IMPACT OF PLATE SHAPE AND SIZE ON INDIVIDUAL FOOD WASTE IN A

UNIVERSITY DINING HALL SETTING

BY

RACHEL McCLAIN RICHARDSON

THESIS

Submitted in partial fulfillment of the requirements

for the degree of Master of Science in Agricultural and Applied Economics
in the Graduate College of the
University of Illinois at Urbana-Champaign, 2019

Urbana, Illinois

Advisers:

Associate Professor Brenna Ellison

Assistant Professor Melissa Pflugh Prescott
ABSTRACT

With an estimated 50% increase in global food demand by the year 2050 (Campanhola &
Pandey, 2019), countries are trying to find ways to increase production and decrease waste to
help meet these needs. Young adults (18-24 years of age) have been identified as a high-wasting
segment of the population (Thyberg & Tonjes, 2016). In the United States, young adulthood often
coincides with the pursuit of postsecondary education. Many students receive housing and meals
through the university. Because of this, university dining facilities make an excellent target for
food waste reduction strategies. The purpose of this study is to evaluate one food waste
reduction strategy: changing the plate size and shape in university dining facilities. Specifically,
this study compares individual food selection, consumption, and waste between round plates (9”
x 9”) and smaller oval platters (9.75” x 7.75”) in a self-serve, all-you-care-to-eat dining
environment. Data was collected at an individual level where diners’ plates were weighed directly
after selection and again before disposal. Results suggest using plates with a smaller surface area
reduces food selection, consumption, and waste. However, the intervention does increase the
odds of a diner selecting seconds, but the amount of waste produced from a second helping could
not be measured. Implementing an intervention such as this in many university dining halls may
contribute to reducing global food waste among a highly wasteful population.

ii
ACKNOWLEDGEMENTS

I am very grateful for all of the support I have received throughout completion of this
project. First, a huge thanks to my advisors Dr. Brenna Ellison and Dr. Melissa Pflugh Prescott for
their continual advice and feedback. Without Dr. Ellison’s interest in working with the University
of Illinois at Urbana-Champaign dining services this study would not have come to fruition.
Moreover, I would like to thank University Housing and dining services for allowing us to conduct
research in their dining halls for several weeks and continuing to assist with this project. Thank
you to Dr. Mary Arends-Kuenning and Dr. Mindy Mallory for taking time to serve on my
committee. Additionally, I’d like to thank the Department of Agricultural and Consumer
Economics for funding my education and for fostering a creative and collaborative learning
environment. Lastly, I’d like to thank my wonderful family and friends for their continued
support.

iii
TABLE OF CONTENTS

CHAPTER 1: INTRODUCTION ................................................................................................1

CHAPTER 2: REVIEW OF SCIENTIFIC LITERATURE ................................................................3

CHAPTER 3: METHODOLOGY & ANALYSIS ...........................................................................8

CHAPTER 4: RESULTS AND DISCUSSION ........................................................................... 14

CHAPTER 5: CONCLUSION................................................................................................. 20

FIGURES ............................................................................................................................ 21

TABLES .............................................................................................................................. 23

REFERENCES ...................................................................................................................... 29

APPENDIX A: DATA COLLECTION PROTOCOL.................................................................... 34

iv
CHAPTER 1: INTRODUCTION

Since the first report to the United States Congress on food waste in 1977 by the United
States Department of Agriculture (USDA), there has been interest in identifying ways to reduce
food loss and waste across the food system (Buzby, Wells, & Hyman, 2014). A recent assessment
on agriculture systems sustainability by The Food and Agriculture Organization (FAO) of the
United Nations (UN) estimates a 50% increase in global food demand by 2050 (Campanhola &
Pandey, 2019). FAO notes that global production already meets this demand, however challenges
with distribution among increasingly urbanized populations, change in dietary preferences, and
decline in health outcomes due to an increase in obesity and micronutrient deficiencies will cause
new strains on our current agricultural systems (Campanhola & Pandey, 2019). Already,
agricultural production uses 80% of available consumptive water in addition to about 50% of land
use (Birney, Franklin, Davidson, & Webber, 2017; Gunders, 2012). Overall, food processing and
agricultural production is estimated to consume 10% – 16% of total US energy (Canning, Charles,
Huang, Polenske, & Waters, 2010; United States Energy Information Administration, 2018). In
developed countries, a majority of food loss occurs at the retail and consumer-levels (Gunders,
2012). Buzby et al. (2014) estimates a combined food loss of 31% of total available food at the
retail and consumer-levels. These groups squander a remarkable amount of consumable water,
energy, and intangible resources input earlier in the supply chain (Birney et al., 2017). The USDA
and United States Environmental Protection Agency (EPA), in alignment with goals set by the UN,
have already set a national goal to reduce food loss and waste 50% by the year 2030 (United
States 2030 Food Loss and Waste Reduction Goal, 2015). However, the mechanisms to reduce
waste are broad. Specific policies and programs aimed at reducing food loss and waste are
continuing to develop to achieve this goal.
This study focused on retail and consumer level food waste rather than the all-
encompassing problem of food loss. While the definitions of food loss and waste vary (for an
overview, see (Bellemare, Cakir, Peterson, Novak, & Rudi, 2017; Ellison, Muth, & Golan, 2019),
we adopt the definition used by the USDA Economic Research Service (Buzby et al., 2014). Under
this definition, food loss includes any edible food consumable by humans lost postharvest due to

1
shrinkage, mold, pests, improper storage, among other loss mechanisms, while food waste is a
subset of food loss that includes any edible food portion left uneaten at the retail and consumer
levels (Buzby et al., 2014). From data reported in 2010, after recycling for paper waste, food
waste alone made up 21% of municipal solid waste (Buzby et al., 2014). With such a high
percentage of food waste occurring at the retail and consumer levels, waste reduction programs
targeting consumers may have a wide-reaching impact. Young adults between the ages of 18 and
24 are seen to waste more food than any other age group (Thyberg & Tonjes, 2016). In the United
States, this age group frequently attends post-secondary education where they are often
provided housing and meals through the university. University dining halls have been estimated
to expend 124.5 g of CO2e per diner (Costello, Birisci, & McGarvey, 2015). As of the 2015-2016
academic year, about 16.98 million students were enrolled in public and private four year
universities (National Center for Education Statistics, 2018), the majority of which will have dining
facilities available to (or even required for) students to obtain meals. 1 Thus, targeting university
dining halls could curb waste from a relatively inattentive population at a large scale.
This purpose of this study is to investigate the impact of replacing round plates with
smaller oval platters on individual selection, consumption, and waste (SCW) in a university dining
hall. This study also examines characteristics that may lead to an increase in individual plate
waste, including demographics and characteristics of the dining environment. Overall, this
research aims to fill gaps in the literature concerning the lack of interventions aimed at reducing
food waste (Reynolds et al., 2019).

1Many four-year universities require students who are living on campus, which is often a requirement for first-
year students, to purchase a meal plan. Thus, a conservative estimate would be that approximately 25% of
students eat in university dining facilities, though it is certainly possible that older students (sophomores, juniors,
seniors, or graduate students) may also opt in to purchasing a meal plan.

2
CHAPTER 2: REVIEW OF SCIENTIFIC LITERATURE

2.1 Overview
There have been a variety of studies measuring ways to alter university dining halls in
order to observe changes in food SCW. Many of these studies focus on improving food choices
by emphasizing healthier options or implementing nutrition (e.g., calorie) labeling in dining
facilities (e.g., Christoph, Ellison, & Meador, 2016; Cioffi, Levitsky, Pacanowski, & Bertz, 2015;
Scourboutakos et al., 2017). However, there are fewer studies dedicated to reducing food waste
in this setting. Of the existing research, food waste interventions within university dining halls
tend to fall within three categories: education campaigns, alternative pricing strategies, and
environmental nudges. Research on plate shape/size within a university dining hall has not yet
been conducted.

2.2 Education Campaigns

Educational campaign strategies tend to present educational materials in the form of
informational signs, table tents or napkin inserts, and customized posters with food waste data
specific to the dining hall or university (Ellison, Savchenko, Nikolaus, & Duff, 2019; Soares Pinto,
Machado dos Santos Pinto, Fochat Silva Melo, Santos Campos, & Marques-dos-santos Cordovil,
2020; Whitehair, Shanklin, & Brannon, 2013). These methods tend to be low-cost and require
minimal maintenance by dining staff (Ellison, Savchenko, et al., 2019). Both the Ellison et al.
(2019) and Whitehair et al. (2013) studies used signage with either informational food waste
statistics or a “call to action” message intended to encourage mindful selection and emphasize
personal responsibility. Similarly, Soares Pinto et al. (2020) created signage aimed at involving
students in waste reduction by having them voluntarily request smaller potions sizes at service
lines. Thus far, the results on educational campaigns are mixed. The campaign to have students
request smaller serving sizes led to a decrease in food wasted, but it was unclear if this was due
to the promoted behavioral change or an increase in consumption due to food waste awareness
(Soares Pinto et al., 2020). Additionally, prompting students to act on their pre-existing beliefs
about food waste through signage has been seen to lead to a decrease in food wasted (Whitehair

3
et al., 2013). However, Ellison et al. (2019) observed a minimal reduction of food waste after an
8-week intervention and noted students admitted they still selected more food than they could
eat despite awareness of food waste as a problem.
Many food waste studies suggest education campaigns as an effective method of food
waste reduction without reproducible results (Reynolds et al., 2019). Overall, educational
campaigns alone may not be the most effective way of reducing food waste as it wholly relies on
independent action of the dining patron. Moreover, any long-term effects of education
campaigns are unknown. As this student population ages, their attitudes and beliefs about food
waste may change (Thyberg & Tonjes, 2016). Educational campaigns implemented while students
receive their meals at university dining facilities may leave an impression reducing their food
waste in the future. At present, more direct methods targeting behavior may lead to larger
reductions in food waste.

2.3 Altering Pricing Strategy

There are no known studies directly comparing food waste between all-you-care-to-eat
and a la carte style university dining halls. However, several studies have identified price and cost
as motivating factors for consumers to reduce or avoid waste (e.g., Graham-Rowe, Jessop, &
Sparks, 2013; Jovanovic, Katare, & Wetzstein, 2018; Thyberg & Tonjes, 2016). A la carte dining
prices items individually where as an all-you-care-to-eat system charges students a one-time
entry fee and allows them to select as much food as they desire while in the confines of the dining
hall (Ellison et al., 2019; Heikkilä, Reinikainen, Katajajuuri, Silvennoinen, & Hartikainen, 2016).
Students identify price as being the chief motivator for food purchase (Kim, Ng, & Kim, 2009),
often making all-you-care-to-eat options more financially advantageous than a la carte. In an all-
you-can-eat fixed price dining hall, Jovanovic and colleagues (2018) found a 15% discount
conditional on finishing all food selected effective in reducing food waste. However, they did not
find a reduction in selection, suggesting students consumed more food than they otherwise
would to receive the discount (Jovanovic et al., 2018). This may raise other concerns related to
overeating or obesity. Kim et al. (2009) concluded that the optimal pricing strategy is one that
gives diners a high level of satisfaction. While changing the pricing structure is likely to generate

4
results in terms of waste reduction, long-term effects of a change in price strategy is unclear and
financial incentives may lead to an increase in negative health outcomes if they are designed to
reduce waste via greater consumption. Further, the all-you-care-to-eat structure is often used as
a marketing tool for many university dining operations when recruiting students, so dining
administrators may be reluctant to make a pricing change.

2.4 Environmental Nudges

Nudges have become an increasingly popular tactic to change food choice behaviors. In
general, nudges are adjustments to the choice environment that can alter behavior without
removing choices or changing economic incentives (Thaler & Sunstein, 2008). In the university
dining hall setting, decreasing standard serving sizes and implementing trayless dining are two
nudges designed to reduce food waste by preventing over-selection of foods. Two studies have
examined the impact of decreasing the standard serving size of French fries in a university setting
(Freedman & Brochado, 2009; Vermote et al., 2018). Freedman and Brochado (2009) reduced
standard serving sizes from 88 grams to 44g over four weeks in 15g increments. Diner SCW and
dining hall pre-consumer food production all decreased over time, with individual consumption
and plate waste significantly decreasing (Freedman & Brochado, 2009). Vermote et al. (2018)
found similar results; most notably, the proportion of French fries eaten by each individual
increased from 88.1% to 95.2% after a decrease in the standard size served. Standard serving size
reduction seems to be a viable and effective strategy in reducing food waste. However, this may
not be a feasible waste reduction solution in all-you-care-to-eat or self-serve dining
environments where pre-portioned servings are generally not provided.
Trayless dining is a similar type of nudge that has been widely adopted on college
campuses in recent years (Foderaro, 2009). Trayless dining removes the ability to carry multiple
dishes at once but does not restrict one’s choices. It can be easily implemented in a variety of
dining environments, generally at low-cost. Several studies found trayless dining decreased food
waste (Kim & Morawki, 2012; Rajbhandari-Thapa, Ingerson, & Lewis, 2018; Sarjahani, Serrano, &
Johnson, 2009; Thiagarajah & Getty, 2012). Over a 6-day period, Kim and Morawski (2012) saw a
32% reduction in food waste after tray removal. Rajbhandari-Thapa and colleagues (2018) found

5
a significant decrease in lunches and drinks served as well as fewer plates with uneaten food
once trayless dining was adopted.2 However, they did not directly measure SCW and instead used
plate types as a proxy for meal items selected (Rajbhandari-Thapa et al., 2018). In a 2012 study,
individual edible food waste significantly decreased by 18.4% after trayless dining was
implemented (Thiagarajah & Getty, 2012). Participants in both the Rajbhandari-Thapa et al.
(2018) and Thiagarajah and Getty (2012) studies selected less food in the first place, suggesting
trayless dining decreases selection which leads to an increase in consumption and decrease in
waste. While most studies concluded trayless dining reduced food waste, one study found the
impact of trayless dining to be more nuanced. Wansink and Just (2013) observed a decrease in
selection of salad, an increase in returning for seconds, and a marginal increase in disposal of
entrée and dessert items when trays were removed.
In addition to trayless dining and reducing standard serving sizes, changing the dishware
available to diners is another nudge that could reduce food waste in the dining hall environment.
In the general population, research has shown that consumers often respond to environmental
ques, such as plate and serving utensil size, when determining how much food is appropriate to
select and consume (e.g., DiSantis et al., 2013; Rolls, Roe, Halverson, & Meengs, 2007; Van
Ittersum & Wansink, 2011; Wansink & van Ittersum, 2013; Wansink, van Ittersum, & Painter,
2006). Both lab and field studies have explored the impact of plate and utensil size on food SCW.
Diners at a buffet style restaurant who were provided larger plates (26.5 cm) selected, consumed,
and wasted more than diners who were given smaller plates (21.0 cm) (Wansink & van Ittersum,
2013). Similarly, Wansink et al. (2006) found participants who received a larger bowl at a self-
serve ice cream bar served and consumed 31% more than those with a smaller bowl. DiSantis et
al. (2013) found a significant increase in lunch quantity self-served and consumed among first-
grade students when participants switch from child sized to adult sized dishware. In a lab setting,
Van Ittersum & Wansink (2011) observed bowl size to have a significant effect on soup portion

2The type of dining environment this study was conducted in was not specified. In a fixed-entry-price all-you-care-
to-eat environment, a decrease in number lunches and drinks served may be beneficial as this would reduce dining
hall production costs. However, if the number of lunches served decreased due to a decline in the number of
students eating at the dining hall this could negatively impact the dining facility’s revenue. Additionally, in an a la
carte environment where customers are charged by item, a reduction in lunches and drinks served would not be
beneficial to dining hall revenue.

6
served. Those with bowls smaller than the control underserved themselves, while participants
with larger bowls overserved themselves (Van Ittersum & Wansink, 2011). Further, Rolls et al.
(2007) saw no difference in food intake between plate sizes in a controlled lab setting. However,
this study did not measure difference in selection or waste and was not in a natural dining
environment. Overall, these results are promising, but evidence of this intervention working in a
dining hall setting has yet to be tested. This research aims to fill this gap in the literature by testing
the influence of plate size and shape on food SCW in a university dining hall setting.

7
CHAPTER 3: METHODOLOGY & ANALYSIS

3.1 Measurement Strategy

There are four standard food waste measurement strategies within dining halls and food
service: direct weight, visual observation, self-reported dietary recall, and digital recording
(Buzby & Guthrie, 2002; Christoph & Ellison, 2017; LizMartins, Cunha, Rodrigues, & Rocha,
2014).3 The direct weight method involves physically weighing a participant’s plate after selection
and before disposal. This strategy was identified as the optimal method for data collection as it
provides the ability to look at food SCW on an individual level and is easily implemented into an
all-you-care-to-eat dining environment. Additionally, direct weight is regarded as the most
accurate means of waste measurement and has the ability to capture individual behavioral
factors (LizMartins et al., 2014).

3.2 Study Design and Data Collection Strategy

Data was collected at two dining halls at a large midwestern university in the Fall of 2018.
Both facilities were all-you-care-to-eat and trayless. Diners were charged a single-entry fee and
were able to select as much food as desired while in the dining area. Both locations used self-
serve stations with each station specializing in a type of cuisine (e.g. American, Italian, Asian,
etc.). Additionally, each facility had a salad and deli bar with options for specialized diets (e.g.
vegetarian, gluten free, vegan, etc.). The researchers and dining hall staff monitored and limited
the dishware available for patron use. The only intervention this study focused on was change in
plate size/shape. Traditionally, the university dining facilities used round plates (9”x9”). In this
study, the intervention was to replace the round plates with oval platters (9.75”x7.75”)

3We use the direct weight method in this study but provide a brief description of the other methods here. Visual
observation method requires researchers to estimate the proportion of food leftover on participant’s trays.
Proportions are often categorized into quarters (e.g. all, ¾, ½, ¼, or none) based on a standard portion served
(Wansink & Just, 2013). Self-reported dietary recall utilizes participant memory to estimate their own SCW (Kubik
et al., 2003). Lastly, digital recording is similar to visual observation where researchers take photographs of
participants plates then later compare their contents and may estimate SCW (Christoph & Ellison, 2017).

8
decreasing the plate’s surface area by 6.76% (63.62 in2 to 59.32 in2).4 See figure 1 for examples
of both plate types. Using a crossover trial design (see table 1), data was collected in each dining
hall for two one-week periods in September and October. Both dining hall menus were on a four-
week cycle; the research team selected data collection dates where the menu offerings would be
the same for both the round plate and oval platter weeks in each dining facility. Lunch service
ran from 11:00am to 1:30pm.
Dining hall patrons were free to choose to participate in the study but were exposed to
the intervention and control plates regardless of participation. Additionally, those with a meal
plan were able to eat at any dining facility on campus, including locations that were not involved
in this study. Diners were eligible to participate if they were over the age of 18 and had not
already participated that day. Participants were able to take part on multiple days and treated as
a new observation each day of data collection.
A group of nine data collectors were trained using a standard protocol (see Appendix A)
before data collection began. Each position included a detailed explanation of their duties.
Researchers were able to review and clarify data collection procedures prior to the start of data
collection in September. Any given day, 3-5 researchers recruited, directly weighed, and digitally
recorded pre-consumption plates. One to two researchers were responsible for post-
consumption weight and digital record. There was at least one researcher experienced in food
waste data collection present each day of data collection to manage all other researchers and
assist with any problems.
Patrons were approached by researchers to participate and incentivized with a later
drawing for a $50 Amazon gift card. Diners were recruited after their plate(s) had been
assembled, but before they sat down at a table. This was to ensure that 1) interacting with a
researcher did not influence food selection and 2) diners had not started eating when asked to
participate. Researchers invited diners to assist with identifying popular food items and meals as
to avoid priming participants on the topic of food waste.5 Once diners gave verbal consent,

4 While the research team preferred to have each dish type offered in isolation, certain menu items (e.g., soups,
cereal) required non-plate dishes like cups and bowls. In addition, one of the dining facilities had a stir fry station
that required a different type of dish that could not be replaced with the round plates or oval platters.
5 University Dining administration was also interested in this information and wanted to collect feedback from

students.

9
researchers then placed each dish on a 5 kg scale (Taylor TE11FT 11lb. Digital Portion Control
Scale) and recorded the pre-consumption weight, a combined weight of the plate and food. A
photo of each plate and corresponding weight was taken to allow identification of non-
compliance and provide a second reference for data input. Only initial plates selected were
measured. We did not record the weight of a second round of plates by participants due to
logistical constraints of the dining halls and researcher limitations. Keeping track of a second
round of plates would require the researchers to closely observe participants, which is difficult
based on the open concept layout of both dining facilities; however, the researchers asked
participants whether they went back for seconds on the accompanying survey, which is described
below.
After pre-consumption weights and photos were completed, researchers gave
participants a survey to take while eating (see figure 2). Participants were asked about their
overall satisfaction with their meal, if they went back for seconds, how many other diners they
ate with, and general demographics. All factors included were thought to potentially influence
individual SCW based on previous studies conducted at the University of Illinois Champaign-
Urbana (Christoph & Ellison, 2017; Christoph et al., 2016; Ellison et al., 2019; Nikolaus, Nickols-
Richardson, & Ellison, 2018) and anecdotal evidence from dining administration and staff
members. To prevent leading or confusing questions, all survey questions were piloted with
several undergraduate students prior to the start of the study. Once done eating, participants
returned their plates and corresponding questionnaire to the researcher located by the dish
return. This researcher removed any inedible items left on the plates (e.g. napkins, straws, bones,
peels, etc). Then, the post-consumption weight was recorded along with a photo of each plate.
During data entry, participants with multiple round plates or oval platters had their SCW
weights combined. To find the true weight SCW per participant, the average plate weight was
subtracted from each observation. The research team weighed five empty round plates (average:
195.6 g) and five empty oval platters (average: 242.2 g) to determine the average plate weight.
It should be noted some round plates were not the same weight as the standard round plates.
After beginning data collection, dining administration estimated one in five round plates were
approximately 100 g heavier than the standard round plate (Etchison & Van Liew, 2018). Due to

10
researchers’ inability to visually differentiate between the standard and heavier round plates, the
research team decided to subtract the average plate weight of the standard round plates for all
round plate observations when calculating SCW.6 The research team felt as though this would
reflect the most accurate measurements and provide results most closely aligned with the true
outcome. Consumption was calculated by subtracting the waste weight from the selection weight
and was not directly observed.

3.3 Model & Data Analysis

Stata/MP Version 15.1 was used for data analysis. To assess the impact of change in plate
size and shape on individual food SCW, we conducted a t-test with unequal variances between
oval platters and round plates by location. To further determine intervention impact and control
for potential confounders, we estimate a linear regression model using survey responses and
individual plate SCW weights:

(1) 𝑌𝑖 = 𝐵0 𝐼𝑛𝑡𝑒𝑟𝑐𝑒𝑝𝑡 + 𝐵1 𝑂𝑣𝑎𝑙𝑖 + 𝐵2 𝐿𝑜𝑐𝑎𝑡𝑖𝑜𝑛𝑖 + 𝐵3 𝑆𝑒𝑐𝑜𝑛𝑑𝑠𝑖 + 𝐵4 𝐹𝑒𝑎𝑚𝑙𝑒𝑖

+ 𝐵5 𝐼𝑛𝑡 ′ 𝑙𝑆𝑡𝑢𝑑𝑒𝑛𝑡𝑖 + 𝐵6 𝑆𝑎𝑡𝑖𝑠𝑓𝑎𝑐𝑡𝑖𝑜𝑛𝑖 + 𝐵7 𝐿𝑢𝑛𝑐ℎ𝑀𝑎𝑡𝑒𝑠2𝑖
+ 𝐵8 𝐿𝑢𝑛𝑐ℎ𝑀𝑎𝑡𝑒𝑠3𝑖 + ε𝑖

Where Y denotes the outcome of interest (food selection, consumption, or waste) for individual
i in grams. For waste, both the weight in grams as well as the percent waste (calculated as waste
weight divided by selection weight) are used as dependent variables. Oval is an indicator variable
where 1 equals an oval platter and 0 equals a round plate. Location is an indicator variable that
equals 1 if the diner ate at Ikenberry Commons Dining Hall and 0 if the diner ate at Pennsylvania
Avenue Dining Hall. The indicator variable Seconds takes on 1 if participants indicated they
returned for more food and 0 if they did not. Female is an indicator variable where 1 is female
and 0 is male. International Student (Int’lStudent) is also an indicator variable with 1 representing

6Inthe future, the research team will go back to the plate photos to identify potential cases where individuals
received the heavier round plate. In these cases, post-photos would reveal a clean plate (0 g waste), but the
calculated waste value would be greater than zero, generally around 100 g as this was the weight difference
between the standard and heavier round plates.

11
an individual having grown up outside of the United States and 0 having grown up within the U.S.
Satisfaction is measured on a Likert scale from 1 being very dissatisfied to 5 being very satisfied
with the meal eaten that day. Finally, the LunchMates2 and LunchMates3 indicator variables
denote how many other people the individual ate with during the lunch period. LunchMates2
indicates that the person ate with 1-2 additional people whereas LunchMates3 indicates that the
person ate with 3 or more people. The reference category for both variables is individuals who
dined alone.  is included as a random error term.
Based on equation (1), we have developed several hypotheses. As waste is our primary
outcome of interest, we present our hypotheses for how each independent variable will affect
waste. First, we hypothesize that B1 < 0, meaning oval platters will result in less waste compared
to round plates. This is because the surface area of the oval platter is smaller than that of the
round plate, reducing the amount of food participants can fit onto their plate(s). Less selection
should ultimately result in less waste. The effect of B2 is ambiguous. We are unsure of differences
in waste at the individual level between locations. We expect B3 < 0, implying that participants
who went back for seconds wasted less food from their first helping than those who did not
return for more food. Selecting seconds would suggest the individual is still hungry after finishing
their first plate(s). Sex is known to impact food SCW as females tend to regulate their food
selection and intake more than men (Beardsworth et al., 2002). Therefore, we hypothesize the
female coefficient, B4, to be negative, meaning females waste more food than males. Females
are expected to consume less of their selected food compared to males, resulting in higher food
waste. It is unclear what sign B5 will take on as food SCW habits between international and
domestic students is unknown. International students may inherently select or consume more or
less food than domestic students. These students may be less familiar with dishes offered in the
dining halls. In this environment, students have indicated over selection of a variety of food as a
way to hedge their satisfaction (Nikolaus et al., 2018); one meal component is bound to be
satiable. Additionally, cultural norms, such as cleaning your plate, may dictate how much food
is wasted between domestic and international students. We hypothesize that the satisfaction
coefficient, B6, will be negative, implying waste decreases as meal satisfaction rating increases. A
higher rating of satisfaction should lead to greater consumption and less leftover edible food.

12
Lastly, we predict coefficients B7 and B8, will be positive, meaning that those who eat with others
will waste more food than when dining alone. The behavior of others has been seen to both
negatively and positively influence individual consumption and waste (Nikolaus et al., 2018).
Nikolaus et al. (2018) noted dining hall goers frequently reported serving more food than they
could consume. In the dining hall environment, socializing may take precedent to mindful eating
and individuals may reflect waste behaviors of peers.
Additionally, we examine the impact of the intervention on the selection of seconds.
Previous studies have found a significant increase in number of diners returning for seconds when
using smaller plates (Rolls et al., 2007); while others saw no difference between plate sizes
(Wansink & van Ittersum, 2013). In this study, it is important to know if there is a significant
difference between plate types. A significant increase in the number of participants taking
seconds when exposed to oval platters may change overall individual plate waste. Understanding
if oval platters increase the likelihood to select seconds may influence future study designs and
areas of focus. Due to our inability to directly measure seconds selected in this study, we estimate
a logistic regression to estimate the probability that an individual went back for seconds, based
on their survey response. We use the same predictor variables from equation (1), with the
exception of Seconds.

13
CHAPTER 4: RESULTS AND DISCUSSION

4.1 Participant Characteristics

There were 1,825 total observations collected. A total of 1,408 observations were kept
for analysis. Observations were excluded if: the participant did not return their plate (n=55);
submitted an incomplete survey (n=100); or only selected food using non-standard dishware
(e.g., only eating a bowl of soup; n=263).7 The most recent enrollment statistics from The
University of Illinois at Urbana-Champaign reported that 7,609 new Freshman students enrolled
in the Fall of 2018, making up about 22% of the undergraduate population (University of Illinois
System, 2018). The vast majority of these new students will live in university housing, which
requires them to purchase a meal plan.
Table 2 presents the characteristics of the sample. There were no significant differences
in sex or number of international students between the intervention and control groups. Our
sample consisted of 40.20% female and 85.16% domestic students. The proportion of female
students is slightly lower than the campus estimate of 46%, but the proportion of domestic
students is in line with campus estimate of 85% (University of Illinois Urbana-Champaign, 2018).
Dining administration estimated an average of 2,186 diners per weekday at Ikenberry Commons
Dining Hall (IKE) and 652 diners per weekday at Pennsylvania Avenue Residence Dining Hall (PAR)
during lunch hours (Etchison, personal communication, Dec. 11, 2018). Most diners reported
being very (38.00%) to somewhat (48.30%) satisfied with their lunch, while only 13.71% reported
feeling neutral, somewhat, or very dissatisfied. Approximately 38.35% of participants ate with
one or two other people and 10.37% reported eating with 3 or more people. Of the participants
exposed to oval platters, 30.92% returned for seconds, while 21.58% of those with round plates
returned for seconds.

7There were no significant differences between included and excluded observations in terms of sex or country of
residence. After excluding observations that included any non-standard dishware, there was no significant
difference in selection between those that returned their plates and those that did not. In addition, students who
did not return a complete survey were not statistically different from the sample in terms of selection.

14
4.2 Impact of Change in Plate Size and Shape
Table 3 presents the overall sample averages for food selection, consumption, and waste.
The average amount selected per participant was 349.6 g, with consumption equaling 291.1 g for
a total average waste of 58.5 g per person. Mean percent of plate waste was 16.0%. Table 4
shows the difference in mean SCW between round and oval platters per person. In this analysis,
we see a 15.4% decrease in selection when oval platters were used. Mean selection with round
plates was 377.4 g whereas oval platters saw an average of 319.4 g per person (p < 0.001).
Similarly, there is a significant decline (8.3%) in consumption. Consumption was at 303.1 g per
person with round plates, but only 278.0 g with oval platters (p < 0.001). On round plate days,
the average participant wasted 74.3 g of their selected meal while oval platter days resulted in a
mean waste of 41.4 g; a significant decrease in grams of food wasted (p < 0.001). Since selection,
consumption, and waste were all lower with oval platters, it is important to determine if the rate
of waste, or percent waste, is also lower. The percent waste for round plates was 19.1% and
12.7% for oval platters (p < 0.001). This suggests a smaller surface area prevents students from
selecting more food than they can realistically consume. These results are in line with findings by
Van Ittersum & Wansink (2011) and DiSantis et al. (2013). Both studies found that larger plates
resulted in greater selection and consumption but did not report on waste. Additionally, a waste
level of 12.7% is more in line with acceptable levels of waste in the National School Lunch
Program (Buzby & Guthrie, 2002).
Using a two-sample t-test with unequal variances, we found a significant difference in
SCW between locations (p < 0.001). For an in-depth analysis, we separated mean SCW by location
(see table 5). In the IKE location, oval platters resulted in significantly lower selection, waste, and
percent waste than round plates (all p < 0.05). At PAR, oval platters significantly reduced all
outcomes of interest compared to round plates (all p < 0.01).
Table 6 presents the linear regression estimates for SCW. For selection, we see that oval
platters are estimated to reduce selection by 55.7 g (p < 0.001) per participant, all else constant.
Additionally, selecting seconds, sex, and satisfaction level all significantly influence selection. The
selection of seconds decreased initial selection weight by 26.1 g, on average (p = 0.001). Females

15
selected 48.8 g less than males, on average (p < 0.001), while satisfaction increased overall
selection (p = 0.001).
Similar to selection, consumption also declines with the introduction of oval platters (-
18.6 g per student; p = 0.003). Going back for seconds, sex, and satisfaction all significantly
influence consumption in addition to location. Returning for seconds is estimated to decrease
consumption by 19.0 g, on average (p = 0.010). Females are estimated to consume an average of
60.0 g (p < 0.001) less than males. Individuals are estimated to consume 26.9 g (p < 0.001) more
with each unit increase in meal satisfaction. Additionally, diners at the IKE location are estimated
to consume 35.3 g (p < 0.001) more than those who eat at PAR, all else constant.
Lastly, waste and percent waste see a significant decline with the oval platter intervention
(-37.1 g and -7.2%, respectively; both p < 0.001). Location, sex, satisfaction, and eating with 1-2
others all significantly impact waste and percent waste. Individuals dining in the IKE waste 31.8 g
(p < 0.001) less than those in PAR; a decline in percent of meal wasted of 6.7% (p < 0.001).
Females are estimated to waste 11.2 g (p = 0.003) more than males; approximately 5.3% (p <
0.001) more of their meal, on average. As hypothesized, those with a higher rating of satisfaction
are less likely to waste food (-11.9 g and -4.1% per unit increase in satisfaction; both p < 0.001).
Dining with one or two others is estimated to increase waste by 11.1 g or 2.7% (both p = 0.004)
relative to eating alone. Unlike selection and consumption, waste and percent waste are not
significantly influenced by returning for seconds.

4.3 Selecting Seconds

Using a chi-squared test, we found a significant difference in those returning for seconds
between round plate and oval platter groups (p < 0.001). To further explore this, we ran a logistic
regression to estimate the odds of selecting seconds (see table 7). The odds of going back for
seconds for participants using oval platters is 1.48 (p = 0.002) times higher than those using round
plates. Other significant variables that change the likelihood of selecting seconds are location,
sex, international student, satisfaction, and dining with 3 or more people. Diners at IKE were less
likely to select seconds than diners at PAR (OR=0.55; p < 0.001). Females were less likely to select
seconds than males (OR=0.43; p < 0.001). Individuals that grew up in the United States were less

16
likely to select seconds than International students (OR=0.49; p < 0.001). Each unit increase in
satisfaction increases the probability of going back for seconds (OR=1.20; p = 0.026). Lastly, those
that dine with three or more people are 1.66 (p = 0.014) times more likely to select seconds.
Even though we could not directly observe those selecting and wasting seconds, we may
estimate the total amount wasted from the number of participants who indicated they selected
seconds and the average amount of food wasted by plate shape. For this calculation, we assumed
the average amount of plate waste is the same from the first serving to the second.8 Those using
round plates wasted an average of 74.3 g with their first plate (table 4), and 21.58% (n=158) of
round plate participants indicated that they returned for seconds. Those using oval platters
wasted an average of 41.4 g with their first plate (table 4), and 30.92% (n=209) of oval platter
participants indicated they returned for seconds. Table 4 reveals that individuals with oval
platters waste 32.9 g less than those with round plates. Multiplying average waste by the total
number of observations ((74.3 g*732 round plate observations) + (41.4 g*675 oval platter
observations)) plus the number of participants selecting seconds based on plate shape ((74.3
g*158 round plate seconds) + (41.4 g*209 oval platter seconds)) gives us total participant waste
(66,124.4 g for round plates; 36,585.6 g for oval platters). This would increase average plate
waste to 90.3 g for round plates and 54.2 g for oval platters, meaning oval platters still result in
less waste (36.1 g) than round plates. Thus, the increased proportion of diners selecting seconds
with oval platters does not appear to offset the waste reduction gains.

4.4 Limitations
While this study makes several contributions to the food waste literature, there are some
limitations that should be acknowledged. First, many participants commented on the difference
in plate shape and size between weeks and locations, noting that they could not fit as much food
on the oval platters as the round plates (figure 1). The difference in plate shape and date of
implementation may have heightened diners’ awareness of the intervention and study
intentions. Similar participant awareness found by Rolls et al. (2007) did not impact the study’s

8This is likely a conservative estimate, as one would expect waste to decrease with the second serving, as diners
are more aware of their current state of satiation and preferred foods.

17
results, so there may not be need for concern. However, to mitigate any negative feedback from
diners, the phasing in of an oval platter after university breaks or between academic years may
produce a more favorable reception with students after not having repeated exposure to other
dishware sizes prior to oval platter implementation. Furthermore, dining hall staff felt as though
there was more food waste found on the floor and on tables when oval platters were used. Due
to the oblong shape, diners may have found it more difficult to carry platters without spillage.
This negative externality was beyond the scope and research capacity of this study but should be
considered in future work.
Second, there was low participation in the study overall. Based on the number of
observations per week and the average number of diners per week in each location, we were
able to estimate population coverage. Although PAR had higher coverage rates than IKE, overall
coverage was still low (range of 8.5% to 11.5% in PAR; 2.9% to 4.0% in IKE). Clearly, higher
coverage rates would have been preferred, but due to participant recruiting protocol and
physical limitations in both dining halls, larger participation was unachievable. Alternative
methods to increase participation may be more useful for attaining higher coverage rates in
future studies.
Additionally, the field nature of this study resulted in some issues that were beyond the
researchers’ control. As noted in the methods, there were two types of round plates, with one
much heavier than the other but visually identical. While the prevalence of these heavier plates
was likely low (estimated to be one in five), the research team proceeded with an analysis
thought to best represent the data in estimating SCW for round plate observations. This could
overstate waste in some cases. Further, this study did not account for SCW of second helpings.
Diners returning for seconds may be beneficial in reducing plate waste in dining halls. Students
returning for seconds may be more attune to their levels of satiation and enjoyment of the foods
served and would only select seconds that would fulfil their needs and preferences. It is also
important to note there may be selection bias in the sample as students opted to participate in
the study. Diners electing to participate in the study may have inherent waste-related
characteristics that this study did not measure, such as anti-waste attitudes and/or various levels

18
of nutrition education, relative to those who declined to participate. Each issue should be
considered in future research.
Lastly, this study did not analyze shifts in meal components selected. Wansink and Just
(2013) noted a decrease in salad selection with trayless dining. The change in plate shape/size
may lead diners to make trade-offs between different dishes. Plate photos allow us to investigate
shifts in food choices and subsequent dietary quality in future research.

19
CHAPTER 5: CONCLUSION

Students continue to disregard personal responsibility when it comes to food waste in

dining halls (Nikolaus et al., 2018). With inattentive patrons, dining administration should
implement strategies that directly reduce food waste. This study shows that introducing smaller
oval platters in a university dining setting can significantly reduce plate waste compared to using
larger round plates. This study confirms transitioning from round plates to smaller oval platters
reduces individual selection, consumption, and waste in all-you-care-to-eat university dining
halls. Of the three categories for waste reduction interventions in university dining halls, findings
from this study have a large impact at a moderate cost as compared to education campaigns,
standard serving size alterations, or pricing strategy adjustments. Future research should study
the impacts of smaller size and alternative shaped plates on selection, consumption, and waste
of seconds. Changes in dietary choices and unintended intervention consequences, like spilled
food, should also be investigated. Likewise, long-term effects of this intervention are unknown.
Supplementary research should test this intervention over a long period of time.

20
 FIGURES

Round Oval
9” x 9” 9.75” x 7.75”
Surface Area: 63.62 in2 Surface Area: 59.32 in2

Figure 1. Main Dish Plates

21
Figure 2. Survey and Weight Record

22
TABLES

Table 1.
Crossover Trial Design.
Week Dates Location Plate Type

Week 1 Sept. 17-21 PAR Oval

Week 2 Sept. 24-28 IKE Round

Week 3 Oct. 15-19 PAR Round

Week 4 Oct. 22-26 IKE Oval

23
Table 2.
Descriptive Statistics (N = 1,408)

N %
Location
IKE 757 53.76
PAR 651 46.24
Plate Type
Round 732 51.99
Oval 676 48.01
Week
1 (PAR, Oval) 374 26.56
2 (IKE, Round) 455 32.32
3 (PAR, Round) 277 19.67
4 (IKE, Oval) 301 21.45
Day
Mon. 279 19.82
Tues. 339 24.08
Wed. 236 16.76
Thurs. 334 23.72
Fri. 220 15.63
Meal Satisfaction
Very satisfied 535 38.00
Somewhat satisfied 680 48.30
Neutral 139 9.87
Somewhat dissatisfied 46 3.27
Very dissatisfied 8 0.57
Lunch Mates
Ate alone 722 51.28
Ate with 1-2 people 540 38.35
Ate with 3 or more people 146 10.37
Seconds a
Yes 367 26.07
No 1,041 73.93
Sex
Male 842 59.80
Female 566 40.20
Country of Residence
International 209 14.84
United States 1,199 85.16
a Participant indicated they returned for additional food.

24
Table 3.

Average selection, consumption, and waste per participant; combined round plates and oval
platters. (N = 1,408)
Mean SD Min Max
Total Selection (g) 349.6 ±134.5 69.8 1186.8
Total Consumption (g) 291.1 ±123.5 12.0 1075.0
Total Waste (g) 58.5 ±72.4 0 461.8
Pct. Waste (%) 16.0 ±0.5 15.1 16.9

Table 4.
Two-sample t-test with unequal variances. Difference in average selection, consumption, and waste
between round plates and oval platters.
Round Plates Oval Platters
Outcome Mean Std. Dev. Mean Std. Dev. p-value
Selection (g) 377.4 ±137.9 319.4 ±124.1 p<0.001
Consumption (g) 303.1 ±127.5 278.0 ±117.7 p<0.001
Waste (g) 74.3 ±80.1 41.4 ±58.4 p<0.001
Pct. Waste (%) 19.1 ±18.2 12.7 ±16.2 p<0.001

25
Table 5.
Two-sample T-Test with unequal variances. Difference in average selection, consumption, and
waste between round plates and oval platters by location.
Round Plates Oval Platters
Outcome Mean Std. Dev. Mean Std. Dev. p-value
IKE
Selection (g) 366.0 ±136.6 344.0 ±121.9 p=0.022
Consumption (g) 314.4 ±135.6 302.7 ±118.3 p=0.207
Waste (g) 51.5 ±54.6 41.4 ±54.0 p=0.012
Pct. Waste (%) 14.7 ±15.1 12.0 ±14.1 p=0.015
PAR
Selection (g) 396.3 ±138.0 299.6 ±122.4 p<0.001
Consumption (g) 284.4 ±110.5 258.1 ±113.6 p=0.003
Waste (g) 111.9 ±99.1 41.4 ±61.7 p<0.001
Pct. Waste (%) 26.2 ±20.4 13.3 ±17.7 p<0.001

26
Table 6.
OLS Regression estimates for average selection, consumption, and waste per participant (N =
1,408)
Variable Selection (g) Consumption (g) Waste (g) Waste (%)
(Std. Error) (Std. Error) (Std. Error) (Std. Error)
Intercept 346.6 210.6 136.0 36.8
(21.8) (19.7) (11.5) (2.8)
a
Oval ‒ 55.7*** ‒ 18.6** ‒ 37.1*** ‒ 7.2***
(7.0) (6.3) (3.7) (0.9)
Location b 3.5 35.3*** ‒ 31.8*** ‒ 6.7***
(7.1) (6.4) (3.7) (0.9)
c
Seconds ‒ 26.1*** ‒ 19.0** ‒7.1 ‒ 1.0
(8.1) (7.3) (4.3) (1.0)
d
Female ‒ 48.8*** ‒ 60.0*** 11.2** 5.3***
(7.2) (6.5) (3.8) (0.9)
e
Int’l Student ‒ 15.2 ‒15.6 0.11 0.6
(9.8) (8.8) (5.2) (1.2)
Satisfaction 15.0*** 26.9*** ‒ 11.9*** ‒ 4.1***
(4.3) (3.9) (2.3) (0.6)
f
Lunch Mates 2 10.9 ‒0.24 11.1** 2.7**
(7.3) (6.6) (3.8) (0.9)
f
Lunch Mates 3 3.5 ‒0.01 3.5 1.0
(11.8) (10.6) (6.2) (1.5)
* Significant at 0.05 level
**Significant at 0.01 level
***Significant at 0.001 level
a Effect relative to round plate
b Effect relative to PAR
c Effect relative to not returning for additional food items
d Effect relative to male
e Effect relative to participants who did not grow up in the United States
f Effect relative to eating alone

27
Table 7.
Odds of participant choosing seconds (N = 1,408)
Variables OR CI 95%
(Std. Error) Lower Upper
Intercept 0.38 0.17 0.87
(0.16)
a
Oval 1.48** 1.15 1.91
(0.19)
Location b 0.55*** 0.43 0.57
(0.07)
Female c 0.43*** 0.33 0.57
(0.06)
Int’l Student d 0.49*** 0.36 0.68
(0.08)
Satisfaction 1.21* 1.02 1.42
(0.10)
Lunch Mates 2 e 1.17 0.89 1.52
(0.16)
e
Lunch Mates 3 1.66* 1.11 2.49
(0.34)
*Significance level at 0.05
** Significant at 0.01 level
***Significant at 0.001 level
a Effect relative to round plate
b Effect relative to PAR
c Effect relative to male
d Effect relative to participants who did not grow up in the United States
e Effect relative to eating alone

28
REFERENCES

Beardsworth, A., Brynan, A., Keil, T., Goode, J., Haslam, C., & Lancashire, E. (2002). Women ,
men and food : The significance of gender for nutritional attitudes and choices. BRITISH
FOOD JOURNAL, 104(7), 470–491. https://doi.org/10.1108/00070700210418767
Bellemare, M., Cakir, M., Peterson, H. H., Novak, L., & Rudi, J. (2017). On The Measurement of
Food Waste. AMERICAN JOURNAL OF AGRICULTURAL ECONOMICS, 99(5), 1148–1158.
https://doi.org/10.1093/ajae/aax034
Birney, C. I., Franklin, K. F., Davidson, F. T., & Webber, M. E. (2017). An assessment of individual
foodprints attributed to diets and food waste in the United States An assessment of
individual foodprints attributed to diets and food waste in the United States.
Buzby, J. C., & Guthrie, J. F. (2002). Plate Waste in School Nutrition Programs Final Report to
Congress. Food Assistance & Nutrition Research Program.
Buzby, J. C., Wells, H. F., & Hyman, J. (2014). The Estimated Amount , Value , and Calories of
Postharvest Food Losses at the Retail and Consumer Levels in the United States, (121).
Campanhola, C., & Pandey, S. (Eds.). (2019). Food and Agricultural Systems at a Crossroads : An
Overview. In Sustainable Food and Agriculture: An Integrated Approach (1st Editio, pp. 3–
10). Academic Press. https://doi.org/10.1016/B978-0-12-812134-4.00001-7
Canning, P., Charles, A., Huang, S., Polenske, K. R., & Waters, A. (2010). Energy Use in the U . S .
Food System, (94).
Christoph, M. J., & Ellison, B. (2017). A Cross-Sectional Study of the Relationship between
Nutrition Label Use and Food Selection, Servings, and Consumption in a University Dining
Setting. Journal of the Academy of Nutrition and Dietetics, 117(10), 1528–1537.
https://doi.org/10.1016/j.jand.2017.01.027
Christoph, M. J., Ellison, B. D., & Nehrling Meador, E. (2016). The Influence of Nutrition Label
Placement on Awareness and Use among College Students in a Dining Hall Setting. Journal
of the Academy of Nutrition and Dietetics, 116(9), 1395–1405.
https://doi.org/10.1016/j.jand.2016.05.003
Cioffi, C. E., Levitsky, D. A., Pacanowski, C. R., & Bertz, F. (2015). A nudge in a healthy direction.

29
 The effect of nutrition labels on food purchasing behaviors in university dining facilities.
Appetite, 92, 7–14. https://doi.org/10.1016/j.appet.2015.04.053
Costello, C., Birisci, E., & McGarvey, R. G. (2015). Food waste in campus dining operations :
Inventory of pre- and post-consumer mass by food category , and estimation of embodied
greenhouse gas emissions. Renewable Agriculture and Food Systems, 31(3), 191–201.
https://doi.org/10.1017/S1742170515000071
DiSantis, K. I., Birch, L. L., Davey, A., Serrano, E. L., Zhang, J., Bruton, Y., & Fisher, J. O. (2013).
Plate size and children’s appetite: Effects of larger dishware on self-served portions and
intake. American Academy of Pediatrics, 131(5), e1451–e1458.
https://doi.org/http://dx.doi.org/10.1542/peds.2012-2330
Ellison, B., Muth, M. K., & Golan, E. (2019). Featured Article Opportunities and Challenges in
Conducting Economic Research on Food Loss and Waste. Applied Economic Perspectives
and Policy, 41(1), 1–19. https://doi.org/10.1093/aepp/ppy035
Ellison, B., Savchenko, O., Nikolaus, C. J., & Duff, B. R. L. (2019). Every plate counts: Evaluation
of a food waste reduction campaign in a university dining hall. Resources, Conservation &
Recycling, 144(December 2018), 276–284.
https://doi.org/10.1016/j.resconrec.2019.01.046
Etchison, T. (2018). Personal Communication.
Etchison, T., & Van Liew, D. (2018). Personal Communication.
Foderaro, L. W. (2009). Without Cafeteria Trays, Colleges Find Saving. The New York Times.
Retrieved from https://www.nytimes.com/2009/04/29/nyregion/29tray.html
Freedman, M. R., & Brochado, C. (2009). Reducing Portion Size Reduces Food Intake and Plate
Waste. Obesity Journal, 18(9), 1864–1866. https://doi.org/10.1038/oby.2009.480
Graham-Rowe, E., Jessop, D. C., & Sparks, P. (2013). Identifying motivations and barriers to
minimising household food waste. Resources, Conservation & Recycling, 84, 15–23.
Gunders, D. (2012). Wasted: How America Is Losing Up to 40 Percent of Its Food from Farm to
Fork to Landfill Author, (august).
Heikkilä, L., Reinikainen, A., Katajajuuri, J., Silvennoinen, K., & Hartikainen, H. (2016). Elements
affecting food waste in the food service sector, 56, 446–453.

30
 https://doi.org/10.1016/j.wasman.2016.06.019
Jovanovic, N., Katare, B., & Wetzstein, M. (2018). Take all You Want, but Eat all You Take:
Effectiveness of a Financial Incentive on Individual Food Waste.
Kim, K., & Morawki, S. (2012). Quantifying the Impact of Going Trayless in a University Dining
Hall (Abstract). Journal of Hunger & Environmental Nutrition2, 7(4), 482–486.
https://doi.org/10.1080/19320248.2012.732918
Kim, W. G., Yen, N. N., & Kim, Y. (2009). Influence of institutional DINESERV on customer
satisfaction, return intention, and word-of-mouth. International Journal of Hospitality
Management, 28, 10–17. https://doi.org/10.1016/j.ijhm.2008.03.005
Kubik, M. Y., Lytle, L. A., Hannan, P. J., Perry, C. L., & Story, M. (2003). The association of the
school food environment with dietary behaviors of young adolescents. American Journal of
Public Health, 93(7), 1168–1173. Retrieved from
http://www.ncbi.nlm.nih.gov/pubmed/12835204
LizMartins, M., Cunha, L. M., Rodrigues, S. S. P., & Rocha, A. (2014). Determination of plate
waste in primary school lunches by weighing and visual estimation methods : A validation
study. Waste Management, 34(8), 1362–1368.
https://doi.org/10.1016/j.wasman.2014.03.020
National Center for Education Statistics. (2018). Total 12-month enrollment in degree-granting
postsecondary institutions, by control and level of institution and state or jurisdiction:
2014-15 and 2015-16. Digest of Education Statistics.
Nikolaus, C. J., Nickols-Richardson, S. M., & Ellison, B. (2018). Wasted food: A qualitative study
of U.S. young adults’ perceptions, beliefs and behaviors. Appetite, 130(July), 70–78.
https://doi.org/10.1016/j.appet.2018.07.026
Rajbhandari-Thapa, J., Ingerson, K., & Lewis, K. H. (2018). Impact of trayless dining intervention
on food choices of university students. Archives of Public Health, 76(1), 61.
https://doi.org/10.1186/s13690-018-0301-5
Reynolds, C., Goucher, L., Quested, T., Bromley, S., Gillick, S., Wells, V. K., … Jackson, P. (2019).
Review: Consumption-stage food waste reduction interventions – What works and how to
design better interventions. Food Policy, 83(February), 7–27.

31
 https://doi.org/10.1016/j.foodpol.2019.01.009
Rolls, B. J., Roe, L. S., Halverson, K. H., & Meengs, J. S. (2007). Using a smaller plate did not
reduce energy intake at meals. Appetite, 49, 652–660.
https://doi.org/10.1016/j.appet.2007.04.005
Sarjahani, A., Serrano, E. L., & Johnson, R. (2009). Food and Non-Edible, Compostable Waste in
a University Dining Facility (Abstract). Journal of Hunger & Environmental Nutrition, 4(1),
95–102. https://doi.org/10.1080/19320240802706874
Scourboutakos, M. J., Mah, C. L., Murphy, S. A., Mazza, F. N., Barrett, N., Mcfadden, B., & Abb,
M. R. L. (2017). Testing a Beverage and Fruit/Vegetable Education Intervention in a
University Dining Hall. Journal of Nutrition Education and Behavior, 49(6), 457–465.
https://doi.org/10.1016/j.jneb.2017.02.003
Soares Pinto, R., Machado dos Santos Pinto, R., Fochat Silva Melo, F., Santos Campos, S., &
Marques-dos-santos Cordovil, C. (2020). A simple awareness campaign to promote food
waste reduction in a University canteen. Waste Management, 76(2018), 28–38.
https://doi.org/10.1016/j.wasman.2018.02.044
Thaler, R. H., & Sunstein, C. R. (2008). Nudge: Improving Decisions About Health, Wealth, and
Happiness. Penuin Books.
Thiagarajah, K., & Getty, V. M. (2012). Impact on Plate Waste of Switching from Tray to a
Trayless Delivery System in a University Dining Hall and Employee Response to the Switch.
JOURNAL OF THE ACADEMY OF NUTRITION AND DIETETICS, 113(1), 141–145.
https://doi.org/10.1016/j.jand.2012.07.004
Thyberg, K. L., & Tonjes, D. J. (2016). Drivers of food waste and their implications for sustainable
policy development. “Resources, Conservation & Recycling,” 106, 110–123.
https://doi.org/10.1016/j.resconrec.2015.11.016
United States 2030 Food Loss and Waste Reduction Goal. (2015). United States Environmental
Protection Agency.
United States Energy Information Administration. (2018). Use of Energy in the United States
Explained: Energy Use in Industry. Retrieved March 10, 2019, from
https://www.eia.gov/energyexplained/index.php?page=us_energy_industry

32
University of Illinois System. (2018). University of Illinois System Data Enrollment. Retrieved
from https://www.uillinois.edu/data/enrollment
University of Illinois Urbana-Champaign. (2018). Fall 2018 Statistical Abstract of First Day of
Enrollment. University of Illinois at Urbana-Champaign,. Retrieved from
http://www.dmi.illinois.edu/stuenr/abstracts/FA18_one.htm
Van Ittersum, K., & Wansink, B. (2011). Plate Size and Color Suggestibility : The Delboeuf
Illusion’s Bias on Serving and Eating Behavior. Jounal of Consumer Research, 39(August),
215–228. https://doi.org/10.1086/662615
Vermote, M., Versele, V., Stok, M., Mullie, P., D’Hondt, E., Deforche, B., … Deliens, T. (2018).
The effect of a portion size intervention on French fries consumption, plate waste, satiety
and compensatory caloric intake: An on-campus restaurant experiment. Nutrition Journal,
17(1), 1–12. https://doi.org/10.1186/s12937-018-0352-z
Wansink, B., & Just, D. R. (2013). Trayless cafeterias lead diners to take less salad and relatively
more dessert. Public Health Nutrition, 18(9), 1535–1536.
https://doi.org/10.1017/S1368980013003066
Wansink, B., & van Ittersum, K. (2013). Portion Size Me : Plate-Size Induced Consumption
Norms and Win-Win Solutions for Reducing Food Intake and Waste, 19(4), 320–332.
https://doi.org/10.1037/a0035053
Wansink, B., van Ittersum, K., & Painter, J. E. (2006). Ice Cream Illusions. Bowls, Spoons, and
Self-Served Portion Sizes. American Journal of Preventive Medicine, 31(3), 240–243.
https://doi.org/10.1016/j.amepre.2006.04.003
Whitehair, K. J., Shanklin, C. W., & Brannon, L. A. (2013). Written Messages Improve Edible
Food Waste Behaviors in a University Dining Facility. JAND, 113(1), 63–69.
https://doi.org/10.1016/j.jand.2012.09.015

33
APPENDIX A: DATA COLLECTION PROTOCOL

“The Impact of Dining Environment Factors on Food Choice Behavior” Protocol

September 17 - 28 & October 15 - 26

Purpose of the study is to test if a change in plate shape and size decreases plate waste at two
University of Illinois dining halls. Further research on nudges within dining halls to reduce plate
waste. Weighing plates before and after consumption will determine the amount students are
wasting. Intervention weeks will be in September and October.

Study Locations: Pennsylvania Avenue Residence Dining Hall 9/17-21 & 10/15-19
Ikenberry Commons Dining Hall 9/24-28 & 10/22-26

PAR 9/17-21 Oval Plates

IKE 9/24-28 Round Plates
PAR 10/15-19 Round Plates
IKE 10/22-26 Oval Plates

Start of lunch Set Up (10:45-11:00):

1. Record clock-in time [see time sheet]
2. Set out signs near dining hall entrance, select dining stations, and data collection tables
(pre- and post-weight tables)

Pre-Weight Table Post-Weight Table

o Signage set up o Set up table near dish return;

o Scale out (set to grams; only use include gloves
5 kg scales) o Trashcan near for napkins &
o Tablet ready for photographs other non-food waste
o Blank surveys and study info o Scale out (set to grams; can use
sheets available 1 kg or 5 kg scales)
o Be sure to wear gloves o Tablets ready for photographs
o Box/folder to store completed
surveys; additional study info
sheets available
o Be sure to wear gloves

End of lunch Break Down (1:45pm-2:00pm):

1. Stop giving out surveys between 1:30 and 1:45pm to allow final students to complete
surveys

34
 2. Make sure there are no remaining students with surveys
3. Wipe down tables and scales
4. Set the tables and signage aside
5. Record clock-out time [see time sheet]
Experiment Procedure:

Pre-Weight Table Post-Weight Table

1. Approach and recruit students waiting in 1. Identify and assist participants that have filled
Role of line or who have full plates. Be sure to out the entire survey and direct them to
Investigators recruit students before they sit down and return their plates
start eating
2. Weigh participants’ plates and record on
2. Weigh participant’s plates and record on survey. Once done, place plates on dish
survey conveyer

3. Assist with weighing while busy and compare

dining site’s stated menu to actual menu on-
site documenting any discrepancies

1. When a student approaches, pick up a 1. Collect survey from student and take their
survey plate
a. Make sure it has a survey ID a. Participants are free to leave once we
2. Either have student place their plate on have taken their plate
scale or if wearing gloves place their plate b. Zero scale before weighing by
on the scale for them pressing ON/OFF/TARE button
a. Zero scale before weighing by 2. Remove any non-edible food waste from
pressing ON/OFF/TARE button their plates (napkins or paper products,
3. Record the weight on the back of the survey bones, peels, etc.)
Protocol for and circle the type of dish weighed 3. Weigh the plate and record the weight on the
Weighing a. There is no need to capture drinks back of the survey
4. Capture a picture with all edges of plate a. There is no need to capture drinks
within the screen making sure it includes 4. Capture a picture with all edges of plate
a. Survey ID number within the screen making sure it includes
b. Weight of the plate a. Survey ID number
c. All items on the plate b. Weight of the plate
5. Return student’s plate and give them the c. All items remaining on the plate
survey for them to fill out while eating; 5. Take student’s plate to dish return
explain where to return plate and survey 6. Place completed survey in folder

35
Notes and Reminders:

If a student approaches with multiple dishes, only weigh the “main entree” dishes — these
should be the round plates or oval platters. In PAR, they also have stir-fry bowls, so you can
circle the “Other” dish type and specify bowl on the back of the survey. Smaller dishes like
soup/cereal bowls or dessert plates do not need to be included in the study. If a student has
multiple main plates, weigh and photograph both plates individually and record on the back of
survey. Indicate there were two plates by writing X2 next to main dish identifier on back of
survey (Dish1 and Dish 2 information, respectively). Make sure to circle both dish types.

Please request that students return their dishes to the “Drop-Off Table” (post-weight table)
before getting seconds. You can also inform students that dishes from second helpings do not
need to be documented.

Do not inform students of what we are measuring. IF they ask, inform them we are helping the
dining hall keep track of what items students like eating, or determining menu changes, etc.
Keep it broad. If they ask specifically about weighing, you can tell them we are trying to help
dining improve their menu forecasting and planning.

It is extremely important to capture participants after consumption. Be sure to monitor

students putting their plates away. If they have a survey in-hand they need to give their plate to
the Post-weight table.

Lead student needs to confirm set up of experiment each day. Make sure the correct plates are
out and are replacing the larger bowls.

**Silver bus will take you from Mumford to PAR Dining Hall**
**#22 Illini bus takes you from Mumford to 4th and W Gregory Dr. close to Ike Dining Hall**

36