Design, development and validation of wearable system to collect in-situ measurements of mood, anxiety and stress for children aged 6-12 years.
Dr Petr Slovak
Lecturer in Human Computer Interaction
HCC Group, Department of Informatics, King’s College London
Dr Johnny Downs, Clinical Senior Lecturer, Department of Child and Adolescent Psychiatry
Professor Edmund Sonuga-Barke, Professor of Developmental Psychology, Psychiatry & Neuroscience
Partners / Collaborators
Professor Katherine Isbister (UC Santa Cruz), Professor Max van Kleek (Oxford), Committee for Children
Innovative evaluation and intervention design methods (such as Micro-RCTs) can accelerate the pace and reduce the cost of intervention development, unpacking the effects of intervention components with better efficiency (e.g., lower samples and shorter studies) than traditional RCTs. Such approaches however require the ability to gather proximal outcomes with high granularity (e.g., hourly or daily), which makes them infeasible in child populations so far. In fact, current methods of collecting in-situ data from children and their families face substantial challenges of scale, and cost (e.g., requiring in-home visits by trained personnel for each data point) or low adherence (e.g., less than 8% response rates for paper questionnaires sent to families). Emerging IoT and wearable technologies have the potential to start addressing these fundamental methodological constraints; imagine, for example, a ecological-momentary assessment (EMA) to detect child immediate anxiety, stress, or mood, relying on a simple 5s interaction with a wearable watch. However, the research on the design requirements, applicability, and practical feasibility of such technologies in child populations is limited, with virtually no work so far done within the context of child mental health.
To develop efficient data collection system that could be widely deployed in-situ across both clinical and non-clinical settings will require addressing a range of technical, psychological, and socio-technical research challenges. These necessitate an interdisciplinary research approach that can combine: (i) user-centred design approaches from Human-Computer Interaction to envision novel data-collection platforms that will be easy to use, are developmentally appropriate, and fit into the everyday lives of children and their families; and (ii) an in-depth expertise from Child Psychiatry to develop/transfer reliable, valid, and clinically meaningful indicators of childrens mental states.
The goal of the PhD scholarship will be to design, develop, and validate the feasibility of a data collection system for in-situ measurements of mood, anxiety and stress, targeted at children 6-12 years old, combining the interdisciplinary expertise brought in by the three co-supervisors (HCI Slovak; Child Psychiatry Downs, Developmental Psychology Sonuga-Barke). Specifically, the PhD candidate is expected to follow a set of established HCI user-centred design techniques, to develop an open-source toolkit platform, capable of collecting a set of validated measures within a case-study context (emotion regulation), but also open to further adaptation, re-use, and modification by the broader research community.
- The candidate will start with a systematic review of current approaches to EMA and other in-situ data collection for children in our age range (drawing on pilot work to date by both co-supervisors).
- This will be followed by a series of user-centred design studies with children and their families (across both clinical and non-clinical sample), as well as psychiatry and psychometrics experts. The candidate will rely on established HCI methods, including co-design workshops, participatory design, and in-depth interviews to identify plausible interaction mechanisms and device types (e.g., smart watches, voice-based assistants such as Amazon Alexa or Google Home).
- The PhD work will then move to design and develop a number of plausible toolkit components (drawing on Double Diamond design processes and parallel prototyping techniques), which will be trialled in qualitative HCI deployments (focussing on acceptability, user experience, and design iteration).
- Finally, the candidate will identify and develop a final set of components (based on empirical data above), which will be validated during the NIHR funded REMAIN study in a cross comparison study (https://www.fundingawards.nihr.ac.uk/award/CS-2018-18-ST2-014)
The PhD work will be positioned within-and be able to draw on-ongoing projects of the two co-supervisors, which provide potential broader infrastructure into which the developed toolkit can be embedded: (1) myHealthE (Downs) provides an existing web-based monitoring system for caregiver and adolescent self-report data, incorporated in CAMHS; (2) Privopticon (Slovak) is based on a collaboration with the Personal Data and Privacy group at Oxford University, with a modular mesh of self-organising IoT components (Raspberry Pis) to enable privacy-preserving data collection in families; and finally (3) Smart Toys (Slovak) is an innovative, in-situ emotion regulation intervention for children delivered through bespoke socially-assistive robot infrastructure. Such interconnections will not only provide pathways for sustainability and impact for the developed toolkit (e.g., by embedding it into myHealthE and/or Privopticon infrastructures), but also bring a range of case study contexts, including on-going trials, to develop and validate the in-situ measurements (e.g., Smart Toys and myHealthE).
Human-Computer Interaction, Emotional Problems, ecological-momentary assessments, IoT Technologies, privacy-preserving methods