Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014 1429

Are you Committed Cathy, Reluctant Rita or Negative Nancy?

Defining User Personas for a Technology-Based
Wrist-Worn Eating Monitor

Michael L. Wilson, MS; James N. Salley, MS; and Eric R. Muth, PhD
Department of Psychology, Clemson University
Clemson, South Carolina, USA

Self-monitoring of energy intake is a critical element of a successful weight loss plan. However, current
methods to monitor energy intake are cumbersome and prone to under reporting. The present study
examined how individuals used a new energy intake monitoring tool, the Bite Counter to adjust their eating
behavior to a targeted bite limit. Data were collected from 30 female participants examining their
compliance with using the device as well as their adherence to eating limits based on bite count. Three
distinct compliance personas were developed based on the shared behaviors and traits of device users:
Committed Cathy (the rapid adopter, seldom misses tracking eating activities), Reluctant Rita (often forgets
device, always has an excuse) and Negative Nancy (will not wear or use the device). These personas will
inform future experimenters on how to improve usage instructions in order to increase participant
compliance with using technology-based eating behavior monitoring tools.

INTRODUCTION

The purpose of the present study was to analyze the usage
habits of Bite Counter wearers. The goal was to identify
common traits and behaviors in order to establish Bite Counter
user personas to inform future weight loss protocols.
The prevalence of obesity in the United States has
reached the point of becoming a national epidemic. The most
recent annual National Health and Nutrition Examination
Survey report shows that two-thirds of the population is
classified as either overweight or obese (Flegal, Carroll,
Ogden, & Curtin, 2010). Self-monitoring is a critical aspect of
both obesity research and successful behavioral treatment of
Copyright 2014 Human Factors and Ergonomics Society. DOI 10.1177/1541931214581298
This study used a new tool developed for the self-
monitoring of eating behavior called the Bite Counter (Dong,
Hoover, & Muth, 2009). The Bite Counter is a mobile health
tool for self-monitoring of energy intake. This wrist-worn
device monitors energy intake by counting the number of
times a person puts food or drink into their mouth, i.e., takes a
bite. The device simply has to be turned on at the start of an
eating activity and off at the end of an eating activity.
While eating, the Bite Counter displays bite count for the
current eating activity in real-time. Between meals, the device
has a user review button which when pressed will display the

obesity.
While energy (caloric) intake can be easily and
objectively monitored in a laboratory setting using direct
observation, monitoring energy intake in the field poses a
challenge. There are a variety of tools available to researchers
for the purpose of monitoring energy intake in a field setting,
each with its own strengths and weaknesses. The most
accurate method available, the Doubly Labeled Water method,
can measure energy intake to within 2% to 8%, but is
extremely costly and impractical for use in studies with a large
sample size (Schoeller, 1988; Black & Cole, 2000). Many
researchers fall back on one of three subjective self-
monitoring tools: the food frequency questionnaire, the 24-
hour dietary recall, and the food diary. However, all three
subjective methods suffer from under-reporting bias
(Livingstone & Black, 2003).
The consistent and frequent self-monitoring of energy
intake is also one of the most critical components of
behavioral interventions for the treatment of obesity (Burke et
al., 2009). Numerous studies have found that compliance with
the use of self-monitoring tools, such as food diaries,
contributes significantly to weight loss (Baker &
Kirschenbaum, 1998; Burke et al., 2009; Helsel et al., 2007).
bite count for the last eating activity and a total bite count for
the day. A time-stamped log of bite count data is stored in
memory for download to a PC to generate a calendar of eating
activities for longer term analysis of eating behavior. Hence,
the device provides data for real-time self-monitoring of
intake during a meal, daily intake self-monitoring, and long
term analysis of week-to-week and month-to-month eating
activities. This allows for behavior change to be targeted at the
single meal, e.g., a cue to stop eating before overeating, as
well as at longer term eating patterns, e.g., eliminating
overeating on weekends compared to weekdays.
The user profile and persona are conceptual models of
targeted user group(s) that serve to promote the shared
understanding that supports User Centered Design throughout
the process of analysis, design, development and
implementation (LeRouge, Ma, Sneha and Tolle, 2011).
LeRouge et al. (2011) further define a persona as a fictional
and super-typical characterization of a user created to
represent a user group. Personas have many benefits including
being able to facilitate effective communication about the
users, enhance identification with the target users and to
increase focus on a specific audience (Miaskiewicz & Kozar,
2011).
Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014
METHODS
Participants
Forty-eight female participants were recruited for this
study from a mid-sized southeastern university faculty, staff
and student population via email, fliers and word of mouth.
The study population was restricted to female participants
only because of a low respondent rate for males during pilot
testing and to have adequate statistical power for at least one
gender. Respondents were screened against the study inclusion
criteria of at least 9 kilograms overweight with no history of
eating disorders. Participants selected were classified as
overweight through obese (BMI between 27 and 50 kg/m2),
ranged in age from 18 to 65 and self-reported as motivated to
lose weight. Participants were paid $25 for their participation
in the study.
Materials
Bite Counter. The Bite
Counter detects a wrist-roll
motion that is characteristic
of taking a bite of food.
The version of the device
that was used for this study
resembles a watch, with a
digital display, an on/off
button, and a display cycle Figure 1. The commercially
button (see Figure 1). available Bite Counter
When bite count mode is
off, the device displays the time. When the device is in bite
count mode, it only displays the bite count for the current
eating activity.
Usability Questionnaire. Usability data were gathered via
experimenter generated questions, and included questions
about frequency and ease of use, e.g. “How easy or difficult
did you find it to use the Bite Counter?”
Procedure
Prior to beginning the study, participants completed an
online screening questionnaire. Participants were excluded if
they had a history of an eating disorder. Upon passing the
initial screening, participants were brought into a laboratory
setting where they completed the University Institution
Review Board approved consent form, given the instructions
for the study and issued a Bite Counter.
Participants were instructed on how to wear and use the
Bite Counter and asked to ensure that they counted all daily-
eating activities by activating the device each time they ate a
meal or a snack. Over the course of the ten-week study period
participants reported to the lab weekly where their weight was
recorded and Bite Counter data were downloaded.
1430
RESULTS
Study Attrition
Of the 48 participants originally enrolled in the study only
30 completed the ten-week usage trial, yielding a dropout rate
of 37.5%. Numerous reasons for attrition were offered
including; “I’m too busy”, “there is too much going on in my
life”, and “I can’t remember to use the device”.
Bite Counter Compliance
Weekly Bite Counter records were examined to identify
individuals who met a minimal compliance standard with
using the Bite Counter. Due to inconsistencies in weight loss
at the weekly level, weight loss was examined at the 5-week
interval. Five-week compliance standards were based on the
weekly compliance definitions developed based on accuracy
rates for other energy intake tracking tools, the food frequency
questionnaire, the 24-hour dietary recall, and the food diary.
Accuracy rates for these measures are reported to average
approximately 86% (Livingstone & Black, 2003). Individuals
were deemed to have met compliance standards if they missed
no more than one day per week of monitoring (6/7 days
(86%), and collected a minimum of 14 eating activities (2
eating activities/day-7days) using the Bite Counter during the
same week. Five-week compliance was defined by selecting
individuals with no more than 1 non-compliant week during
the 5-week evaluation period. Using these screening criteria,
19 of the 30 (63%) participants using the Bite Counter were
determined to meet compliance standards during their initial
5-weeks with the device.
Adherence to Bite Targets
Another area of interest examined during this study was
the ability of the Bite Counter to provide self-monitoring cues
that would result in stopping a participant from overeating. In
a recent meta-analysis of self-monitoring’s effect on weight
loss, Burke found that of the 15 studies he evaluated that
focused on dietary self-monitoring and weight loss, all found
significant associations between self-monitoring and weight
loss. In fact, the evidence in these studies showed that the
amount of monitoring and amount of weight loss were
strongly positively correlated (Burke, Wang, & Sevick, 2011).
In the current study, participants were given a meal-level
bite target value in the form of an audio alarm that beeped
when the bite target value was reached, and continued to beep
during each successive bite until the Bite Counter was
switched off. To quantify how often study participants
stopped eating when the alarm sounded, data from the
participants who were deemed compliant with using the
device (n=19) were examined. The total number of eating
activities that met or exceeded the alarm setting was counted
for each participant over the 5-week time period. Next, the
number of times that the participant stopped eating at or
within 3 bites after the alarm sounded (alarm value +3 -0
bites) were counted. This analysis revealed that on-average
Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014
participants stopped eating when the alarm sounded 65.3% of
the time.
Usability Data for the Bite Counter
Participants found the Bite Counter easy to use because
they only had to press a button to turn it on and off. Some
people liked that it was on the wrist, easily portable,
functioned as a watch and could be strapped to a lunch bag or
the refrigerator handle. Charging was described as similar to
charging a cell phone. They described using the device as
“easy”, a “no brainer”, “user friendly”, and that it “became
second nature”. The audible and visual feedback was helpful
for knowing when the device was turned on and off. Some
participants liked being asked about the device by friends and
coworkers so that they could tell them about their participation
in the study. Participants liked that it increased their awareness
of eating.
Overall, the most difficult aspect of the Bite Counter was
remembering to turn it on and off. Some participants had
difficulty remembering to wear the device. In terms of
physical discomfort and appearance, the device was described
as unattractive, uncomfortable, “too big”, “bulky”,
“cumbersome”, not “trendy”, and “ugly”. Some participants
found the Velcro to be irritating, and several participants
disliked having something on their wrist. Some participants
reported that their friends and coworkers often asked about the
device, but some participants disliked describing their “weird-
looking watch” to others. When asked how the device could
be improved, participants suggested a smaller device with a
curved back, a thinner non-Velcro wristband, optional
beeping, different colors, additional watch features like the
date and a stop watch, syncing to devices like the iPhone,
water-resistance, and automatic detection of eating.
DISCUSSION
The results of this study indicate that overall,
participants reported their use of the Bite Counter as a positive
experience. There were however, three main areas of concern;
attrition, compliance, and adherence (device usage within the
weight loss protocol). When examining the first issue, study
attrition, the dropout rate of 37.5% was higher than the 20%
dropout rate encountered during pilot testing. Second,
although the majority of users report that the Bite Counter was
easy to use, only 63% of participants were compliant in
consistently wearing and using the device. Finally, when given
a bite target cue to stop eating, users only stopped an average
of 65.3% of the time when the alarm sounded.
Device Usability
The majority (75%) reported the Bite Counter easy to use.
When asked about their preference between the Bite Counter
and a written food-log, 79% preferred the Bite Counter. Many
of the problems that participants listed with the Bite Counter
were largely aesthetic rather than with using the device itself.
The present study has a few limitations. First, there is no
comparable data available for studying compliance in wrist-
1431
worn eating monitors. Second, the usability data gathered was
largely self-report. Future studies should focus on objectively
comparing the usability of multiple self-monitoring tools
within a behavioral intervention.
The user acceptance of self-monitoring tools carries broad
implications, and the principles of Human Factors should be
applied to them. If a self-monitoring tool is unobtrusive and
easy to use, then users will be more likely to use the tool
effectively and routinely. If such a tool can be shown to be
accurate as well, then it can be used to improve the success of
behavioral interventions and help combat the obesity
epidemic.
Human Factors Approach
As the study progressed it became apparent that the
individual study participants shared behavioral traits that could
be easily aggregated into user personas. Normally employed
as a tool to assist designers and engineers in product
development goals, the establishment of user personas is an
invaluable tool to assist in describing common user behaviors.
A logical application of Human Factors principles to the
development of the Bite Counter, user personas were
established to describe the common positive and negative
behaviors observed during in-person weekly follow-up
sessions during this study. By examining how these traits
influenced key behavioral factors, these user personas can be
used as a predictor of potential success, as well as an analysis
tool for future studies. The personas developed are described
below.
Personas
Committed Cathy. Cathy
is excited and cannot wait to
get started. She asks questions
about the Bite Counter and
how she can use it to help her
lose weight. During your
interview you notice that
Cathy is knowledgeable about
food and calories and has
experience with weight loss
and dieting. She leaves the
office wearing the device and Figure 2. Committed Cathy
will usually have it on her wrist when she comes in to the lab
for follow-up appointments. Cathy will rapidly adopt the Bite
Counter, she seldom misses recording an eating activity and
she eats a consistent diet (times and amounts similar). You
notice that Cathy usually stops eating at the Bite Target Value
or alarm and usually loses weight each week. When Cathy
overeats she does so with the Bite Counter on, counting her
bites and can give you the details of why she overate when
asked. Cathy is highly motivated and excited about her weight
loss success and may ask if she can participate in an exercise
program to increase her rate of weight loss. Cathy also has the
full support of her friends and family, has professed her
intention to lose weight to everyone and can often be seen
explaining her Bite Counter to anyone who asks about it.
Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014 1432

Reluctant Rita. Rita has experience a tool for researchers. This chart is included in Table 2 below.
with weight loss and dieting and claims Key attributes were selected to help describe differences in
to be knowledgeable about food. She users based on an observation of the study participants
knows the diet lingo and professes to attitudes and habits during the combined 10-week study
have “tried them all”. Rita reports that period.
she is ready to lose weight, but
struggles with it. During follow-up Table 1. Bite Counter persona comparisons.
appointments you note that Rita never
has the Bite Counter on her wrist. She BITE COUNTER PERSONA COMPARISONS
usually keeps it in her purse or pocket. COMMITTED RELUCTANT NEGATIVE
ATTRIBUTE
She admits that she only puts the device CATHY RITA NANCY
on when she eats, and removes it Figure 3. Reluctant
Excitement
High Moderate Low
immediately. Her weight changes vary Rita
level
Moderate -
weekly with weight gains or losses of Diet Moderate - High - nothing
rarely
several pounds interlaced with weeks of no progress. Rita’s Experience some success
successful
works
diet is not consistent (times and amounts vary widely). She Only during
often forgets to wear the Bite Counter, but always has an Wear of Device All of the time Always forgets
meals
excuse about why she didn’t use the device. Some weeks Rita Daily Usage of Seldom forgets
Frequently
More missed
appears compliant with instructions, but there is a moderate Device to record meals
forgets to
than recorded
weight gain (you suspect she is not recording all eating record meals
Consistency of Extremely
activities). She is quick to offer explanations about the Eating Times
Very consistent Inconsistent
inconsistent
problems that interfere with her ability to follow diet protocol. Consistency of Amounts are Amounts vary Amounts vary
Often, Rita lacks the support of friends and family or may Amount Eaten fairly constant moderately greatly
even report that they sabotage her efforts. Casual
Low Moderate High
Negative Nancy. Nancy is Consumption
an experienced dieter who has Weight Loss Yes Possible No
struggled with weight loss for Public Aware
Yes Possible No
years. She is knowledgeable of Diet
about food, calories and the Family Support Yes Possible No
benefits of exercise. Nancy Rarely Criticizes the Constantly
Complaints
knows that she has problems, complains device and diet complains
but uses them as an excuse. Accepts blame
Seldom accepts Always blames
Nancy listens to the Bite Figure 4. Negative Nancy Excuses blame for something/one
for failures
failures else
Counter explanation, but complains about its size and
appearance immediately. You notice that when Nancy reports
to the lab for subsequent screening appointments that she has
to “dig” the Bite counter out of the bottom of her purse. Future Work
During a data review you immediately notice several
consecutive days of no eating activity during the past week. These user personas can be used to aid future researchers
When she remembers to use the Bite Counter you notice that in the early identification of study participants who may be at
Nancy’s eating patterns are not consistent (times and amounts a higher risk for dropping out of the study or those who may
vary widely). Nancy’s weight varies weekly with a weight have trouble with Bite Counter compliance or adherence to a
gain or loss of several pounds. She almost never stops eating diet protocol based on bite count.
at a Bite Target Value or when the alarm sounds. She is Ideally, it would be beneficial to determine performance
excited when she has a successful week, but alibis her conduct predictors based on behavior screening questionnaires that
when she gains weight. When you question Nancy about her measure some easily adapted construct such as motivation,
use of the Bite Counter she repeats her criticism about the self-efficacy or body image and to correlate screening scores
device’s appearance noting that it is ugly and clunky. Nancy with these personas. This would give future researchers the
claims that the Bite Counter doesn’t fit her lifestyle (or eating ability to effectively classify and screen users during study
style). During follow-up interviews Nancy admits that she is intake and initial phases of the study, instead of needing weeks
embarrassed to wear the device because it lets everyone know of observation to identify problematic participants.
that she is trying to diet. Nancy also may lack the support of In conclusion, this study establishes three distinct user
friends and family and has not disclosed her weight concerns personas for the Bite Counter and provides researchers with a
or intention to lose weight to anyone. tool that can be used as a performance predictor for future
study participants.
Persona Comparisons

The personas were aggregated into a side-by-side

comparison chart of the study participants’ attributes to create
Proceedings of the Human Factors and Ergonomics Society 58th Annual Meeting - 2014 1433

REFERENCES

Baker, R. C., & Kirschenbaum, D. S. (1998). Weight control during the

holidays: Highly consistent self-monitoring as a potentially useful
coping mechanism. Health Psychology, 17(4), 367.
Burke, L. E., Styn, M. A., Glanz, K., Ewing, L. J., Elci, O. U., Conroy, M. B.,
Keating, A. L. (2009). SMART trial: A randomized clinical trial of
self-monitoring in behavioral weight management-design and
baseline findings. Contemporary Clinical Trials, 30(6), 540-551.
Burke, L. E., Wang, J., & Sevick, M. A. (2011). Self-monitoring in weight
loss: A systematic review of the literature. Journal of the American
Dietetic Association, 111(1), 92-102.
Dong, Y., Hoover, A., & Muth, E. (2009). A device for detecting and counting
bites of food taken by a person during eating. Paper presented at
the IEEE International Conference on Bioinformatics and
Biomedicine, 2009, 265-268.
Flegal, K. M., Carroll, M. D., Ogden, C. L., & Johnson, C. L. (2002).
Prevalence and trends in obesity among US adults, 1999-2000.
JAMA: The Journal of the American Medical Association,
288(14), 1723-1727.
Helsel, D. L., Jakicic, J. M., & Otto, A. D. (2007). Comparison of techniques
for self-monitoring eating and exercise behaviors on weight loss in
a correspondence-based intervention. Journal of the American
Dietetic Association, 107(10), 1807-1810.
LeRouge, C., Ma, J., Sneha, S., & Tolle, K. (2011). User profiles and personas
in the design and development of consumer health technologies.
International journal of medical informatics.
Livingstone, M. B. E., & Black, A. E. (2003). Markers of the validity of
reported energy intake. The Journal of Nutrition, 133(3), 895S-
920S.
Miaskiewicz, T., & Kozar, K. A. (2011). Personas and user-centered design:
How can personas benefit product design processes?. Design
Studies, 32(5), 417-430.
Schoeller, D. A. (1988). Measurement of energy expenditure in free-living
humans by using doubly labeled water. The Journal of Nutrition,
118(11), 1278-1289.