1:30 - 1:50 pmSunday, September 29
LK 130
Simplifying mobile phone food diaries with food indexes
LK 130
Simplifying mobile phone food diaries with food indexes
Scientist, ARO
Background Mobile phone-food diaries could be useful tools against the growing obesity crisis. However, one of the challenges of keeping a digital food diary is related to finding specific foods in a database.... Read more

Description

Background

Mobile phone-food diaries could be useful tools against the growing obesity crisis. However, one of the challenges of keeping a digital food diary is related to finding specific foods in a database. The work we present here begins to address a solution of designing food diaries that do not depend on databases for self-monitoring. We investigate the viability, appropriateness and value of food indexes to improve the food logging experience on mobile phones.

Method

We identified two food indexes that varied in complexity. One was the Healthy Eating Index-2005 (HEI), which consists of 12 components that reflect both food groups and nutrients. The other was the Food-Based Quality Index (FBQI), which consists of 7 food group components. We designed and implemented a mobile-phone food diary reflecting each of these food indexes. We compared usability measures of each of these food diaries to usability measures of a traditional, database-based food diary (TRAD). 12 participants entered given food items into all three mobile phone food diaries. We report on time to create an entry, correctness of entry, self-reported workload measures for each food diary, and overall preference.

Results

Time to create entries was significantly different with each interface (F2,10 = 27.49, p<0.01). All pairwise comparisons are significant (Holm’s sequential Bonferroni procedure, α = 0.05). The FBQI  food diary was significantly faster than the HEI food diary, which was significantly faster than the TRAD food diary. Two TLX workload measures showed significant differences: Mental Demand and Success. A Friedman test was significant for Mental Demand (χ2=10.56, N=12, df=2, p<0.01); follow-up showed no significant pair-wise differences. A Friedman test for Success was significant (χ2=16.89, N=12, df=2, p<0.01). Follow-up showed that HEI is significantly lower than TRAD (z=-2.831, p<0.01) and FBQI is significantly lower than TRAD (z=-2.825, p<0.01), but no significant difference in self-reported Success between FBQI and HEI. Finally, a χ2 test of proportions showed that significantly more participants preferred HEI and TRAD; no participants preferred FBQI.

Conclusions

Our investigation found that as the food diary interfaces became less detailed but required more mental activity to create a food entry, time to create the entry still decreased. This indicates that when considering food diary designs, designers can offload more mental activity onto the user rather than force the user to capture it in the food diary. This is particularly important when considering mobile phone food diaries, which are typically used on the go and in time- and attention-constrained situations. Correctness of the entry is impacted by the complexity of the food index, or amount of mental activity necessary to create the entry, but the complexity is preferred by users when compared to an overly-simplified approach. Users also liked the traditional food diary approach, but qualitative feedback revealed this was due to the certainty they experienced when creating entries. This was a weakness of the current study: all food entries existed in the traditional food diary database, so participants never experienced a more typical failure to find a desired food in the database.

Contact Us

We're not around right now. But you can send us an email and we'll get back to you, asap.

Not readable? Change text. captcha txt

Start typing and press Enter to search