Behaviour change for when you can’t stop getting worse
By ucjubil, on 4 April 2017
By Dr Andrew McNeill, a researcher at PaCT Lab in Northumbria University
Doris vacantly stared at the Sudoku puzzle on her iPad. She used to love doing these puzzles; her daughter said they would help keep her mind active. It had worked for a while, but now the time it took her to complete them was getting longer. The system calculated a score based on the time taken to complete the puzzles and then told her every week how agile her mind was. At first, she was excited by her progress. But for the past few months, when she looked at the chart of her progress, all she saw was a steady decline. It was a visible reminder that she was getting old. She wondered if it was the onset of dementia.
Next door, George rubbed his knees tenderly. His arthritis was getting worse, and while he had been very active in the past, the pain was getting too much for him. A faint “ding” caught his attention. Glancing down at his phone, he read the message: “Hi George! You’re slowing down compared to this time last year! Why not go for a walk today and beat your high score?” He groaned. Why did it have to rub it in that he was getting worse?
The scenarios described above are fictional, but they make an important point. Digital behaviour change interventions have often assumed that their goal is to maintain or increase the health of their users, but in the case of older adults in decline, the paradigm can no longer hold true. In their case, the goal has to be to minimise the acceleration of decline while maintaining their self-efficacy as much as possible. To accomplish this, two concepts ought to be considered: privacy and identity.
Firstly, privacy. Privacy is often considered as the relationship between information about the self and the access of others to that information. But, in the context of health-monitoring systems, the system retains information about the self, separate from the user’s awareness. Does this open up a new aspect of privacy? We have argued that privacy can also involve the desire of a person to keep information about themselves hidden from both themselves and others (1). Health-monitoring systems thus have to be designed to consider the extent to which users want to know information about themselves. While they often find this information interesting, when decline is underway, they often do not want to see this data, particularly if the decline is irreversible.
This brings us to our second consideration: identity. If people often seek to avoid information about irreversible decline, this is often because it reinforces a negative future identity (3). People seek to maintain an identity that provides self-esteem and this being the case, feedback from a system about decline that is perceived by the user as irreversible can cause the user to lose self-esteem (2). To maintain their positive identity (both present and future), the user may ignore the information contained about them in the system because it is too threatening.
But how can privacy management and identity management be incorporated into a system for behaviour change? Perhaps we should allow active information seeking so that the user has to seek out their data to see if they are improving or declining? That way we might avoid the problem of confronting them with identity-threatening information. And perhaps we need to present the data differently – perhaps comparing trends within a shorter time period would allow us to estimate the upper threshold of the user’s performance and thus encourage them to achieve realistic targets rather than unrealistic ones from the previous year. That way, privacy management would involve the system hiding information from the user by presenting a smaller window of data. But the question is still open: How is it possible to present encouraging feedback to older adult users when they are in a period of irreversible decline?
References
- McNeill, A., Coventry, L., Pywell, J., & Briggs, P. (2017). Privacy Considerations when Designing Social Network Systems to Support Successful Ageing. In Proceedings of CHI 2017. ACM. http://doi.org/10.1145/3025453.3025861
- McNeill, A., Briggs, P., Pywell, J., & Coventry, L. (2017). Functional privacy concerns of older adults about pervasive health-monitoring systems. In Proceedings of PETRA 2017. ACM. http://doi.org/10.1145/3056540.3056559
- Vignoles, V. L., Manzi, C., Regalia, C., Jemmolo, S., & Scabini, E. (2008). Identity Motives Underlying Desired and Feared Possible Future Selves. Journal of Personality, 76(5), 1165–1200. http://doi.org/10.1111/j.1467-6494.2008.00518.x
BIO: Dr Andrew McNeill is a researcher at PaCT Lab in Northumbria University where he conducts research on the ACANTO Project. The project seeks to design a recommender system for older adults to encourage them to be more physically and socially active.
Acknowledgement
The project described here was funded by the European Union’s Horizon 2020 research and innovation programme under grant agreement No 643644 (ACANTO: A CyberphysicAL social NeTwOrk using robot friends).
Join the discussion about this post on our LinkedIn group.
Want to be a guest blogger? Please email Aiste for more information.