Abstract
For individuals trying to lose or maintain weight, self-tracking their weight, diet, or exercise is important. In the past, different tracking modes have been examined, like paper and pencil, memory, or personal digital assistants. But the recent advancement and adoption of mobile technologies could also result in easier and simpler self-tracking. However, little is known about self-trackers, their tracking modes, and the absolute or relative contribution of each tracking mode at the population level. This study fills this gap by (a) comparing self-trackers’ characteristics across tracking modes and against nontrackers and (b) testing the relationship between mobile self-tracking and tracking outcomes using a representative sample of data from the Pew Internet and American Life Project from 2012. Controls in the model include demographics, technology use, and health indicators. Results suggest that mobile self-trackers are younger and more educated and that mobile self-tracking is a positive contributor and the best tracking mode.
Notes
1 Overweight is defined as a body mass index greater than or equal to 25; for obese, it is greater than or equal to 30 (World Health Organization, Citation2016).
2 There has been a lot of buzz about the quantified-self movement (Feiler, Citation2014), which maintains a website, holds local meetings and a yearly conference (see http://quantifiedself.com), and counts on a group of steadfast devotees (Smarr, Citation2012).
3 Even though these are low response rates, they are within the acceptable norm (Pew Research Center, Citation2016).
4 The most used measures for frequency of tracking in the literature are the number of diaries completed (for paper-and-pencil tracking modes), the frequency of log-ins, or the frequency of self-reported weight (Burke, Wang, et al., Citation2011).
5 What might be at stake here is the complexity of the behavior. For complex or tedious behaviors (e.g., specific caloric intake or concrete physical activity), devices that track those behaviors are far more reliable than self-reports (Lichtman et al., Citation1992). However, for less demanding behaviors (e.g., self-reports with a broader framework for the frequency, like daily, twice a day, or every week), the correlation between self-reports and actual behavior is significantly higher (Boase & Ling, Citation2013).
6 This survey was administered before the Affordable Care Act, also known as Obamacare, had launched. Therefore, this item followed the format of previous surveys and did not inquire about additional purchase options for health coverage.