Aims: The aim of this symposium is to inform the future design and
implementation of digital interventions by considering analyses of how people use them. The
symposium will consider how user characteristics such as education and health literacy level
may affect attitudes, usage and outcomes. The symposium also illustrates novel techniques of
carrying out detailed usage analyses that can reveal which elements of digital interventions
are and are not used and how this relates to outcomes of the intervention. Rationale: Usage
analyses are vital for understanding engagement with digital interventions – in particular for
investigating who are the people that dropout and why. They can also be valuable for
identifying the effective ingredients of behaviour change interventions, as well as who
benefits most from which ingredient. Summary: Morrison commences by demonstrating how
visualisation techniques can be used to explore and describe patterns of digital intervention
usage in three different trials of a weight management intervention (POWeR), examining usage
with and without human support, and with and without access to both website and app-based
content. Reinwand also looks across multiple interventions, analysing factors predicting
dropout across a range of interventions. Interestingly, her paper reports that people with a
lower educational level appreciate digital interventions but nevertheless tend to drop out from
them. Muller examines how people with differing levels of health literacy use and benefit from
an interactive digital intervention, compared with traditional static text. Her paper analyses
the effects on health literacy outcomes of behavioural change techniques such as motivational
quizzes and planning. Yardley then describes the factors mediating and moderating outcomes of a
hand-washing intervention, exploring which elements were used and which predicted attitudes,
intentions and behaviour. Finally, Crutzen will lead a discussion of what lessons for
intervention design and implementation we can learn from these analyses.
