TrackMe: A Mobile Application for User-Created Self-Report Measures

Self-tracking is acknowledged as one of key components for people’s successful behavior change. Self-tracking helps people to be aware of their situations and motivated toward attaining goals. People’s self-tracking often involves self-reporting to some degree in collecting data. Although drawbacks of it exist including reporting burden and potential inaccuracy and thus automatic data collection using sensor technology is often recommended as an alternative to be pursued, self-reporting is considered to have considerable benefits. Self-reporting allows people to track virtually any aspects of their daily life with no or little cost. Self-reporting is an essential approach in measuring person’s subjective states and experience. It is also considered enhancing people’s awareness, self-reflection, and engagement, which may be reduced with automatic data collection. Acknowledging the benefit of self-tracking and the needs of manual input, I designed an application, ‘TrackMe’, that allows users to create their own self-reporting measures by providing essential elements for capturing and reflecting various aspects of daily life.

Users of TrackMe do, in general, the following three tasks: (1) Create a measure, (2) Log data, and (3) View collected data. In creating a measure, users can choose one of the five data types: number, timestamp, time duration, list, scale. I attempted to pick the most indispensable ones. Among other possible types, the binary type was not included (e.g., ‘yes/no’) as users may satisfy their purposes using the number (e.g., enter ‘0’ for no, ‘1’ for yes) or list type (e.g., ‘yes’ and ‘no’ choices). The order or rank type was not included, as it seems rarely involved in people’s self-tracking, and as well can be covered with the current types if necessary. Collected data can be viewed in two different granularities: individual entries and a week/month chart. For number and time duration data, daily total, averaged, minimum, or maximum values are drawn in a bar chart. For list and timestamp data, users can check frequency of each option (for list data) or time window (for timestamp data) chosen. Scale data can be viewed in the both ways.

Please check out a wireframe prototype.

Meanwhile, this interest in supporting users’ own creation of self-reporting measures has extended to the concept of knowledge-based design support for users’ more reliable data collection, and our team’s currently initial idea will be presented and explored in the CHI 2017 workshop, ‘Digital Health & Self-Experimentation: Design Challenges and Provocations’. As described in our workshop paper, we plan to conduct the initial formative study with key stakeholders including novice and extreme self-experimenters. We intend to explore:

  1. What is the current process that self-trackers use when creating their own self-report measures?
  2. What difficulties do self-trackers have in creating measures, logging, and reviewing collected data?
  3. How will they respond to our concept of having knowledge-based design support?

Based on results from this formative study, we plan to design and evaluate a proof-of-concept prototype.