JMIR Publications

ABSTRACT

Background:

Young adults are the leading users of self-managed dietary reporting applications. However, scant research has examined the user experience of different measurement approaches for mobile dietary reporting applicatioins when dealing with a wide variety of food shapes and container sizes that reflect everyday life conditions.

Objective:

Field user experience testing was conducted under actual meal conditions to assess the accuracy, efficiency, subjective reaction of three food portion measurement approaches embedded in a developed mobile app. Each of the three approaches featured a unique user interaction design. Key-in-based assessment (KBA), commonly used in many current apps, relies on the user’s ability to accurately and consistently estimate serving sizes or weights. Photo-based assessment (PBA) extends traditional assessment methods using life size visual aids, allowing users to scroll, observe and select a representative image from a set of reduced size options. Gesture-based assessment (GBA) is a new experimental approach developed based on user finger movements to describe rough food boundaries accompanied with a background reference.

Methods:

A group of 124 young adults aged 19-26 were recruited for a head-to-head randomized comparison, and divided into three equal groups: a KBA (n=41) control group, and PBA (n=42) and GBA (n=41) experimental groups. Three meals (breakfast, lunch, and dinner) were served in a university cafeteria. Participants were provided with twenty-five dishes and beverages with a variety of food shapes and containers that reflect everyday life conditions. The accuracy and time spent for realistic interaction during food portion estimation, and subjective reaction of each aid were recorded and analyzed.

Results:

Among the twenty-five food dishes and drinks provided in the cafeteria, participants (n=41) in the KBA provided the highest accuracy for round-shaped pieces, such as hash brown weight (p=0.0039), and also outperformed PBA or GBA for many soft drinks. PBA (n=42) had the best results for cylindrical hot dog (p<0.000), irregular shaped pork chop (p=0.0008), and beverages in cups, such as green tea (660ml, p=0.0005), while GBA (n=41) outperformed PBA for most drinks and KBA for some vegetables. The GBA group spent significantly more time assessing food items than the KBA and PBA. For each approach, the overall subjective reaction based on the score of System Usability Scale (SUS) was not significantly different. Moreover, the score of KBA and PBA surpassed the benchmark and were not significantly different.

Conclusions:

Experimental results show that each portion size aid had some distinguishing advantages for some food shapes and containers. As for user acceptance, participants considered the three aids to be useable, with respective pros and cons. Furthermore, users’ subjective opinions on measurement accuracy contradicted empirical findings. Future work will consider the use of each aid based on food or container shape, and integrate the advantages of the three aids for better results. Our findings on the use of the portion size aid are based on realistic and diverse food items, providing useful reference for future application development and improvement of an effective, evidence-based feature, and acceptable feature in mHealth applications. Clinical Trial: http://www.isrctn.com/ISRCTN36710750