Session: Innovative Nutrition Education Interventions Across Populations (Oral 21)
(OR21-01-24) Understanding the Learnability of a Novel Mobile App for Personalised Food Guidance Leveraging Persuasive Technology and Computer Vision: A Mixed-Methods Study
Research Fellow University of Wollongong Wollongong, New South Wales, Australia
Disclosure(s): No relevant financial relationship(s) with ineligible companies to disclose.
Objectives: With the rapid pace of digital technology innovation, understanding the learnability of digital dietary interventions plays a pivotal role in facilitating the adoption and use of digital dietary tools for eating behaviour changing. The aim of this study was to explore the learnability of a novel mobile app for personalised food guidance.
Methods: A progressive web app prototype was collaboratively developed with end-users and Accredited Practising Dietitians applying the design science paradigm. The food layer of the Nutrition Care Process, the strategies from cognitive behavioural theory and the Australian Dietary Guidelines were translated into the prototype features guided by the Persuasive Systems Design model. User-driven, rule-based reminders based on the Elaboration Likelihood Model of persuasion was created to support users in achieving their goals for food choices. Expert cognitive walkthroughs were conducted to evaluate the learnability of the prototype (n=4). Convenience sampling from adults in Australia was asked to use the prototype for seven consecutive days (n=9). A survey using the Mobile Application Rating Scale and semi-structured in-depth interviews were performed to examine the usability of the prototype.
Results: Although both the experts and end-users identified tracking food choices as having the highest number of learnability problems, end-users reported that this feature was the main driver for improving their eating behaviour. The experts identified that the provision of interactive feedback for each action at the interface level could improve the human interaction design to enhance learnability. The end-user's previous experience with the nutrition app influenced their behaviour in using the prototype. The end-user suggested that the benefits of using artificial intelligence in nutrition apps outweigh any potential harms, particularly in terms of improving their accessibility and inclusiveness.
Conclusions: Providing interactive feedback is crucial for improving the learnability of novel digital dietary interventions. A novel approach is required to engage end-users with digital dietary tools and facilitate their learning for eating behaviour change.
Funding Sources: Illawarra Health and Medical Research Institute Young Investigator Award and Multiple Sclerosis Australia Postdoctoral Fellowship
