You can’t “nudge” nuggets: An investigation of college late-night dining with behavioral economics interventions
You can't ªnudgeº nuggets: An investigation of college late-night dining with behavioral economics interventions
Samuel Bevet 1 2
Meredith T. Niles 0 2
Lizzy Pope 0 2
0 Department of Nutrition and Food Sciences, University of Vermont , Burlington, Vermont , United States of America
1 Department of Food Systems, University of Vermont , Burlington, Vermont , United States of America
2 Editor: Rebecca A. Krukowski, University of Tennessee Health Science Center , UNITED STATES
A mixed-methods approach was used to evaluate and improve the ªlate-night diningº options in a university dining hall. Surveys assessed student desires for late-night offerings, and evaluated students' habits and motivations during late-night dining. Two interventions were implemented to see if students could be ªnudgedº into different choice patterns. In the first, a ªveggie-heavyº entreÂ e was added at the beginning of the entreÂ e line, so that students would substitute an entreÂ e containing vegetables for the alternatives. In the second, a snack-food bar was set up to cater to students who didn't want to stand in the long entreÂ e line, and preferred a snack. Data on food choice was collected during the interventions. Survey responses showed significant differences in the reasons females and males utilized late-night dining (p<0.001). We also found that students at late-night dining had a lower emphasis on health than the general student population. Even students at late-night who reported being healthconscious showed no difference in food selections from students who said health was not important (p = 0.883). Veggie-heavy entreÂ es had mild success in increasing vegetable selection. However, veggie-heavy entreÂ es were largely ignored when the other option was chicken nuggets. The snack bar was very popular. EntreÂ e placement and convenience lines may have mild impacts on food selection in a late-night dining environment.
Data Availability Statement: All relevant data are
within the paper and its Supporting Information
Funding: The study was funded by the University
of Vermont Dean of Student's Office. The funders
had no role in study design, data collection and
analysis, decision to publish, or preparation of the
Competing interests: The authors have declared
that no competing interests exist.
The American College Health Association reports that 33% of college students are overweight
or obese [
], and American undergraduates often gain weight while at college [
]. People who
become obese or develop poor eating habits during childhood and young adulthood are more
likely to struggle with these problems in adulthood [
]. This can lead to a variety of illnesses
including diabetes, cardiovascular disease, and cancer .
Consumption of ªjunk foodº and evening snacks may be large contributing factors in
college weight gain [
]. Evidence indicates between 32.5%-72.8% of college students report that
they often/always have evening snacks [
]. Some research suggests calories consumed late at
night contribute to greater weight gain than if the calories were consumed during the day, due
to the body's cyclical metabolism . This means unhealthy late-night meals may be especially
detrimental for diet-related health. A study of Korean college students found that living in
dormitories was associated with significantly increased calorie intake at night compared to living
at home, especially from fried chicken and flour-based foods [
]. However, little research has
analyzed late-night eating at American colleges.
Students living in dormitories on American college campuses often consume meals in
college dining halls as part of a pre-paid meal plan. These plans often are structured around ªall
you care to eatº dining experiences where a student can take and eat as much or as little food
as they would like each time they enter the dining hall. Dining hall intervention studies have
demonstrated the potential for cafeterias to encourage healthy eating choices. These
interventions are informed by behavioral economics, which examines the reasons why consumers
make decisions in the short-term that may conflict with long-term goals, such as remaining
]. The positioning of food within both the serving line and cafeteria has been
shown to influence the amount and types of food people will take [11±14]. These interventions
have been called ªNudging Interventions,º because they subtly push consumers towards
healthier choices without removing unhealthy options [
]. While many studies have looked
at applying behavioral economics to dining halls, to our knowledge no studies have looked at
using these interventions during late-night dining. To address this research gap, the objectives
of this study were to examine college students' perceptions of health and late-night dining,
while also implementing two nudging interventions in a late-night dining environment to
determine their impact on food choices.
Research was conducted at the University of Vermont (UVM), which enrolled 9,786
undergraduate students in the 2016±2017 school year. ªLate-night diningº was held Monday-Wednesday
from 10:00pm-12:30am at an all-you-care-to-eat cafeteria on campus. Minimal food service
staffing during these hours restricted the types of food that could be easily served. Options were
usually limited to fried processed foods and dishes that could be premade and quickly reheated,
such as chicken nuggets, corn dogs, and pulled pork.
The study was approved by the UVM Committee on Human Research in the Behavioral and
Social Sciences and deemed exempt. Participants completing surveys could read a study
information sheet before choosing whether or not to complete surveys. The IRB approved our request
for a waiver of documentation of consent, as participants' willingness to complete the survey was
considered consent and no identifying information was collected with research data. After IRB
approval was received, a survey about late-night dining was administered through a survey
delivery program, Campus Labs. The survey was developed collaboratively to collect information of
interest to both campus administrators, Dining Services, and researchers. Participants were
recruited through an email to all 4,994 students living in on-campus housing. Survey completion
was voluntary, and incentivized by the chance to win one of five $25 gift cards towards
on-campus dining. Students were asked about their frequency and reasons for attending the university's
late-night dining option. Students used a 7-point Likert scale to rate the importance of health on
their late-night food choices (ªHealth scoreº), ranging from ªNot at allº to ªVery important.º
Students were also asked about other foods they would like to see offered (if any). Responses were
coded into three categories: 1) those wanting healthy options, 2) those wanting less healthy
options, or 3) those seeking ªmoreº options [
]. Two researchers (including a Registered
Dietitian) coded each response and then compared codes and resolved any coding disagreements by
discussing the particular food item and consulting nutrition facts. Requests for fruits, vegetables,
lean protein, and whole grains such as a salad bar, fruit bar, or grain bowls were classified as
healthy options whereas options high in fat, sugar, or calories such as ice cream, fries, and
2 / 12
chicken nuggets were classified as less healthy. Students who requested both healthy and less
healthy options or whose requests could not be easily classified as healthy or unhealthy without
more information, such as those requesting cereal were classified as wanting ªmoreº options.
Table 1 displays survey questions and variables created for analysis.
Following the online survey, a second survey was administered in-person in the cafeteria
during late-night dining. This survey contained a subset of the questions from the emailed
survey. Additionally, students reported the food they had selected to eat that night. These reported
food choices were coded by two researchers as healthy, less healthy, both, or unknown using the
same criteria as the Pre-Survey. Due to the anonymous nature of both surveys, it is unknown
how many students completed both the Pre-Survey and the At-Late-Night Survey.
Researchers and dining staff worked together to implement two behavioral
economicsbased interventions at late-night dining during the spring semester of 2017. In the first
intervention, vegetable-heavy entreÂes were added at the beginning of the self-serve line. These
entreÂes were vegetable lasagna (Mondays), broccoli mac-and-cheese (Tuesdays), veggie-egg
scramble and a root-vegetable hash (Wednesdays); they were added to the usual options of
chicken nuggets (Mondays), pulled pork sandwiches (Tuesdays), and pancakes and sausage
(Wednesdays). The intervention options were not necessarily lower in calories and saturated
fat or substantially higher in key micronutrients than the traditional options. They were
designed to be 1) easy for the small late-night staff to prepare, and 2) appealing enough to
latenight diners so that we could test the concept that students could be nudged into taking
vegetable-containing entreÂe options by placing these options at the beginning of a serving line. The
intervention was carried out for three weeks, for a total of nine days (late-night dining only
occurred on three days out of the week). During this time, researchers stood near the serving
line to record the food choices and gender of everyone taking food. Additionally, the number
of people coming into the dining hall each night was recorded electronically through the cash
In the second intervention, a snack-food convenience line called the ªCrunchy Munchy
Barº was added. It was placed beside the salad bar, which Dining Services reported had
minimal foot traffic prior to the intervention. Snack foods included chips and salsa, hummus,
1±7 Likert Scale:
1 = Not at all, 7 = Very
3 / 12
popcorn, trail mix, yogurt, and pre-cut fruit. The Crunchy Munchy bar was designed to appeal
to students who did not want to wait in a long entreÂe line, and might only be looking for a
snack at late night. Researchers tallied how many students took something from the snack
food line, broken down by gender. This intervention was also carried out for three weeks, for a
total of nine days. During this time, the veggie-heavy entreÂe intervention was discontinued so
that each intervention could be looked at in isolation. Initially, we hoped to assess the
effectiveness of both interventions using Dining Services production reports. Unfortunately, these
reports were less accurate than we anticipated. Therefore, we did not get the depth of
quantitative food production data that would have allowed us to compare our observational data with
all food served every night.
Survey responses were analyzed using STATA 15 [
]. Since the majority of the variables
were either binary or ordinal, Kendall's tau correlations were used to determine relationships
between variables on the Pre-Survey and At-Late-Night Survey. Statistically significant
differences were examined for binary variables using Chi-Square tests. An analysis of variance with
Scheffe's multiple comparison tests [
] was utilized to explore varying outcomes based on
different groups and pairwise comparisons among variables with Likert or other numerical
outcomes. Finally, to assess the multiple potential variables related to a student's perceived health
score, ordered logistic regressions were used on both the Pre-Survey and the At-Late-Night
Survey. Results for the model are reported in odds ratios, which can be interpreted that any
coefficient below 1.00 is a reduced odds and anything with a coefficient higher than 1.00 is an
Descriptive statistics for both surveys are reported in Table 2.
Pre-survey. Our Pre-Survey received 681 responses for a response rate of 13.6%. Students
provided open-ended answers to the question ªAre there any particular foods you would like
to see offered at Late-Night?º Of these responses (N = 409), 39.6% of students requested only
healthier options, 30.1% of students requested only less healthy options, and 29.8% requested a
combination of both. Additionally, 10.8% of all these responses made explicit requests for
more vegetarian/vegan options. The mean Health Score for respondents was 3.94, with 14% of
respondents reporting a Health Score of 1 meaning that health was not at all a factor in their
late-night food choices. An ANOVA test compared Student Health Scores between those that
reported attending late-night at least weekly (mean Health Score = 3.75, SD = 1.88) and those
who reported going less than weekly/never (mean Health Score = 4.23, SD = 1.94), which was
statistically significant difference between the groups (F = 9.68, p<0.01).
ANOVA results also examined whether gender was correlated with different reasons for
attending late night. Overall, 49% of males compared to 25% of females reported attending late
night for a meal, a statistically significant difference (Chi2 = 53.89, p<0.001). Conversely, fifteen
percent of males compared to 33% of females reported attending late-night for socializing
(Chi2 = 4.91, p<0.001). There was no significant difference in gender for those attending late
night for a snack or because they were bored. Additionally, ANOVA results examined the
relationship between variables (Desires), (Satisfaction), and (Health Score). Students that
exclusively wanted less healthy options (Desire_Unhealthy) were significantly more likely to have a
lower Health Score (M = 2.66, SD = 1.57) than other students (M = 4.72, SD = 1.77, F = 124.28,
p<0.001). Fifty-seven percent of students who wanted less healthy options (Desire_Unhealthy)
were satisfied with the food offered at late-night dining, compared to 21% for other students,
which was statistically significant (Chi2 = 52.12, p<0.001). Conversely, students who requested
4 / 12
Note. aValues in parentheses denote the subset of respondents who reported attending late-night at least once per week.
bCategorical variables reported as percentages.
healthier options (Desire_Healthy) had significantly higher Health Scores (M = 5.35, SD = 1.60)
than students who requested other types of food (M = 3.28, SD = 1.73, F = 147.67, p<0.001).
Only 19.7% of students who requested healthy options (Desire_Healthy) were satisfied with the
food at late-night dining, compared to 40.5% of other students, which was significantly different
(Chi2 = 19.24, p<0.001).
Pre-survey model. A logistic regression model was run to look at factors influencing a
student's Health Score (Table 3). We found that students who more frequently ate late-night
meals when home from college (Home_Habits), students who were more satisfied with the
current late-night offerings (Satisfaction), and students who attended late-night dining
primarily for snacking or socializing (Reason_Snack; Reason_Socialize), all had significantly
greater odds of having a lower Health Score (p<0.05). Students who desired healthy options
(Desire_Healthy) and students who desired vegetarian options (Desire_Vegetarian) both had
increased odds of having a higher Health Score, but only Desire_Healthy was statistically
5 / 12
At-late-night survey. One hundred and twenty eight students agreed to take the
in-person survey conducted during late-night dining. Descriptive statistics are reported in Table 2.
Although we did not explicitly ask for class year on this survey, the majority of respondents
were likely to be freshman and sophomores, as they almost exclusively are the students who
live on campus and would be attending late-night dining. The mean Health Score of students
surveyed at late-night dining was 3.48. Notably, 24% of respondents at late-night dining chose
a Health Score of 1, indicating health was ªnot at allº a factor in their late-night dining choices
(as compared to 14% in the Pre-Survey).
ANOVA tests were used to compare each student's Health Score to their actual reported
food choices. There was no statistically significant difference between a student's Health Score
and the foods they actually took at late-night dining, (F = 0.39, p = 0.883), indicating that how
important health was for their late-night dining options did not relate to actual entreÂe choice.
A logistic regression model was run to look at factors influencing a student's Health Score at
late-night dining (Table 3). Students who reported being unsatisfied with the offerings at
latenight dining (Satisfaction) had significantly increased odds of having a higher Health Score
(p<0.001). No other factors from the At-Late-Night Survey had any significant influence on a
student's Health Score.
Vegetable-heavy entreÂes were added to the main entreÂe line for three weeks, from March
27-April 12. During this time, researchers observed 2,397 trips through the entreÂe line, 28% by
females and 72% by males. Student food choices during the vegetable-heavy entreÂe
intervention period are shown in Fig 1. March 27, April 3, and April 10 are all Mondays when chicken
nuggets were served alongside vegetable lasagna. A sharp contrast between those three
Mondays and the other six dates can be seen. On Tuesdays and Wednesdays, between 54%-79% of
students incorporated a vegetable-heavy entreÂe into their late-night dining selection
(VeggieHeavy + Both). On chicken nugget Mondays, only between 9%-14% of students took a serving
of the vegetable lasagna.
After the vegetable-heavy entreÂe intervention concluded, the Crunchy Munchy healthy
snack food bar was implemented. Over the course of three weeks, students made 1,975 trips to
6 / 12
Fig 1. Vegetable-heavy intervention observations. Percentage breakdowns of daily student food choice during our first intervention period.
the snack bar. Gender breakdown for usage was 51% female, 49% male. Qualitative assessment
of student feedback (shown through quotes in Table 4), suggest that the Crunchy Munchy Bar
led to increased selection of (and excitement about) healthy foods.
From our two surveys, we gain some important insight into the way college students think
about late-night food and health. We find key differences between the responses of students
completing the Pre-Survey away from late-night dining compared with students taking the
Student Quotes at Crunchy Munchy
I just ate mushrooms at late night, it was
an incredible experience.
This is awesome
Actually, like, decent food
They have crunchy munchies!
7 / 12
At-Late-Night Survey while in the late-night environment. We also see no correlation between
students' health goals and their behaviors during late-night dining.
The mean student Health Score on the Pre-Survey was 3.94, and 3.75 for students who
reported attending late-night dining at least once per week, while the mean Health Score of
students at late-night dining was 3.48. While we cannot compare these statistically since the
samples are not the same, this suggests that in the moment while at late-night dining, students
place a lower emphasis on healthy eating than they do at other times. This could be due to
projection bias, which is when a person incorrectly estimates how they will react in a future
]. Students taking the Pre-Survey in a ªcoldº logical state imagine that health will be a
very important factor in their future food choices; this ends up not being true when they arrive
at late-night dining in a ªhotº visceral state [
]. When not at late-night dining, students
overestimate the importance of health on their decision-making. After a long day of classes and
homework, students place less emphasis on health and are more interested in fulfilling their
immediate cravings for comfort food.
Student food choices during late-night dining were not significantly impacted by student
Health Scores. Students were just as likely to choose vegetable-heavy options if they said health
was ªvery importantº or ªnot at all importantº to their food choices at late-night dining.
Students may not have perceived the vegetable-heavy entreÂes to be that much healthier than the
original options. Alternatively, this could also be due to what O' Donoghue & Rabin [
to as present-biased preferences: students put more weight on their immediate preferences
(eating chicken nuggets) than their long-term goals (eating more vegetables). The tendency to
make decisions in the present that are immediately rewarding versus making decisions that
might lead to greater gains in the long term can also be explained by hyperbolic discounting
theory. Hyperbolic discounting refers to the tendency for consumers to pick smaller-sooner
rewards rather than larger-later rewards that they would need to wait for, especially if the
smaller-sooner reward is immediately available [
]. In this study, students in the late-night
dining environment seemed more likely to indulge their hedonic or taste preferences rather
than make choices that might be consistent with their long-term health goals, effectively
discounting the importance of the health goals. This can present a challenge for dining services
trying to satisfy student desires. In ªcoldº states, students request that healthier options be
offered; however, when they are in ªhotº states they walk right past the healthy options and
head for the junk food. This disconnect between students' stated desires and actions can be
frustrating for dining service administrators, and may encourage the dining service to simply
continue catering to students' ªhotº states.
Students' present-biased preferences could potentially be exacerbated by intoxicants such
as alcohol or marijuana. As Ajzen [
] notes, ªperformance of a behavior is a joint function of
intentions and perceived behavioral controlº ([
], p. 185). While students may have the
intention of eating healthfully, intoxicants may reduce their ability to regulate their own behaviors.
We did not collect data on rates of student intoxication, but multiple students were overheard
saying they were currently under the influence of marijuana, and another study identified
alcohol as an influencing factor in students' late-night food consumption [
]. Future research on
the role intoxicants play in student food choice is needed.
We observed an interesting gender divide between the entreÂe station and the Crunchy
Munchy bar. Only 28% of students using the entreÂe line were female, while 51% of Crunchy
Munchy bar trips were made by females. This is consistent with our survey responses, where
women reported being much more likely to attend late-night dining for a snack or to socialize,
while men were much more likely to go for a meal. Counihan [
] attributes gendered eating
differences in America to cultural food norms, where men are socially encouraged to consume
large amounts of hearty food while women are encouraged to more sparingly eat healthy
8 / 12
items. However, Wichianson, Bughi, Unger, Spruijt-Metz, & Nguyen-Rodriquez [
identified stress as a common reason for college students' nighttime food consumption, and found
that among their sample, men were more likely than women to use maladaptive eating
practices to try to manage stress. Understanding the gender division in late-night dining would be
beneficial both for student health and college dining services.
Our nudging interventions appear to have been partially successful in increasing the choice
of vegetable-heavy entreÂes and snack foods. Based on our observational data, we know that
students were incorporating more vegetables into their diet than they otherwise would have, since
prior to the intervention no vegetables were served in the late-night entreÂes. However, one
unintended consequence of these interventions may have been that the additions of new
vegetable and snack options just led to more food selection, rather than a reduction in less healthy
selections. It is likely that if students took snacks from the Crunchy Munchy Bar and a
latenight entreÂe they would be eating additional food. Similarly, taking a large portion of snacks,
even healthy ones, from the Crunchy Munchy Bar may result in the same health/weight
outcomes as just eating the original entreÂes. Increased consumption of vegetables and fruits is often
suggested as a way to promote healthy weight due to their low energy-density and high fiber
]. However, Djuric et al. [
] observed a six-pound weight gain among women who
only focused on increasing vegetable and fruit consumption without also focusing on reducing
fat intake. Another study found that some vegetables were associated with weight loss, while
others were associated with weight gain [
]. Solely emphasizing vegetable and fruit
consumption may not be enough to positively influence college student health in dining halls. We would
need more concrete data on how much food was served to draw conclusions about the
interventions' effectiveness at improving student health.
Chicken nugget Mondays appear to have been mostly impervious to nudging interventions.
Although we did see a drop in nugget servings per student, around 90% of students on Mondays
ignored the vegetable lasagna in favor of the chicken nuggets and french-fries. Several factors
might be in play here. The first is that self-serving the lasagna from the tray took a bit more
effort than scooping up nuggets. The lasagna was pre-cut, but students had to use a spatula to
separate and serve pieces. This could slow down their progression, while dozens of other hungry
college students waited behind them. Research has shown that even mild increases in the effort
needed to access food can reduce selection [
]. The second factor is that students may just
have a strong preference for chicken nuggets, or a strong preference against vegetable lasagna,
that cannot be overridden by nudging interventions. One study targeting elementary students
tried to increase fruit consumption over french-fry consumption by making apple slices the
default item served, but given the option, 96% of students switched their apple slices for fries
]. As long as french-fries were available, students took them. Similarly, when chicken nuggets
were offered, students were able to ignore the nudging intervention, skipping right over the
vegetable lasagna and loading up their plates with piles of nuggets. Finally, there could have been a
ªMonday effectº for chicken nugget choice where the stress of starting a new week and
anticipating the week ahead led to more students choosing chicken nuggets and being impervious to
the nudging intervention. These results demonstrate some potential limitations for nudging
interventions to positively influence consumer health.
Our interventions concluded at the end of the spring semester. The following fall semester,
dining services continued to offer both vegetable-heavy entreÂes and the Crunchy Munchy bar.
Instead of using the nudging intervention, vegetable-heavy entreÂes are the only option served
on some nights. The campus head chef reported that this may potentially reduce food costs,
because students were previously over-serving themselves the less healthy options. He also
noted that the Crunchy Munchy bar has continued to be popular with students. Dining
9 / 12
services also decided to only serve chicken nuggets occasionally, rather than every Monday, to
make nugget night less of a habitual weekly event.
This research had several limiting factors. Our Pre-Survey and At-Late-Night Survey had
population overlap; therefore while we were able to note differences between the two groups,
we were unable to compare them statistically. Students were asked about the role health plays
in their decision-making, but ªhealthº was self-defined by each student. We focus on physical
health here, but students may have considered mental health in their answers. To gain a greater
understanding of how student preferences change between ªhotº and ªcoldº states, future
research should include more nuanced definitions of health, and either track the same student
responses over time or ensure statistically independent samples. Our conclusions were also
limited by problems with the Dining Services production report data, which made it difficult
to assess our interventions' effects on less healthy food selection. More robust tracking of the
food served and wasted by students would be beneficial for evaluating cafeteria nudges. The
vegetable-heavy entreÂes we chose to offer may still have been high in fat and calories, and
therefore not seen as ªhealthyº by students at late-night dining. A nutrition analysis of generic
versions of each entreÂe indicated that the veggie-heavy entreÂes were lower in fat, sugar, and
sodium as well as higher in Vit A and Vit C than generic versions of the normal entreÂes, but
similar in calories and saturated fat. Using entreÂes that are lower in fat and calories may better
illustrate whether late-night dining food choice is associated with overall interest in one's
health. Finally, our study was limited to repeated menu offerings of the same foods on the
same days each week. Randomized control trials evaluating different food pairings on different
days of the week, and comparison treatments using other non-vegetable entrees, could help
control for additional factors influencing student food choice. This would allow for better
evaluation of the role food-positioning plays in food-selection. The biggest strength of our research
was taking a mixed-methods approach to investigate a relatively unstudied area of student
dining. Through a mixture of quantitative, qualitative, and observational data, we were able to
create a baseline understanding of students' late-night dining behavior that can inform future
From our survey data, we concluded that the stated importance of health on food selection did
not have a relationship to actual student food choice. We also found that, on average, students
not at late-night dining placed a higher value on health than students attending late-night
dining. We found a significant difference in the reasons males and females attended late-night
dining, with males more likely to go for a meal and females more likely to go to socialize.
Although we do not know whether our nudging interventions decreased less healthy food
selection, they were effective at increasing vegetable selection in at least some contexts. The
exception to this was during chicken nugget nights, where students demonstrated their
overwhelming preference for nuggets. For colleges and dining services looking to positively impact
student health, it is important to assess the strengths, but also the limitations, of nudging
interventions within the dining hall.
S1 Dataset. Pre-survey anonymousÐThis file contains the data from the survey conducted when participants were not at late-night dining.
10 / 12
S2 Dataset. At-late-night survey dataÐThis file contains the data from the survey conducted at late-night dining.
S3 Dataset. Late night dining survey dataÐThis file contains data about participants
choice of entreÂe during the veggie-heavy entreÂe first in line intervention.
S4 Dataset. Crunchy Munchy comparison dataÐThis file contains data about food choice during the Crunchy Munchy Bar intervention.
Conceptualization: Samuel Bevet, Lizzy Pope.
Data curation: Samuel Bevet, Meredith T. Niles, Lizzy Pope.
Formal analysis: Samuel Bevet, Meredith T. Niles.
Investigation: Samuel Bevet, Lizzy Pope.
Methodology: Samuel Bevet, Lizzy Pope.
Project administration: Samuel Bevet, Lizzy Pope.
Writing ± original draft: Samuel Bevet, Lizzy Pope.
Writing ± review & editing: Samuel Bevet, Meredith T. Niles, Lizzy Pope.
11 / 12
1. Pope L , Hansen D , Harvey J . Examining the Weight Trajectory of College Students . J Nutr Educ Behav . 2016 ; 1±6 . Available from: http://linkinghub.elsevier.com/retrieve/pii/S1499404616308727
2. Serdula MK , Ivery D , Coates RJ , Freedman DS , Williamson DF , Byers T. Do obese children become obese adults? A review of the literature . Vol. 22 , Preventive medicine. 1993 . p. 167 ± 77 . Available from: http://www.sciencedirect.com/science/article/pii/S0091743583710145 PMID: 8483856
3. Nelson MC , Story M , Larson NI , Neumark-Sztainer D , Lytle LA . Emerging Adulthood and College-aged Youth: An Overlooked Age for Weight-related Behavior Change . Obesity . 2008 ; 16 ( 10 ): 2205 ± 11 . Available from: http://doi.wiley. com/10 .1038/oby. 2008 .365 PMID: 18719665
World Health Organization. Diet , Nutrition and the Prevention of Chronic Diseases . Geneva; 2003 .
5. Levitsky D , Halbmaier C , Mrdjenovic G. The freshman weight gain: a model for the study of the epidemic of obesity . Int J Obes . 2004 ; 28 : 1435 ± 42 .
6. Driskell JA , Kim YN , Goebel KJ . Few differences found in the typical eating and physical activity habits of lower-level and upper-level university students . J Am Diet Assoc . 2005 ; 105 ( 5 ): 798 ± 801 . https://doi. org/10.1016/j.jada. 2005 . 02 .004 PMID: 15883559
7. DeBate RD , Topping M , Sargent RG . Racial and Gender Differences in Weight Status and Dietary Practices Among College Student . Adolesence . 2001 ; 36 ( 144 ): 819 ± 33 .
8. Garaulet M , GoÂmez-AbellaÂn P . Timing of food intake and obesity: a novel association . Physiol Behav . 2014 ; 134 : 44 ± 50 . https://doi.org/10.1016/j.physbeh. 2014 . 01 .001 PMID: 24467926
9. Jun YS , Choi MK , Bae YJ . Night eating and nutrient intake status according to residence type in university students . J Korean Soc Food Sci Nutr . 2015 ; 44 ( 2 ): 216 ± 25 .
10. Thaler RH , Sunstein CR . Nudge: improving decisions about health, wealth, and happiness . New Haven, Conn.: Yale University Press; 2008 .
Wansink B , Hanks AS . Slim by Design: Serving Healthy Foods First in Buffet Lines Improves Overall Meal Selection . PLoS One . 2013 ; 8 ( 10 ): 1 ± 5 .
12. Hanks AS , Just DR , Smith LE , Wansink B . Healthy convenience: Nudging students toward healthier choices in the lunchroom . J Public Heal (United Kingdom) . 2012 ; 34 ( 3 ): 370 ± 6 .
13. Thorndike AN , Sonnenberg L , Riis J , Barraclough S , Levy DE . A 2-phase labeling and choice architecture intervention to improve healthy food and beverage choices . Am J Public Health . 2012 ; 102 ( 3 ): 527 ± 33 . https://doi.org/10.2105/AJPH. 2011 .300391 PMID: 22390518
14. Rozin P , Scott S , Dingley M , Urbanek JK , Jiang H , Kaltenbach M. Nudge to nobesity I: Minor changes in accessibility decrease food intake . Judgm Decis Mak . 2011 ; 6 ( 4 ): 323 ± 32 .
15. Corbin JM , Strauss A . Grounded theory research: Procedures, canons, and evaluative criteria . Qual Sociol . 1990 ; 13 ( 1 ):3± 21 .
16. StataCorp. Stata Statistical Software: Release 15 . College Station, TX: StataCorp LLC; 2017 .
17. Jaccard J , Becker MA , Wood G. Pairwise Multiple Comparison Procedures : A Review. Psychol Bull . 1984 ; 96 ( 3 ): 589 ± 96 .
18. Loewenstein G , O'Donoghue T , Rabin M.Projection Bias in Predicting Future Utility . Q J Econ . 2003 ; 118 ( 4 ): 1209 ± 48 . Available from: https://academic.oup.com/qje/article-lookup/doi/10.1162/ 003355303322552784
19. Liu PJ , Wisdom J , Roberto CA , Liu LJ , Ubel PA . Using behavioral economics to design more effective food policies to address obesity . Appl Econ Perspect Policy . 2014 ; 36 ( 1 ):6± 24 .
20. O ' Donoghue T , Rabin M. Doing It Now or Later . Am Econ Rev . 1999 ; 89 ( 1 ): 103 ± 24 .
21. Laibson D. Golden Eggs and Hyperbolic Discounting . Q J Econ . 1997 ; 112 ( 2 ): 443 ± 78 . Available from: https://academic.oup.com/qje/article-lookup/doi/10.1162/003355397555253
22. Ajzen I. The theory of planned behavior . Orgnizational Behav Hum Decis Process . 1991 ; 50 : 179 ± 211 .
23. Nelson MC , Kocos R , Lytle LA , Perry CL . Understanding the Perceived Determinants of Weight-related Behaviors in Late Adolescence: A Qualitative Analysis among College Youth . J Nutr Educ Behav . 2009 ; 41 ( 4 ): 287 ± 92 . Available from: https://doi.org/10.1016/j.jneb. 2008 . 05 .005 PMID: 19508935
24. Counihan CM . Food Rules in the United States: Individualism, Control, and Hierarchy. Anthropol Q. 1992 ; 65 ( 2 ): 55 . Available from: http://www.jstor.org/stable/3318134?origin=crossref
25. Wichianson JR , Bughi SA , Unger JB , Spruijt-Metz D , Nguyen-Rodriquez ST . Perceived stress, coping and night-eating in college students . Stress Heal . 2009 ; 25 ( 3 ): 235 ± 40 .
26. Rolls BJ , Ello-Martin JA , Tohill BC . What Can Intervention Studies Tell Us about the Relationship between Fruit and Vegetable Consumption and Weight Management? Nutr Rev . 2004 ; 62 ( 1 ):1± 17 . PMID: 14995052
27. Djuric Z , Poore KM , Depper JB , Uhley VE , Lababidi S , Covington C , et al. Methods to Increase Fruit and Vegetable Intake With and Without a Decrease in Fat Intake: Compliance and Effects on Body Weight in the Nutrition and Breast Health Study . 2002 ; 43 ( 2 ): 141 ± 51 .
28. Bertoia ML , Mukamal KJ , Cahill LE , Hou T , Ludwig DS , Mozaffarian D , et al. Changes in Intake of Fruits and Vegetables and Weight Change in United States Men and Women Followed for Up to 24 Years: Analysis from Three Prospective Cohort Studies . PLOS . 2015 ; 1 ± 20 .
29. Brunner TA . It takes some effort. How minimal physical effort reduces consumption volume . Appetite [Internet] . 2013 ; 71 : 89 ± 94 . Available from: https://doi.org/10.1016/j.appet. 2013 . 07 .014 PMID: 23932949
30. Just DR , Wansink B. Better School Meals on a Budget: Using Behavioral Economics and Food Psychology to Improve Meal Selection . Choices . 2009 ; 24 ( 3 ):1± 6 .