JOURNAL REVIEW

Untuk Memenuhi Salah Satu Tugas Mata Kuliah Journal Review

DISUSUN OLEH :
Inas Humaid (P1337431215081)
Nur Amalia Faidiyati (P1337431216004)
Sophia Danarena (P1337431216012)
Vina Fitrotun Nisa (P1337431216021)
Ajeng Cahya Pramesti (P1337431216028)
Christya Agustina Widianingrum (P1337431216039)

PROGRAM STUDI D IV GIZI

KELAS REGULER A SEMESTER V
POLITEKNIK KESEHATAN KEMENKES SEMARANG
2018 / 2019
 CHECKLIST CONSORT PADA PENELITIAN EKSPERIMEN

Topik Nomor Checklist Dilaporkan

pada
halaman
Title and 1a Sudah teridentifikasi sebagai randomized controlled trial dalam judul 1
Abstrack artikel penelitian

1b Sudah menuliskan ringkasan dari penelitian berupa abstrak yang 1

meliputi latar belakang, tujuan, desain penelitian, hasil dan kesimpulan
serta kata kunci dari penelitian tersebut. Penulisan abstrak sudah
terstruktur.

Introduction 2a Sudah menjelaskan latar belakang masalah secara ilmiah dan rasional
Background yang didasarkan pada referensi jurnal-jurnal penelitian sebelumnya 1-2
And Objectives seperti BMC Public Health, Diabetologia, American Journal Clinical
Nutrition, dll. Latar belakang penelitian dijelaskan di bab Pendahuluan
paragraf 1 sampai dengan paragraf 3.

Sudah menuliskan tujuan penelitian dalam bab Pendahuluan paragraf 4

2b yaitu untuk mengevaluasi kemungkinan perbedaan pengaruh pada 2
massa lemak visceral dan komposisi tubuh, serta variabel biokimia dan
klinis, termasuk komponen dari sindrom metabolik yang diakibatkan
dari pemberian diet rendah lemak tinggi karbohidrat (LFHC) dan
sangat tinggi lemak rendah karbohidrat (VHFLC) pada kalori yang
sama dan pembatasan energi sederhana pada jumlah protein dan asam
lemak tidak jenuh rantai panjang (PUFAs) yang sama.
Methods 3a Sudah mendeskripsikan desain percobaan yang digunakan yaitu
Trial design parallel design pada bab Metode di sub bab Desain Percobaan 2
paragraf 1. Pencapaian alokasi acak dengan alternating treatments by
days .

Tidak ada alasan penentuan kriteria kelayakan peserta penelitian hanya

menyebutkan adalah orang dewasa.. Kriteia inklusi disebutkan pada
3b bab Metode, sub bab Mempelajari sampel dan pengaturan paragraf 1.
2

Participant 4a Belum mencantumkan kriteria peserta penelitian, tetapi mencantumkan 6 dan 2-3
jumlah peserta penelitian yaitu sebesar 443 klien yang terbagi menjadi
4b
205 klien under the Front treatment dan 238 klien back treatment.

Tempat penelitian sudah dijelaskan pada bab Metode subbab Peserta

dan pengaturan penelitian paragraf 1, yaitu penelitian dilakukan di 2
sebuah pantry makanan di Negara Bagian New York. Sedangkan
waktu penelitian dijelaskan pada bab experiment design yaitu pada 28
Oktober 2014 hingga 18 November 2014 yang diselenggarakan pada
setiap hari Selasa jam 1:00 to 2:00 siang.
Interventions 5 Intervensi yng diberikan pada klien sudah dijelaskan secara secara urut
dan rinci serta bagaiman, kapan, dan dimana pemberian intervensi
pada bab metode dan experimen design.Intervensi yang diberikan
adalah :
1. Intervensi dilakukan di pantry makanan di New York dengan
berafiliasi bersama anggota gereja yang menjadi bagian Feeding
2
America.
2. Gereja menyediakan pantry makanan setiap hari Selasa pukul
13.00 hingga 14.00. Pantry dijalankan oleh 15 – 20 relawan, yang
mana banyak dari relawan yang merupakan paroki gereja. Serta
pantry melayani 7269 klien dengan total 150 klien (orang dewasa) 3
yang terlayani pada setiap sesinya.
3. Klien berjalan dan dapat memesan pada meja yang dimana relawan
menawarkan produk.
4. Pada setiap meja, klien diijinkan untuk mengambil lagi produk
tergantung dengan ketersediaan produk dan perkiraan jumlah
peserta, Namun produk tidak dapat dipertukarkan pada semua 3
kategori seperti klien tidak diijinkan untuk melewati makanan
penutup untuk mendapatkan mendapatkan buah yang lebih.
5. Intervensi difokuskan pada meja makanan penutup (dessert) yang
mana klien dpat memilih satu dessert dari produk yang tersedia
6. Intervensi yang diberikan pada uji experimental adalah memesan
dan mengemas (pengemasan) 3
7. Sebelum dilakukan intervensi, harus menerima persetujuan dari
lembaga sponsor
8. Intervensi pemesanan, protein bar ditempatkan pada sisi belakang
atau bagian akhir di makanan penutup (dessert) dan minggu
3
berikutnya protein bars ditempatkan pada bagian depan baris
dessert. Intervensi dilakukan pengulangan selama empat kali
berturut-turut (28 Oktober – 18 November 2014)
3
Intervensi pengemasan, pantry menawarkan produk dengan
kemasan asli yaitu dalam bentuk box. Namun sebagai alternatif,
pantry menawarkan protein bar yang dibungkus sendiri dengan
kantong plastik bening. Pada sesi pertama yaitu dua kali berturut
turut ( 28 Oktober dan 4 November) dengan kemasan box. Lalu
pada sesi terakhir yaitu dua kali berturut-turut ( 11 dan 18
3
November 2014) tanpa kemasan box

Outcomes 6a Sudah dijelaskan secara spesifik mengenai sumber dan cara

pengambilan hasil primer dan sekunder serta waktu pengambilan
datanya, diantaranya :
1. Pengambilan data dilihat berdasarkan pengamatan yang dilakukan
oleh petugas. Petugas mengamati dan melakukan proses
pengaturan terhadap urutan meja penyajian yang difokuskan pada
makanan penutup dan jenis makanan penutup yang akan disajikan
pada penduduk gereja yang akan menikmati
2
2. Petugas membuat sebuah perauturan berupa pada setiap meja, klien/
penduduk gereja diizinkan untuk mengambil sejumlah produk makanan
penutup yang telah disediakan sesuai dengan perkiraan jumlah peserta.
 Klien tidak diijinkan untuk melewati meja makanan penutup melainkan
harus mengantri, sebagian besar klien lebih memilih mengambil buah
lebih banyak
3. Penelitian ini sudah mendapat persetujuan dari IRB 2
dari lembaga sponsor (14-065 EX 1403 dan
1303003756)

Sample size 7a Penentuan besar sampel penelitian disebutkan dalam bab Metode sub
bab design experiment paragraf 1, yaitu berdasarkan kriteria yang
2
datang ke pantry makan pada hari belajar. Semua peserta tidak
diberikan penjelasan mengenai intervensi.

Dalam penelitian ini sudah disebutkan analisis sementara yang

disebutkan dalam bab Metode sub bab Statistik paragraf 1, yaitu
7b
analisis per protocol (PP) untuk 443 peserta yang lengkap data
penelitiannya. Selain itu juga dilakukan analisis intention-to-treat
(ITT) untuk semua peserta yang diacak.
Mengenai alasan penghentian sampel akan dilakukan apabila
2
pengunjung sudah berhenti datang pada hari itu atau produknya sudah
habis
 Journal of Public Health | Vol. 39, No. 2, pp. 366–372 | doi:10.1093/pubmed/fdw043 | Advance Access Publication May 12, 2016

Food pantry selection solutions: a randomized controlled

trial in client-choice food pantries to nudge clients to
targeted foods
Norbert L.W. Wilson 1, David R. Just 2, Jeffery Swigert3, Brian Wansink4
1
Department of Agricultural Economics and Rural Sociology, Auburn University, Auburn,
AL, USA 2Dyson School of Applied Economics and Management, Cornell University,
Ithaca, NY, USA 3Policy Analysis and Management, Cornell University, Ithaca, NY, USA

Downloaded from https://academic.oup.com/jpubhealth/article-abstract/39/2/366/3002970 by guest on 22 January 2019

4
Dyson School of Applied Economics and Management, Cornell University, Ithaca, NY
14850, USA Address correspondence to Brian Wansink, E-mail:
fblsubmissions@cornell.edu

ABSTRACT

Background Food pantries and food banks are interested in cost-effective methods to encourage the selection of targeted foods without
restricting choices. Thus, this study evaluates the effectiveness of nudges toward targeted foods.

Methods In October/November 2014, we manipulated the display of a targeted product in a New York State food pantry. We evaluated the
binary choice of the targeted good when we placed it in the front or the back of the category line ( placement order) and when we presented the
product in its original box or unboxed ( packaging).

Results The average uptake proportion for the back treatment was 0.231, 95% CI ¼ 0.179, 0.29, n ¼ 205, and for the front treatment, the
proportion was 0.337, 95% CI ¼ 0.272, 0.406, n ¼ 238 with an odds ratio of 1.688, 95% CI ¼ 1.088, 2.523. The average uptake for the

unboxed treatment was 0.224, 95% CI ¼ 0.174, 0.280, n ¼ 255, and for the boxed intervention, the proportion was 0.356, 95% CI ¼ 0.288,

0.429, n ¼ 188 with an odds ratio of 1.923, 95% CI ¼ 1.237, 2.991.

Conclusions Nudges increased uptake of the targeted food. The findings also hold when we control for a potential confounder. Low cost and
unobtrusive nudges can be effective tools for food pantry organizers to encourage the selection of targeted foods.

Trial Registration Number: NCT02403882.

Keywords behavioral economics, food banks, food pantries, nudges, nutrition

Background
Despite significant efforts, both public and private, food inse-
curity persists within the USA. Feeding America, which serves
a vast network of food banks and pantries nationwide, is the
largest domestic feeding program in the USA, providing food
to 46.5 million unique individuals or 15.5 million households
annually.1 Of the households that benefit from Feeding
America, 82.3% were at or below 130% of the federal po-
verty line, which is one of the criteria for eligibility for the
Supplemental Nutrition Assistance Program (SNAP).2 Nearly
84% of these households were classified as food insecure,
while in 2014 the US rate of food insecurity was 14.0%.3 From
the literature on food insecurity, many of these households ex-
perience diet-related diseases.4 – 9 Thirty-three percent of
# The Author 2016. Published by Oxford University Press on behalf of Faculty of Public Health. All rights reserved. For permissions, please e-mail:
journals.permissions@oup.com.
households participating in Feeding America programs have
at least one member diagnosed with diabetes, and 58% of the
households have at least one member who is hypertensive.1
Food pantry organizers aim to give clients the choice of
which foods they receive to both honor the dignity of their
clients and to encourage clients to take products that they ac-
tually want and will end up using, hopefully mitigating waste.
As a result, organizers have created client-choice food
Norbert L.W. Wilson, Professor
David R. Just, Professor
Jeffery Swigert, Doctoral Candidate
Brian Wansink, John S. Dyson Professor of Marketing
366

pantries. With greater choice, however, a greater risk exists
that patrons will choose food bundles that are nutritionally
wanting. Given the demonstrated prevalence of poor dietary
habits among clients, organizers of client-choice food pantries
are interested in encouraging clients to make healthier choices
at their pantries. In this study, we offer two examples of be-
havioral economic techniques that nudge clients without
costly or time consuming effort on the part of the food
pantry (or food bank).
In our analysis, we use two interventions in a client-choice
pantry to determine whether we can nudge clients to a tar-
geted product. We hypothesize that placing a targeted product
first in a presentation of goods will increase the uptake of the
product. Similarly, we hypothesize that presenting the product
in its original packaging will encourage the selection of the tar-
geted product relative to the repackaged product. Food pan-
tries often repackage products to ensure equitable distribution
of products. This intervention is designed to determine the
implications of such repackaging on selection.
Method
Study sample and setting
Our interventions took place at a food pantry in New York
State. The pantry is affiliated with a church and is a member
of the Feeding America network. The church hosts a food
pantry every Tuesday from 1:00 to 2:00 pm. The pantry is
run by 15 – 20 volunteers, many of whom are parishioners of
the church. In 2013, the pantry served 7269 clients and aver-
aged ~150 clients served each session. All clients (n ¼ 443)
who shop the pantry are adults.
The pantry offers clients’ products from a series of tables.
The convention of this pantry is for clients to walk in the
order they arrived along tables where volunteers offer pro-
ducts. At each table, the clients are permitted to take a fixed
amount of the product depending on product availability and
estimates of the number of participants. Products cannot be
exchanged across categories; e.g. clients are not permitted to
skip the desserts to get more fruit. For the interventions we
implement, we focused on the dessert table. Clients can select
one dessert from the products available. The typical mix of
desserts includes frosted sheet cakes, trays of brownies, cookies,
pies, assorted pastries, etc. While the mix of products shifted
from week to week, the general composition of products did
not vary. During our preliminary visits, we noticed protein bars
as one of the choices in the dessert section. Though not neces-
sarily a ‘health food’, the choice of a typical protein bar provides
more protein and fiber and less fat than an equivalent portion
of frosted chocolate cake.10 Furthermore, empirical evidence
suggests that the individually wrapped protein bars may lead to
FOOD PANTRY SELECTION SOLUTIONS 367
lower calorie consumption relative to the chocolate cake
because of the set portions size.11
Experiment design
The pantry offered an assortment of bars (Luna Bars, Clif Bars,
Rickland Orchard Greek Yogurt Covered Protein Bars and
Quaker Oats Chewy Granola Bars (Before the pantry day, if
stocks were low, the research team supplemented the pantry
with the targeted good when available quantities were low)),
which we used for the interventions. We implemented two inter-
ventions, order and packaging, on the dessert section of the
Downloaded from https://academic.oup.com/jpubhealth/article-abstract/39/2/366/3002970 by guest on 22 January 2019
pantry. Prior to the interventions, we received IRB approval
from the sponsoring institutions (14-065 EX 1403 and
1303003756). For the order intervention day, we placed the bars
at the end of the line of desserts (Back). The following week we
implemented the treatment in which we placed the bars at the
beginning of the line of desserts (Front) for the pantry session.
We repeated these interventions for the next two sessions in
reverse order. The interventions took place over four successive
Tuesdays, 28 October 2014 through 18 November 2014. We
analyzed the data in December 2014. We randomized treatments
with parallel assignment by day of the interventions. We achieved
random allocation by alternating treatments by days. We deter-
mined the sample size (n ¼ 443) by the number of clients who
attended the food pantry on the study days. We stopped collect-
ing data if we ran out of either the targeted or the other dessert
products. We did not inform clients about the interventions.
The interventions were observable though not obvious.
For the packaging interventions, we offered the protein bars
in the original packaging (Boxed). As an alternative treatment,
we offered the individually wrapped bars repackaged in sealed,
clear plastic bags (Unboxed). When the product was offered in
the box, the pantry provided 12-count boxes of the products.
The pantry did the same with the unboxed products; however,
for the Chewy Granola Bars, the pantry offered 15-count bags.
For the first two pantry interventions (28 October and 4
November), the pantry offered the boxed products. For the last
two interventions (11 and 18 November), the pantry offered
the unboxed product. We placed the targeted product at the
front of the dessert section on the first and third sessions (28
October and 11 November), and we placed the product at the
back of the section for the second and fourth sessions (4 and
18 November). We measured the take rate, the proportion of
clients that selected the targeted item, and the binary choice to
determine whether the intervention increased selection.
Results
In total, 443 clients visited the pantry over the four pantry
sessions. We had 205 clients under the Front treatment and

238 in the Back treatment. The allocation ratios are 0.46
(Front) and 0.54 (Back). For the packaging intervention, 188
clients visited the pantry under the Boxed treatment and 255
visited under the Unboxed treatment. The allocation ratios
are 0.42 (Boxed) and 0.58 (Unboxed) (Fig. 1 for enrollment).
Based on the binomial distribution of the outcome of selec-
tion of the protein bars, the average uptake of the bars under
the Front treatment is 0.337 with a 95% confidence interval,
CI ¼ 0.272, 0.406 (Table 1). For the Back treatment, the
average uptake rate is 0.231 with a 95% CI ¼ 0.179, 0.290.
The pairwise difference in proportions is 20.105 with a 95%
CI ¼ 20.189, 20.0215. The odds ratio for the intervention
is 1.688 with a 95% CI ¼ 1.088, 2.523. The average uptake
under the Boxed treatment is 0.356 with a 95% CI ¼ 0.288,
0.429, and under the Unboxed treatment, it is 0.224 with
a 95% CI ¼ 0.174, 0.280. The pairwise difference in propor-
tions is 20.133 with a 95% CI ¼ 20.218, 20.047. The
odds ratio of the intervention is 1.923 with a 95% CI ¼
1.237, 2.991. Figure 2 has the means and confidence interval
of the proportion uptake of the targeted food for the order
and the packaging interventions.
Because the two interventions overlap, we assess the effects
of four different combinations of the treatments: Front-Boxed
Assessed for eligibility (n = 443)
Enrollment
Randomized (n = 443)
Allocation
Allocated to Front Boxed intervention (n = 59)
Received allocated order intervention (n = 59)
Did not receive allocated intervention (give reasons)
(n = 0)
Allocated to Back Boxed intervention (n = 146)
Received allocated order intervention (n = 146)
Did not receive allocated intervention (give reasons)
(n = 0)
Analysis
Analysed (n = 59) Analysed (n = 146)
Excluded from Excluded from
analysis (give reasons) analysis (give reasons)
(n = 0)
(n = 0)
Fig. 1 Enrollment and randomization procedure.
FOOD PANTRY SELECTION SOLUTIONS 368
(n ¼ 59), Front-Unboxed (n ¼ 129), Back-Boxed (n ¼ 146)
and Back-Unboxed (n ¼ 109) (On each pantry day, the pantry
had on average 35 units of the targeted product in a mix of
~200 total desserts offered). The mean uptake of the protein
bars for Front-Boxed is 0.525 with a 95% CI ¼ 0.391, 0.657.
For Front-Unboxed, the mean uptake is 0.260 with a 95%
CI ¼ 0.191, 0.339. The mean uptake for Back-Boxed is 0.279
with a 95% CI ¼ 0.204, 0.365. For Back-Unboxed, the mean
uptake 0.174 with a 95% CI ¼ 0.108, 0.259.
We assess the four possible pairwise comparisons and then
consider a comparison of them to the referent intervention
Downloaded from https://academic.oup.com/jpubhealth/article-abstract/39/2/366/3002970 by guest on 22 January 2019
of Back-Unboxed. For Front-Boxed and Front-Unboxed, the
difference in proportions is 20.265 with a 95% CI ¼ 20.411,
20.119. Compared with the referent of Back-Unboxed, the
odds ratio for Front-Boxed is 5.224 with a 95% CI ¼ 2.436,
11.291, and for Front-Unboxed compared with Back-
Unboxed, the odds ratio is 1.667 with a 95% CI ¼ 0.895,
3.105. For Back-Boxed and Back-Unboxed, the difference in
proportions is 20.105 with a 95% CI ¼ 20.210, 0.0004.
The odds ratio for Back-Boxed is 1.834 with a 95% CI ¼
0.974, 3.453. We find that the difference in proportions for
Front-Boxed and Back-Boxed is 20.246 with a 95%
CI ¼ 20.395, 20.097 while the difference in proportions
Excluded (n = 0)
Not meeting inclusion criteria (n = 0)
Declined to participate (n = 0)
Other reasons (n = 0)
Allocated to Front Unboxed intervention (n = 129)
Received order intervention (n = 129)
Did not receive allocated intervention (give reasons)
(n = 0)
Allocated to Back Unboxed intervention (n = 109)
Received order intervention (n = 109)
Did not receive allocated intervention (give reasons)
(n = 0)
Analysed (n = 109) Analysed (n = 129)
Excluded from Excluded from
analysis (give reasons) analysis (give reasons)
(n = 0) (n = 0)
 FOOD PANTRY SELECTION SOLUTIONS 369

Table 1 How placement order and packaging influence proportion of uptake of the targeted good—means and odds ratios

Intervention Mean (95% CI) Pairwise difference of proportions (95% CI) Odds ratio (95% CI)

Placement order
Placed at the back of the line (Back) 0.231 (0.179, 0.290) 20.105** (20.189, 20.0215) 1.688* (1.0879, 2.523)
Placed in the front of the line (Front) 0.337 (0.272, 0.406)
Packaging
Kept in original package (Boxed) 0.356 (0.288, 0.429) 20.133** (20.218, 20.047) 1.923** (1.237, 2.991)
Removed from original package (Unboxed) 0.224 (0.174, 0.280)
Combined

Downloaded from https://academic.oup.com/jpubhealth/article-abstract/39/2/366/3002970 by guest on 22 January 2019

Front-Boxed 0.525 (0.391, 0.657) 20.265*** (20.411, 20.119) 5.224*** (2.436, 11.291)
Front-Unboxed 0.260 (0.191, 0.339) 1.667 (0.895, 3.105)
Back-Unboxed 0.174 (0.108, 0.259) 20.105 (20.210, 0.0004) 1.000
Back-Boxed 0.279 (0.204, 0.365) 1.834 (0.974, 3.453)
Front-Boxed Back-Boxed 20.246** (20.395, 20.097)
Front-Unboxed Back-Unboxed 20.0860 (20.187, 0.015)

Back-Unboxed is the referent for the estimates of both odds ratios of the combined interventions.

*P . 0.5.

**P . 0.01.

***P . 0.001.

(a) (b)
1
Proportion uptake of the targeted good

1
Proportion uptake of the targeted good

.9
.9
.8
.8
.7
.7
.6
.6
.5
.5
.4
.4
.3
.3
.2
Back Front .2 Unboxed Boxed
.1
Placement order intervention .1 Packaging intervention
0
0

Back Front Unboxed Boxed

Fig. 2 How placement order and packaging of pantry foods influence proportion of uptake of the targeted good, (a) placem ent order; ( b) packaging.

for Front-Unboxed and Back-Unboxed is 20.0860 with a Because the pantry does not randomize the line, clients who
95% CI ¼ 20.187, 0.015. are at the beginning of the line have first pick of a limited
These findings suggest that placing the protein bars at the
beginning of the line of desserts increased the uptake relative
to placing the bars at the end of the line. The packaging inter-
ventions are also statistically different. However, the overlap-
ping interventions have statistically significant odds ratios only
for Front-Boxed to Back-Boxed.

Odds ratios from the logistics regression

 FOOD PANTRY SELECTION SOLUTIONS 370
supply, so they may face a different product mix than later
clients. Additionally, clients may self-select to be first. These
early clients will have a lower Queue Number than other
clients. Both of these factors may have influence on the
selec- tion of product. The declining diversity of product
may en- courage selection of the targeted product regardless
of intervention. If clients are self-selecting to be first to pick
the best products, nudges may be less effective on their
selection. Therefore, we control for the effect of Queue
Number in the estimation of the logistics odds ratio. For the
Queue Number, we used a categorical variable that indicates
the quartile of the queue that the client visited the pantry. A
client in the first

25% of the clients is the first quartile. For the second quartile, a
client visited the pantry between the 25th and 50th percentile
of the clients that day. Clients who visited the pantry between
the 50th and 75th percentiles of the clients that day are in the
third quartile, while the fourth quartile has clients who were in
the last 25% of clients who visited the pantry that day.
We estimated the odds ratios for the binary choice of the tar-
geted product with a logistic model with controls (Table 2). We
also ran the linear probability model; both models produced
similar results. The outcome variable is selection of the targeted
product (1) or not (0). We hypothesized that Queue Number
may influence the selection of the targeted product outside of
the intervention. Therefore, we controlled for the direct effect
of Queue Number with the Quartile Queue Number. We also
estimated the models with the ordinal Queue Number, which
ranges from 1 to 147. The estimates are similar for both mea-
sures of the Queue Number, but statistical significance of the
overlapping treatments changed. Additionally, we clustered the
standard errors by the Queue Number. We also ran the models
without clustering the standard errors by the Queue Number,
and the results were nearly identical.
In Table 2, we present four estimates of the logistics regres-
sion. Column 1 (2) reports the odds ratios for the model of the
Order (Packaging) intervention with the control of Queue
Number. In Column 3, we present the odds ratios of control for
both the Order and Packaging interventions along with the
Table 2 Odds ratio of interventions with 95% CI in parentheses below
1 2
Interventions (Referent ¼ Back or Unboxed)
Order placement (Base: Front) 1.769** (1.132, 2.763)
1.828** (1.188, 2.812)
Packaging (Base: Boxed)
Overlapping interventions (Referent ¼ Back-Unboxed)
Front-Boxed
Front-Unboxed
Back-Boxed
2nd Quartile Queue Number 3.663* (1.169, 11.477) 3.660* (1.152, 11.625)
Quartile Queue Number (Referent ¼ 1st quartile)
3rd Quartile Queue Number 4.676** (1.531, 14.279) 4.558** (1.520, 14.334)
4th Quartile Queue Number 9.202*** (3.320, 25.508) 8.692*** (3.133, 24.116)
Constant 0.049*** (0.017, 0.142) 0.051*** (0.019, 0.136)
2
LR x2 26.97*** 33.82***
Pseudo R2 0.076 0.077
n 443 443
We cluster standard errors by Queue Number.
*P . 0.5.
**P . 0.01.
***P . 0.001.
FOOD PANTRY SELECTION SOLUTIONS 371
Quartile Queue Number. Column 4 reports the odds ratios for
the overlapping interventions with Back-Unboxed as the referent
and the Quartile Queue Number. For the Order model (Column
1), the odds ratio for Front is 1.769, 95% CI ¼ 1.132, 2.763. In
the Packaging model (Column 2), the odds ratio for Boxed is
1.828, 95% CI ¼ 1.188, 2.812. In both models, the odds ratios,
for the second, third and fourth Quartile Queue Number relative
to the first quartile, are statistically significant and increasing.
Controlling for both Order and Packaging (Column 3), we
found larger odds ratios for both intervention types while pre-
serving the order of the magnitude. The odds ratio for the
Downloaded from https://academic.oup.com/jpubhealth/article-abstract/39/2/366/3002970 by guest on 22 January 2019
Order intervention (Front to Back) is 2.140, 95% CI ¼ 1.333,
3.438, while the odds ratio for Packaging intervention (Boxed
to Unboxed) is 2.204, 95% CI ¼ 1.396, 3.481. Again, we see
that the Quartile Queue Number has a statistically significant
odds ratios. Across the three columns, the odds ratios for the
Quartile Queue Number do not change, though the point
estimates differ by small amounts.
In the final model (Column 4), we consider the overlapping
interventions while controlling for Queue Number. Relative to
the Back-Unboxed treatment, the odds ratio for the Front-
Boxed is 4.739 with a 95% CI ¼ 2.269, 9.898. For Front-
Unboxed, the odds ratio is 1.988 witha 95% CI ¼ 1.060, 3.729.
The odds ratio for Back-Boxed is 2.045 with a 95% CI ¼
1.080, 3.872. For each of the quartiles of the Queue Number,
the odds ratios are statistically significant and increasing.
3 4
2.140** (1.333, 3.438)
2.204*** (1.396, 3.481)
4.739*** (2.269, 9.898)
1.988* (1.060, 3.729)
2.045* (1.080, 3.872)
3.706* (1.158, 11.863) 3.709* (1.157, 11.897)
4.691** (1.486, 14.813) 4.690** (1.478, 14.875)
8.664*** (3.056, 24.567) 8.550*** (3.029, 24.136)
0.0325*** (0.011, 0.095) 0.034*** (0.012, 0.101)
42.51*** 43.02**
0.098 0.098
443 443

Discussion
Main finding of this study
The results of this intervention contribute to the literature at
the intersection of behavioral economics, nutrition and po-
verty in three important ways: (i) We provide evidence of
the effectiveness of behavioral interventions in food choice
among adults in a food pantry. Products in the first position
may make selection cognitively easier.12,13 Further, keeping
products in the original box may remove the stigma of receiv-
ing the product from the food pantry. (ii) We furnish findings
that suggest that the overlaying of interventions may strength-
en their effectiveness—e.g. Front-Boxed has the highest odds
ratio. (iii) We find evidence that suggests the effectiveness of
interventions may change over the implementation period. As
a result, low cost and simple interventions may improve the
selection of targeted products at food pantries. However, the
context, such as product availability and point of selection,
may matter in implementation of interventions.
What is already known on this topic?
An extensive literature suggests that behavioral interventions
may alter food choice. This literature presents challenges to
pricing and educational interventions that often have little
impact on day-to-day choices either in food or in finance.14
Despite best efforts, individuals make mistakes, especially
when long-term outcomes are largely determined by short-
run decisions.14 – 19 Behavioral techniques have been effective
in shaping food decisions of children15,17,20 – 23 and adults24,25
and often at little cost. Most of these interventions were in
schools, however, leading to uncertainty in whether these
effects generalize to non-student populations. Also, while
some of the schools in these studies had a relatively high per-
centage of low-income children, none of the studies consider
populations with the high concentration of poverty that the
current study considers. Only one study explores the possibil-
ity of nudges in food pantries.13
What this study adds?
We provide evidence that behavioral interventions are effect-
ive in the promotion of products among this high-poverty
population, and the impact that interventions have on clients’
choices persists as the choice set each subsequent client
encounters changes.
Limitations of this study
A potential limitation of this study is the noisiness inherent in
the setting. Food pantries rely, in large part, on in-kind dona-
tions and have limited resources, so they offer different pro-
ducts from session to session. The number and diversity of
FOOD PANTRY SELECTION SOLUTIONS 372
clients also vary over sessions. In this study, the dessert
section fluctuated in product offerings over the study period,
though there were no major changes in the types of products
offered. The number of clients recorded during the study
ranged from 59 to 146 clients. As these data are observational,
and the clients move through the line at a steady pace, we did
not collect detailed information on the characteristics of the
clients.
As the results in Table 2 suggest, Queue Number or the
Quartile Queue Number has a statistically significant effect on
the odds of uptake of the targeted product. This result is con-
Downloaded from https://academic.oup.com/jpubhealth/article-abstract/39/2/366/3002970 by guest on 22 January 2019
sistent with the observation that the diversity of the products
available decreases throughout the day, because the odds ratio
increases for each subsequent quarter. Subjects are more likely
to select the targeted product later in the day as the relative
presence of the targeted products increases. Thus, the increas-
ing odds ratio follows from the point of view of product diver-
sity. Despite this effect, the odds ratios still indicate a
statistically significant effect. Pantries can resolve this problem
by ensuring a more stable balance of products for all clients.
Conclusions
When given a choice of desserts, protein bars may not be the
first choice among frosted cakes, pies, doughnuts and cookies.
However, this study provides evidence that even within this
choice set, clients at food pantries may be successfully nudged
toward selecting the lesser of two (or three, four, five .. .)
evils. We find that placing protein bars at the front of the
display relative to the end of the display leads to an odds ratio
of 2.140 (95% CI ¼ 1.333, 3.438), controlling for queue
number. Similarly, keeping the product in the original pack-
age has an odds ratio of 2.204 (95% CI ¼ 1.396, 3.481).
Overlapping interventions by placing the targeted product up
front in the original package (Front-Boxed) has an odds ratio
of 4.739 (95% CI ¼ 2.269, 9.898) over placing the product at
the end of the line out of the original package, controlling for
queue number. The simplicity of these interventions suggests
that implementation of nudges is feasible for any pantry or-
ganizer.26 This study is the first to provide evidence for the ef-
fectiveness of simple behavioral interventions in the food
pantry setting to encourage the selection of targeted foods.
Acknowledgements
The research team thanks the Food Bank of the Southern
Tier and the volunteers and clients of the food pantry at
Immaculate Conception Catholic Church. The research team
appreciates the editorial assistance of Rosemarie Hanson.
Trial Registration name: Behavioral Economic Interventions
FOOD PANTRY SELECTION SOLUTIONS 373
 FOOD PANTRY SELECTION SOLUTIONS 374

at Food Pantries. Trial Registration Number: NCT02403882. Date of Registration: 26 March 2015.

Conflict of interest statement

N.L.W.W. conducted the research for this paper while on sabbatical leave at Cornell University.
D.R.J received grant funding from Feeding America to explore nudges in food pantries.
However, those funds were not used in this project.

Funding
This work was supported by the Agricultural and Food Research Initiative Competitive Program of the
US Department of Agriculture (USDA) National Institute of Food and Agriculture (NIFA) (grant
number 2014-67024-21849).

References
1 Weinfield NS, Mills G, Borger C et al. Hunger in America Report 2014. Chicago, IL: Feeding America, 2014.
2 Food and Nutrition Service. Supplemental Nutrition Assistance Program (SNAP), Eligiblility. 2014.
http://www.fns.usda.gov/snap/ eligibility (22 February 2015, date last accessed).
3 Coleman-Jensen A, Rabbitt MP, Gregory C et al. Household Food Security in the United States in 2014.
Washington, DC: US Department of Agriculture ERS, 2015.
4 Townsend MS, Peerson J, Love B et al. Food insecurity is positively related to overweight in women. J Nutr
2001;131(6):1738 – 45.
5 Adams EJ, Grummer-Strawn L, Chavez G. Food insecurity is asso- ciated with increased risk of obesity in
California women. J Nutr 2003;133(4):1070 – 4.
6 Lynn M, Valerie T, Tony Jinguang L. Improving the nutritional status of food-insecure women: first, let them
eat what they like. Public Health Nutr 2007;10(11):1288 – 98.
7 Olson CM, Strawderman MS. The relationship between food insecur- ity and obesity in rural childbearing
women. J Rural Health 2008;24(1):60 – 6.
8 Handforth B, Hennink M, Schwartz MB. A qualitative study of nutrition-based initiatives at selected food
banks in the Feeding America network. J Acad Nutr Diet 2013;113(3):411 – 5.
9 Ding M, Wilson NLW, Garza KB et al. Undiagnosed prediabetes among food insecure adults. Am J Health
Behav 2014;38(2):225 – 33.
 FOOD PANTRY SELECTION SOLUTIONS 375

10 US Department of Agriculture, Agricultural Research Service, Nutrient Data Laboratory. USDA

National Nutrient Database for Standard Reference, Release 28. Version Current: September
2015. http://www.ars.usda.gov/nea/bhnrc/ndl.
11 Geier A, Wansink B, Rozin P. Red potato chips: segmentation cues can substantially decrease
food intake. Health Psychol 2012;31(3):398 – 401.
12 Wansink B. Change Their Choice! changing behavior using the CAN approach and activism
research. Psychol Mark 2015;32(5): 486 – 500.
13 Wilson NLW. When the cupboards are bare: nudging food pantry clients to healthier foods. J

Downloaded from https://academic.oup.com/jpubhealth/article-abstract/39/2/366/3002970 by guest on 22 January 2019

Assoc Consum Res 2016;1(1):125 – 133.
14 Just DR, Payne CR. Obesity: can behavioral economics help? Ann Behav Med 2009;38:47 – 55.
15 Hanks AS, Just DR, Wansink B. Preordering school lunch encourages better food choices by
children. JAMA Pediatr 2013;167(7):673 – 4.
16 Just DR, Mancino L, Wansink B. Could Behavioral Economics Help Improve Diet Quality for
Nutrition Assistance Program Participants? Washington, DC: US Department of Agriculture,
Economic Research Service, 2007.
17 Wansink B, Just DR, Hanks AS et al. Pre-sliced fruit in school cafeterias: children’s selection and
intake. Am J Prev Med 2013;44(5): 477 – 80.
18 Thaler RH, Sunstein CR. Libertarian paternalism. Am Econ Rev
2003;93(2):175 – 9.

19 Thaler RH, Sunstein CR. Nudge: Improving Decisions About Health, Wealth, and Happiness.
New Haven, CT: Yale University Press, 2008.
20 Wansink B, Just DR, Payne CR et al. Attractive names sustain increased vegetable intake in
schools. Prev Med 2012;55(4):330 – 2.
21 Olstad DL, Goonewardene LA, McCargar LJ et al. Choosing healthier foods in recreational sports
settings: a mixed methods investigation of the impact of nudging and an economic incentive.
Intl J Behav Nutr Phys Act 2014;11(1):1 – 30.
22 Hanks AS, Just DR, Smith LE et al. Healthy convenience: nudging students toward healthier
choices in the lunchroom. J Pub Health 2012;34(3):370 – 6.
23 Wansink B, Just DR, Payne CR. Can branding improve school lunches? Arch Pediatr Adolesc Med
2012;166(10):967 – 8.
24 Thorndike AN, Sonnenberg L, Riis J et al. A 2-phase labeling and choice architecture
intervention to improve healthy food and bever- age choices. Am J Public Health
2012;102(3):527 – 33.
25 Payne CR, Niculescu M, Just DR et al. Shopper marketing nutrition interventions. Physiol
Behav 2014;136(0):111 – 20.
26 Wansink B. Healthy profits: a retail framework for increasing the sales of healthy foods. J
Retailing 2016;93:1(September), forthcoming.