivity data across disciplines
• Map 1: mapping existing social connectivity data against stakeholder needs
• Data collection specications to ll existing gaps including methodological, ethical, and practical challenges

Objective 2: Understand public acceptability of social connectivity data collection

Deliverables:

• Summary of findings on public acceptability including factors that positively or negatively in€uence people’s decisions to participate/share their data
• Map 2: Calibrating Map 1 with acceptable solutions.

Objective 3: Specify data collection infrastructure including scaling up and integration into existing surveys and studies.

Deliverables:

• Final specication for data collection and outputs, including methods, denitions and data requirements, scalability (including ball-park costs), and suggested next steps.
• Map 3: Integrated Map of connectivity data infrastructure.

Project plan

STRIKE consists of 6 stages. Methodologically it is a combination of desk-based/review and participatory research with data users, providers and participants.

Stage 1 Months 1-3: Existing data review

Review of existing data available on social contacts, interactions and relationships including definitions and the methodologies used to collect these data.

This will include UK and international datasets. The latter will be reviewed to identify the kinds of data other countries collect and in what ways, and how UK data might be aligned with other countries’ to support cross-national research. Datasets we are already aware of include epidemiological data from the POLYMOD and CoMIX surveys, data on social capital collected on a number of government surveys, such as the Community Life Survey (repeat cross-sectional); Understanding Society (longitudinal); and Scottish Household Survey (repeat cross-sectional); the University of Dundee’s Multimodal Focused Interaction dataset, the ONS and UK (http://k/) Time Use surveys, and the Multinational Time Use Study.

The review will identify the mode(s) of data collection used in these surveys and alternative data collection methods, such as the use of sensors and apps, and the linking of different kinds of data, such as individual’s survey responses with their social media accounts (e.g. Stier et al, 2019, Kreuter et al 2020, Sloan et al, 2020, Hadler et al, 2022) to learn more about people’s online interactions (e.g. Bian, 2020).

Stage 2 Months 2-4: Data user needs

Consultation with stakeholders about the use and value of the data.

We will hold a workshop and individual meetings with potential users of the data including our project partners (O}ce for National Statistics (ONS) and the Behavioural Insights Team (BIT) and a wider stakeholder group involving representatives from the O}ce for Health Improvement and Disparities (OHID), the UK Health Security Agency (UKHSA), Public Health Scotland, Public Health Wales, Public Health Agency (Northern Ireland), UK BioBank, the What Works Centre for Ageing Better, the Scottish and Welsh Governments and the Northern Ireland Statistics and Research Agency (NISRA) to discuss potential uses of social connectivity data and the kinds of research questions such data could support. Data requirements will be unpacked, exploring issues such as the kinds of social contacts and connections to be captured; from whom e.g. adults, children, households; geography e.g., UK residents, interactions taking place in the UK or involving UK residents; over what time period; whether there is a need for cross sectional and/or longitudinal data; data granularity required e.g. timing, length/intensity, type of interaction (physical, non-physical), level of information needed about those involved in the interaction (e.g. age, gender, ethnicity), mode of interaction (in-person, telephone, online, post, text, video).

The workshop will be held online with up to 24 participants. We will then carry out 8 follow up interviews with participants.

Stage 3 Months 1-6: Scoping challenges and solutions

Gathering information on social contacts and connections poses ethical, methodological and practical challenges. STRIKE will identify these challenges, drawing on the experiences of members of the research team with studies such as CoMIX and HomeSense (Jiang et al 2019), the experiences of our partners and stakeholders, and the research literature. One important set of challenges relate to response burden. Studies such as POLYMOD and CoMIX involved members of the public being asked to recall occasions when they had contact with others over a specified recall period or keeping a diary of contacts. Diary methods in particular are known to be cognitively burdensome and prone to measurement errors related to diary fatigue (e.g., Dillman and House, 2013; Crossley and Winter, 2015). STRIKE will consider the burden on participants, particularly around recording habitual interactions such as those between people who live and/or work together. This will include looking at options to combine interaction data from different sources and to use data from sensors and apps, as alternatives to or in combination with survey data.

Another important challenge relates to minimising non-response bias. As survey response rates decline (Groves et al, 2009) efforts have been made to link different sources of information on individuals to address nonresponse and expand the data available for analysis (e.g. Sakshaug et al, 2012; Al Baghal, 2016; Mostafa and Wiggins, 2018). Linking data involves seeking the consent of research participants, with consent being dependent on context i.e., what information is to be shared with whom (e.g., Sala et al, 2012; Saskhaug et al, 2012; Al Baghal, 2016). STRIKE will identify challenges with linking data, e.g. social media and survey data, and current best practices for asking for consent to link data.

At the end of this stage (Month 6) we will have a mid-project partner and stakeholder workshop in London to consolidate findings.

Stage 4 Months 4-8: Public acceptability

Assessing the extent to which data needs can be met.

This stage will consider what information participants would be willing to share, with whom, and for what purposes. This will involve qualitative work (workshops) with members of the public, including those from groups less likely to be included in research, such as those from minority ethnic groups, young people and children, and those at risk of digital exclusion (UKSA, 2021). The workshops will explore factors that may impact willingness to participate such as the nature of the information being collected and how it is collected e.g. answering survey questions or installing an app on their smartphone.

We will hold 6 workshops (3 online, 3 in-person) in different parts of the UK, Scotland, north of England, and London, each with 10 members of the public. Workshop data will be analysed to identify design features that may limit participation and the types of people who may be more or less willing to consent to different data collection requests. Eight follow up interviews will be used to explore in more detail the factors influencing participation.

Findings from this stage of work will be compared to the data needs expressed by stakeholders (stage 2), to assess the extent to which it may be possible to collect data on the topics of interest, to the level of granularity required.

Stage 5 Months 9-11: Data priorities

Stakeholder workshop to identify data priorities.

A second online workshop with stakeholders who contributed to stage 2 will take place after stages 3 and 4 are complete, to share and discuss the findings  from these stages, with the aim of identifying data priorities that reflect the feasibility of collecting relevant connectivity data.

Stage 6 Months 11-12: Data Infrastructure Specification

Development of a specification for data collection infrastructure.

Based on the outputs from stages 1-5, a specification of data collection infrastructure will be developed and a final report will be produced, bringing together findings  from stages 1-5. It will include discussion of issues relating to data privacy, ethics, data deposit and documentation, and scalability. The report will also propose next steps, setting out a workplan for development and implementation of a data collection infrastructure, including estimates of costs.

STRIKE is a project led by the University of Glasgow with the University of Surrey, the London School of Hygiene and Tropical Medicine (LSHTM), and the National Centre for Social Research (NatCen), and with the Office of National Statistics and the Behavioural Insights Team as partners.