Biases in Physical Activity Tracking

Project description

Personal informatics and the quantified self in general and physical activity tracking is particular are all about collecting, tracking and reflecting on personal data. Assessing and reflecting on personal data is a key stage, which is recognised in personal informatics frameworks [1], for continuous engagement with the tracking system and to facilitate behaviour change. However, this stage involves uncertainty that comes from many sources (incomplete data, information overload), hindering the reflection stage [2]. The fact that data is personal and the tracking process involves uncertainty indicates that sense-making and decision-making upon self-tracking data are prone to systematic errors and may involve cognitive biases. For instance:

- Anchoring bias is exhibited when there individuals rely on the first piece of information encountered (the anchor), which affects the judgements that follow. In self-tracking is manifested for example when a self-tracking tool recommends a daily goal of 10,000 steps or 30 minutes of exercise. These anchors become reference points to assess progress and success, preventing users from establishing their own baselines.

- Confirmation bias is the natural propensity toward maintaining existing beliefs. Self-trackers may select or interpret evidence that supports their beliefs.

But there may be more including illusory correlations, availability heuristic, status-quo bias etc. The goal of this PhD project is to:

1. Explore and systematically identify the cognitive biases that may exist in physical activity tracking. This will involve a literature review on physical activity tracking, cognitive biases and phycology literature on experimental designs to identify biases.
2. Run experiments to identify the biases and how they manifest in physical activity tracking.
3. Design interventions for debiasing purposes and evaluate them through field studies.

Consequently, the project would be appropriate for a candidate who is skilled in at least two of the following: software development, experimental design and data science. If one skill was missing, the candidate must show commitment to acquire it during the course of the project. Also, it is expected that the candidate will publish academic papers in SIGCHI venues (CHI, DIS, EICS), UMAP and Ubicomp/IMWUT.