Cultivating Motivation through Data Visualization 

In the first phase, I reviewed previous data visualization design strategies for health and fitness mobile apps and conducted a design strategy analysis in the form  of affinity clustering. The mobile apps were chosen based on the frame below. I collected data visualizations from the selected apps (n=95) by taking screenshots. These screenshots were materials used in this analysis.

• Apps are from "Popular Apps" in the iTunes App Store (iOS platform) and from "Recommended for You" in the Google Play (Android platform), under the category of "Health & Fitness".

• According to the rating distribution of each category in the App Store, the average rating for the "Health & Fitness" category is around four (Colin Eberhart 2014). To obtain "quality" apps for analysis, only apps with a rating higher than four and with more than 1000 reviews were considered.

• Target apps should be free of budget concerns.

• Target apps do not require other devices to start to use their core functions.

• The language used in the target apps is in English.

To categorize the screenshots, I adopted the data visualization categories from The Graphic Continuum Card Set (Ribecca & Schwabish 2018). By the end of this analysis, a list of common data visualization design strategies under four categories with several subcategories has been defined.

In the second phase, using the predefined design strategies from the affinity clustering, I performed  semi-structured interviews to uncover how those strategies could affect the motivation of doing exercise at different stages of exercise adoption. Through a screening survey, interview participants (n=33 out of approximately  150) were recruited from three undergrad classes at North Carolina State University in Fall semester 2019 and were sorted into 3 groups: Preintenders (n=8), Intenders (n=12), and Actors (n=13). During the interview, participants were asked to arrange the order of the predefined design strategies in terms of their motivational affordance. The interview questions focused on:

• Participants' interpretations of each data visualization design strategy from the screenshots;

• User-centered values of each data visualization design strategy;

• Participants' preferences of those pre-defined design strategies in terms of motivating them to exercise;

• Participants' preferred, desired, and concerned design features.

Preliminary Findings

Design strategies that work best for each of the groups are specified based on the reflection from the interview data and reflection notes. There are 23 emergent concepts under 8 categories generated. Preliminary findings from the interview suggest that it is necessary to design for stage-specific mindsets. For example, Preintenders need to learn positive information from the data, Intenders benefit from comparing with their own goals, and Actors need more detailed information to keep them on track. For the deliverables, design suggestions/guidelines will be formatted as posters and implement scenarios.

Press & Presentation

• Design4Health (D4H, 2017) paper accepted.

• The 9th International Conference on Design Computing and Cognition (DCC'20) poster presentation.