HomeSense: digital sensors for social research

What actually happens within households? We know that men are increasingly sharing in domestic duties and parenting; but does this mean that these activities are being done with their partners or are they taking turns? Do families eat together and talk to each other, or do they have separate meals in different rooms while talking on social media to their friends? It is hard to observe households, and research on these issues is done through self-reporting, with people answering questions and filling in diaries, or with highly invasive methods such as video recording. There is another way.

Digital devices are becoming more sophisticated. A modern mobile phone can measure position and movement, as well as what the phone is being used for. Many people wear sensors for heart rate, sleeping patterns, and physical activity. And fixed sensors in houses can be simply plugged in to measure sound and energy use. Using such sensors effectively would reduce the need for questionnaires and interviews, reducing the amount of work for respondents and providing potentially more accurate reporting.

However, there are technical problems to be solved. What can be measured by these devices? How can the data be converted into meaningful descriptions of activities? How reliable are these descriptions? There are also ethical concerns. How can the datasets be securely stored and for how long? How does consent work if people forget the devices are there? When should consent be obtained from people who are monitored but not intentionally included in the research, such as visitors?

This project will examine these technical and ethical issues. We will develop guidelines for social researchers who want to use digital sensing devices in their research. These will be based on expert advice and discussion with members of the general public, as well as the experience of household members and researchers in a trial study.
The data collected in the trial study will be used to compare, contrast and integrate the use of sensor devices with existing research methods. The trial data and comparison of methods will be the foundation to develop analysis tools that help researchers to interpret and understand the rich data that can be collected with these methods, to answer questions about what happens within households.

The project relates directly to methodology development for digital sensing devices but addresses much wider ESRC and NCRM commitments.

First, digital sensing devices can be used to support cohort and longitudinal studies by easing or augmenting data collection. Longitudinal data collections focus on the interplay of structure (such as demography, geography or social position) and social outcomes (such as health or employment), and these can be supplemented by detailed data about micro processes to provide an understanding how and why these correlations come about.

Secondly, digital sensor data is able to contribute to the ESRC's priority area of Influencing Behaviour and Informing Interventions. Through unprecedented access to a variety of aspects of home life, digital data will provide a unique insight into the routines, practices and habits which are basic building blocks of everyday life. The understanding of the habitual linking of activities, the coping strategies employed and the interaction patterns of people in households obtained through the methodological integration of digital device data can provide a foundation for interventions in particular with regards to energy consumption, health and childcare, and also less obvious issues such as barriers to employment and wellbeing.

The project will develop algorithms and software tools to visualise and understand data generated by digital sensing devices, particularly as it relates to activity sequences, location and interaction. These algorithms may have direct economic and social impact, particularly for the Quantified Self movement and for companies marketing home automation systems. The sale of personal monitoring devices and data analysis subscription services are small, but significant and growing, industries. Richer analysis strengthens these industries and also increases the potential health benefit of such monitoring, as better monitoring and diagnostic tools are developed that use the data. As sensors are introduced into homes as part of the spread of the 'internet of things', companies selling such devices may benefit from the work we will do on issues of privacy, security, consent and ethics.