Food Insecurity in Community College
Students in the Central Valley of California
Chapter 1: Introduction to the Study
The cost of a college education has increased dramatically over the years. As a
result, college students sometimes struggle with having enough money to support
themselves while going to college, and students sometimes must sacrifice basic living
needs such as food. Food insecurity occurs when there is unwanted reduced food intake
or disrupted eating patterns (U.S. Department of Agriculture, Economic Research Service
[USDA ERS], 2020). In a study of 34 colleges, 48% of students were food insecure, and a
second study found 36% of students in 66 colleges were food insecure, both of which are
higher than the national average of 12% of households being food insecure (Meza et al.,
2019). A systematic review of food insecurity studies and college students found 32.9%
of U.S. students were food insecure (Bruening et al., 2017). The consequences of food
insecurity are far-reaching and can influence all aspects of a person’s life. When students
do not have enough food to eat, their daily lives are disrupted, academic success is lower,
and their psychosocial health is impacted (Meza et al., 2019). Therefore, educational
leaders must begin considering food insecurity as a component that may contribute to
academic success. Studies have been conducted in recent years on food insecurity in
college students in general. However, there needs to be more in-depth research on college
students in different regions of the country, on different types of college students, such as
community college students, and on what factors and demographic characteristics are
associated with food insecurity.
The goal of this study was to develop a list of characteristics food insecure
students have in common. This list of common characteristics will serve as a student
profile that may reveal which community college students may be at risk for being food
insecure. Once a profile of a food-insecure student is identified, colleges can offer
referrals and support services for students to help prevent them from experiencing food
insecurity and the associated consequences. Helping support students’ basic need for food
can help promote their academic success—a goal of higher education. Higher education
leaders will benefit from knowing the results of this study so they can implement change
in their colleges. Chapter 1 provides an overview, purpose, and background of this study
on food insecurity and community college students. The framework, nature, and scope of
the study, assumptions, limitations, and significance are also described in this chapter.
Background
In the United States, food insecurity is a growing problem; one out of seven
households is food insecure (Martinez et al., 2016). With so many households
experiencing food insecurity and many colleges reporting a higher rate of food insecurity
among college students, the University of California created the UC Global Food
Initiative in 2014 to determine the scope of food insecurity within the UC campuses
(Martinez et al., 2016). In 2015, the University of California’s 10 campuses conducted a
Student Food Access and Security Study and found that 48% of undergraduates and 25%
of graduate students were food insecure, and they found a higher number of older
students, first-generation students, and ethnic minorities were food insecure (UC Global
Food Initiative: Food and Housing Security at the University of California, 2017).
Students who are food insecure are less successful academically because they are
less able to focus on school (Martinez et al., 2016). As a result, colleges have a vested
interest in identifying and helping prevent food insecurity of students, thus helping to
improve students’ academic success (Frank, 2018; Henry, 2017). Colleges, students, and
society will benefit from having college students who are food secure because students
will be more likely to complete their academic programs. Studies show students are less
likely to complete their programs of study when they are food insecure (Vasquez et al.,
2019). Students will also benefit by avoiding the harmful physical and emotional stress of
being food insecure. Society will benefit from having a more educated workforce and
community. Since food insecurity is related to poor health outcomes, an additional benefit
to society will be a healthier college student population (Martinez et al., 2019).
There is a gap in the literature regarding food insecurity in community college
students and the support needed to help prevent food insecurity in at-risk students
(Vasquez et al., 2019), especially for California community college students (Ilieva et al.,
2019), which this study will address. More research is needed on community colleges
students and food insecurity since community colleges tend to enroll more disadvantaged
at-risk students, including students of color (Vasquez et al., 2019). Community colleges
are affordable public higher education institutions that offer certificate and 2-year degree
associate degree programs. The California community college system is the largest higher
education system in the country, with an enrollment of over 2.1 million students at 115
colleges (California Community Colleges Chancellor’s Office, 2020). Given the breadth
of students served by California community colleges, the profile of a food-insecure
California community college student must be identified, especially considering under
40% of community college students complete their program within 6 years (Mann, 2018).
Problem Statement
There is a lack of knowledge of the profile of a food-insecure community college
student and the relationships between ethnicity, gender, being a first-generation college
student, being a non-traditional student, being a single parent, and food insecurity in
students at Hiker’s Community College (HCC; pseudonym) in the Central Valley of
California. Food insecurity is common among students in California, particularly among
community college students and students with low incomes (Institute for College Access
& Success, 2018). In the Central Valley of California, poverty rates are among the highest
in the nation. According to the University of California Student Food Access and Security
Study, approximately 42% of students in the University of California System are food
insecure (Martinez et al., 2016). Similarly, Ortiz et al. (2019) found over half of students
attending a Central Valley university were food insecure, with minorities affected at a
greater rate. Rates of food insecurity are higher in the Central Valley, presumably because
rates of poverty are higher. Colleges are taking steps to help address food insecurity in
students; however, identifying the students most at risk remains a challenge; therefore,
many students remain hungry (Institute for College Access & Success, 2018). Having
basic needs met, such as enough food to eat, contributes to collegiate success because
students can focus on school and complete school in a timely manner (Institute for
College Access & Success, 2017). But lack of money for all costs associated with
attending college, including lack of money for food, contributes to fewer students
enrolling in college and graduating from college (Institute for College Access & Success,
2018). Understanding the profile of a food-insecure student will enable leaders at HCC to
identify students in need of support, which will address a gap in practice.
Purpose of the Study
The purpose of this quantitative correlational study was to determine the profile of
a food-insecure community college student and any relationships between food insecurity
and ethnicity, gender, being a first-generation college student, being a non-traditional
student, and being a single parent at HCC in the Central Valley of California. Food
security was the dependent variable, and ethnicity, gender, being a first-generation
college student, being a non-traditional student, and being a single parent were all
independent variables. These variables were chosen because they were each significant
predictors for food insecurity in other studies, as described in the literature review. For
example, Spaid and Gillet-Karam (2018) found that ethnic minorities, women,
nontraditional students, and single parents had high rates of food insecurity in college
students in Maryland. My study determined if these variables were predictors in
community college students in the Central Valley of California, a very different location.
If HCC has a profile of food-insecure students, college leaders can target supportive
services to food-insecure students. Surveys were emailed to the total non-dual enrollment
student population at HCC to investigate the problem in this study. By using an ordinal
logistic regression test, relationships between study variables were analyzed to identify
the profile of food-insecure students.
Research Question and Hypothesis
Research Question: Is food insecurity related to ethnicity, gender, being a
firstgeneration college student, being a non-traditional student, or being a single parent in
HCC students?
H0: There is no relationship between food insecurity and ethnicity, gender, being a
first-generation college student, being a non-traditional student, or being a single parent
in HCC students.
Ha: There is a relationship between food insecurity and ethnicity, gender, being a
first-generation college student, being a non-traditional student, or being a single parent
in HCC students.
Theoretical Foundation
Abraham Maslow created a humanistic theory of human motivation in the 1970s
to describe the positive potential of people (Beers, 2020). According to Maslow’s theory,
behavior is influenced by a hierarchy of five categories of needs: physiological, safety,
belongingness, esteem, and self-actualization (Farmer, 1984; Kretchmar, 2019). These
lower-order physiological needs, such as food, must be met before higher-order needs can
be achieved (Kretchmar, 2019). Based on this theory, eliminating hunger can improve
students’ physical and emotional health and potentially their academic performance
(Smith, 2018).
The humanistic theory of motivation was used as the framework of this study
because meeting students’ physiological needs for food and water must be met before
students can move on to achieve higher-level needs on Maslow’s hierarchy of needs.
Achievement, including educational attainment, is considered in Maslow’s esteem
category of needs and can only be met after physiological, safety, and belongingness
needs are met (Kretchmar, 2019; Smith, 2018). There has been other research on food
insecurity and college students, which showed that college students who were food
insecure had lower grade point averages (GPAs) than students who were food secure
(Raskind et al., 2019). Studies have shown the need to identify food-insecure students and
for colleges to provide supportive services to help promote academic success.
Nature of the Study
To identify if relationships exist between multiple variables, researchers should
use a quantitative correlational design (Curtis et al., 2016). The nature of this study was
quantitative and consisted of an anonymous survey that was sent to all non-dual
enrollment students at HCC in the Central Valley of California. An ordinal logistic
regression model was used to analyze the data to determine if relationships existed
between food insecurity and ethnicity, gender, being a first-generation college student,
being a non-traditional student, and being a single parent. Results of this study can inform
community college leaders. The dependent variable was food insecurity. The independent
variables were ethnicity, gender, being a first-generation college student, being a
nontraditional student, and being a single parent. SPSS Version 27 was used to analyze
the data.
Definitions
Community college: Colleges that offer 2-year associate degree programs as well
as vocational programs (EducationUSA, 2015).
Ethnicity: Defined as either Asian, Black or African American, Hispanic/Latino,
Native American or other Pacific Islander, or White (Office of Civil Rights, 2003).
First-generation College Student: A student whose biological parents have not
completed a 4-year degree (Center for First-Generation Student Success, 2017).
Food insecurity: Reports of reduced quality, variety, or desirability of diet (USDA
ERS - Definitions of Food Security, 2020).
Food security: High food security refers to no reported indications of food-access
problems or limitations (USDA ERS - Definitions of Food Security, 2020). Marginal
food security refers to one or two reported indications—typically of anxiety over food
sufficiency or shortage of food in the house. Little or no indication of changes in diets or
food intake (USDA ERS - Definitions of Food Security, 2020). Low food security: reports
of reduced quality, variety, or desirability of diet. Little or no indication of reduced food
intake (USDA ERS - Definitions of Food Security, 2020). Very low food security refers
to multiple indications of disrupted eating patterns and reduced food intake
(USDA ERS - Definitions of Food Security, 2020).
Gender: Defined as either male or female (National Center for Education
Statistics, 2021).
Non-traditional student: A student who begins college more than a year after high
school graduation, has family and financial responsibilities, such as being a single parent,
works full time, has earned a GED, or attends college part time (National Center for
Education Statistics, n.d.). In this study, a non-traditional student was defined as someone
25 and older.
Single parent: A person who is raising at least one child without a partner (U.S.
Census Bureau, 2021).
Assumptions
Assumptions are critical conditions in the research design that are taken for
granted and out of the researcher’s control (Burkholder et al., 2016). In this study, the
first assumption was that students would understand the survey questions and answer the
survey questions accurately and honestly. Second, I assumed that respondents were
representative of the entire student population at HCC. The third and final assumption
was that Maslow’s human theory of motivation is fundamentally true and applies to food
insecurity in college students. Further, ordinal logistic regression tests have four
assumptions that must be met in order to obtain valid results: the dependent variable must
be an ordinal variable, there must be at least one categorical or continuous variable, no
multicollinearity can exist between two independent variables, and there should be
proportional odds that each independent variable has the same impact (Udo &
Akwukwuma, 2019).
Additionally, as a registered dietitian, I am biased to believe food insecurity is
related to many negative physical, mental, and social outcomes. In relation to this study, I
had a preconceived belief that certain ethnic groups, being a woman, being a
firstgeneration college student, and being a single parent are all factors that can relate to
food insecurity. To reduce bias, I analyzed all data objectively and did not include
subjective information.
Scope and Delimitations
The scope of a study describes the population being studied; therefore, the scope
frames the population the results are applicable to and helps determine the title of the
study (Burkholder et al., 2016). The setting for this study was a large community college
in a large Central Valley city in California. The study site was in a poor agricultural and
oil-based city that has one of the highest rates of poverty in the United States. The scope
of this study was to identify the profile of a food-insecure community college student in
the Central Valley of California. This college was chosen based on its large student
population, high poverty rate in the country (20.5% versus 11.8% nationally), and low
college education completion rate of only 16% of the population having a 4-year degree
or higher (U.S. Census Bureau, 2020). Having a population that is educated can improve
unemployment and poverty rates in the county. There were approximately 40,000
students at this college, 10,000 of which were high school dual enrollment students who
were excluded from the study.
Delimitations define the boundaries of a study in terms of what will not be
included in the study (Burkholder et al., 2016). In addition, the definition of the study
population and the delimitations of the study must be in alignment. The delimitations of
this study included health outcomes, grades, and graduation rates of students who were
food insecure. Those items were excluded because the study would have been too large
with those variables included. The study also excluded high school dual enrollment
students because they are generally minors living with their guardians, which could skew
the results. Understanding the psychological impact of food insecurity on college students
would be useful; however, adding a qualitative component to the study would have
expanded the scope and time needed to complete the research beyond what this study
intended.
Limitations
Limitations are uncontrollable weaknesses in research design that can influence
study findings, help to present findings in the proper context, and can influence the
research quality (Ross & Bibler Zaidi, 2019). The limitations of this study included using
a standardized survey instrument where further clarification on questions was not able to
be provided to respondents. Survey responses were anonymous to ensure confidentiality
and encourage students to respond honestly. Additionally, the survey was only given to
students at one community college in California, so the responses and results are not
generalizable to all California community college students, especially for college students
outside of the California Central Valley. Another limitation in this study was that only
community college students were included and not 4-year university students.
Determining correlations between research variables does not mean one variable caused
another; it merely means a relationship exists. Therefore, it is important not to assume
causation from correlational research. A significant limitation to food insecurity research
is the lack of knowledge about why students are food insecure. Lastly, there is a
difference between food insecurity causing lack of nutrition and variety of food choices
versus food insecurity that causes a lack of having enough food to eat to meet basic
physiological needs. This study followed the USDA definitions and does not offer
explanations as to why food insecurity existed.
Significance
Multiple researchers have identified a gap in the literature regarding the
prevalence of food insecurity in community college students (Ilieva et al., 2019; Vasquez
et al., 2019). This study addressed a local problem by focusing on increasing knowledge
in the field about the existence of food insecurity in community college students.
Educators can benefit from this research because it provides an increased understanding
of the characteristics of California community college students in the Central Valley of
California who are food secure. The findings contribute to positive social change by
identifying the high prevalence of food insecurity in community college students in the
Central Valley and by providing a profile of high-risk college students most likely to be
food insecure. Education professionals can take this information on food-insecure
students and design interventions to help address food insecurity for college students.
Summary
Chapter 1 introduced the problem of food insecurity in college students in a
community college in the Central Valley of California. The background and history of
food insecurity and college students were provided in Chapter 1. Chapter 1 also outlined
the research question and hypothesis, provided the theoretical foundation and nature of
the study. Additionally, definitions, assumptions, scope and delimitations, limitations,
and significance were addressed in Chapter 1. The purpose of this research was to
determine the profile of a food-insecure community college student in the Central Valley
of California and determine any relationships that existed between food insecurity and
ethnicity, gender, being a first-generation college student, being a non-traditional student,
and being a single parent. The literature review in Chapter 2 provides greater in-depth
analysis of the existing literature related to food insecurity and its impact and identifies
future research needs.
Chapter 2: Literature Review
It is important to identify a profile of community college students most likely to
be food insecure. The purpose of this research was to provide this profile of a
foodinsecure community college student in the Central Valley of California as well as
determine the relationships between food insecurity and ethnicity, gender, being a
firstgeneration college student, being a non-traditional student, and being a single parent.
Chapter 2 includes a comprehensive review of the literature related to food insecurity and
college students.
Literature Search Strategy
I used multiple databases to ensure a comprehensive search of sources when
conducting the literature review on food insecurity and college students. The databases
included EBSCO, Education Source, ERIC, Google Scholar, Medline, ProQuest,
PubMed, SAGE Journals, Science Direct, and the Walden Library. The journals included
the American Journal of Health Education, American Journal of Public Health, Annals of
Anthropological Practice, BMC Public Health, California Agriculture, Canadian Journal
of Higher Education, Education, Journal of the Academy of Nutrition and Dietetics,
Journal of Behavioral and Social Sciences, Journal of College Student Development,
Journal of Family and Consumer Science, Journal of Health Psychology, Journal of
Hunger and Environmental Nutrition, Journal of Nutrition Education and Behavior,
Nutrients, Public Health Nutrition, and the South African Journal of Clinical Nutrition.
To find adequate resources for the literature review, I continued to expand the key
terms I used in my searches over the last 4 years of researching food insecurity. The final
list of key terms used in my search included: academic success, community college,
college student, ethnicity, first-generation college student, food insecurity, gender, higher
education, hunger, lack of food, minority, single parent, and student success. No date
limits were used during my searches because of the need to ensure seminal works were
included. In my final research, only peer-reviewed research under 5 years old was
included, with the exception of seminal works. One technique used to identify additional
sources was to review the bibliographies of each research article for relevant sources that
could be included. After an exhaustive literature search and analysis, a comprehensive
review of the literature on food insecurity and college students was conducted. The
sources were sorted by topic, study variables, type of college student, and location in the
United States. Each source was thoroughly evaluated before being included in the
literature review.
Theoretical Foundation
Maslow was a psychologist who turned his observations of famous people into a
theory widely used in education and social science fields (Mason, 2016). In the 1970s,
Maslow developed the humanistic theory of human motivation to explain positive
potential (Beers, 2020). Maslow believed a hierarchy of five categories of needs
influences human behavior (Kretchmar, 2019). The five categories of needs include
physiological, safety, belongingness, esteem, and self-actualization (Farmer, 1984).
Maslow’s humanistic theory of human motivation is based on the need to have the most
basic human needs met before higher-level needs of belongingness, esteem, and
selfactualization can be achieved (Kretchmar, 2019). The most basic physiological needs
are for food and water, which must be satisfied before people will be motivated or
interested in satisfying any other needs and fulfill their potential (Maslow, 1943;
Gialamas et al., 2017; Sarici Bulut, 2018). For example, if someone is chronically hungry,
they cannot focus on anything but food, and this hunger drive not only changes their
current view of their priorities, but it changes the way they look at the future because all
that matters to them is having enough food to eat (Maslow, 1943). Experiencing hunger
can even influence their sense of belonging, self-esteem, and ability to connect to people
(Mason, 2016). Hunger can also contribute to the stressful and draining feeling of being
overwhelmed and in survival mode (Gialamas et al., 2017). If the need for food (and
shelter) is satisfied, the hungry person can begin having higher-level social goals,
including education, which fulfills the desire to know and understand (Maslow, 1943).
Without meeting basic physiological needs, functioning at an optimal level is challenging,
and progressing up the hierarchy of needs toward self-actualization is difficult (Mason,
2016; Mcleod, 2020).
Based on Maslow’s theory, students who experience food insecurity cannot focus
on school and do not perform as well academically (Smith, 2018). According to
Maslow’s theory, education would be in the esteem category and cannot be met until
physiological needs have been met. Considering a holistic approach to education can help
improve student outcomes by ensuring students are not hungry (Mcleod, 2020). Lack of
food can impact students’ psychological health and well-being, which may negatively
affect their academic achievement. Furthermore, food-insecure students may be
embarrassed to ask for help; therefore, they do not seek available support resources.
Results have shown that almost half of the South African students surveyed preferred to
go hungry instead of disclosing their lack of resources, and they would rather go hungry
than receive food aid (Sabi et al., 2019). For this reason, many schools offer help to food
resources to all students instead of singling out food-insecure students. These same South
African students reported they missed classes or could not concentrate in class due to
hunger, which led to a lack of focus, missed classes, and lack of participation in academic
activities; therefore, students did not live up to their potential academically (Sabi et al.,
2019). This is an illustration of the importance of satisfying basic needs as described by
Maslow. Students cannot be expected to achieve academic success if they lack the basic
necessity of food.
Using Maslow’s hierarchy of needs as the framework for this study on food
insecurity fits with the existing research on food insecurity that identifies basic needs as
paramount before success in higher education can be achieved. Sabi et al. (2019) used
Maslow’s hierarchy of needs and Piaget’s theory of cognitive learning as the construct for
their study. Piaget’s theory of cognitive learning states that teachers do not give
knowledge; instead, learners construct knowledge depending on internal and external
factors, such as physiological well-being and motivation (Berkeley Graduate Division,
n.d.). Since many students have external factors that prevent them from having enough
adequate food to eat, learning becomes secondary. Without the feeling of having
physiological needs met, students will not be able to move to higher levels of
achievement, nor will they have satisfaction until their basic needs are met (Mason,
2016). Raskind et al. (2019) also conducted a study using Maslow’s hierarchy of needs to
determine if psychosocial health mediates a relationship between the GPA of college
students in Georgia and food insecurity. Results showed food insecurity impacted
academic performance and, therefore, student success. Therefore, Maslow’s hierarchy of
needs supported this study in determining a profile of food-insecure community college
students in order to encourage student support for achieving food security, which can
help improve academic success by allowing students to move past the physiological need
for food and consider higher-level goals.
Literature Review Related to Key Concepts and Variables
Food Insecurity Defined
Food security occurs when there is not enough food to eat on a daily basis to live a
healthy life. Food insecurity occurs when a household does not have enough food for
each person to eat because of a lack of money or resources (USDA ERS, 2020). There are
different levels of food insecurity. The USDA ERS (2020) categorizes food security and
insecurity in four categories:
1. High food security: no reported indications of food-access problems or
limitations.
2. Marginal food security: one or two reported indications—typically of anxiety
over food sufficiency or shortage of food in the house. Little or no indication
of changes in diets or food intake.
3. Low food security: reports of reduced quality, variety, or desirability of diet.
Little or no indication of reduced food intake.
4. Very low food security: multiple indicators of disrupted eating patterns and
reduced food intake. This is the most significant type of food insecurity and
can have detrimental effects both physically and mentally. Hunger results
from very low food security. Hunger is physiological condition that results
from food insecurity and results in discomfort, weakness, illness, or pain.
Prevalence of Food Insecurity
In 2019, approximately 10.5% of households were food insecure, and 4.1% of
households were very food insecure (Coleman-Jensen et al., 2020). The very food
insecure households had at least one person who did not have enough food to eat due to a
lack of resources. As another indicator of the prevalence of the need for food support,
approximately 25% of Americans participate in one or more USDA food and nutrition
assistance programs during any given year (USDA ERS, 2019). In California, 21% of
K12 students are food insecure (Smith, 2018). Given the high rate of household and K-12
student food insecurity in the United States, the likelihood of college students being food
insecure is high due to their limited incomes and increased expenses of school and living.
Multiple studies have shown high rates of college students who are food insecure.
In a study of students at eight universities in the United States, 19% of students
experienced food insecurity, and an additional 25.3% of students were at risk of
becoming food insecure with ethnic minorities, living off-campus, having parents who
did not attend college, receiving a Pell grant, and not participating in a meal plan as
commonalities of being a food-insecure student (El Zein et al., 2019). Further studies
identified a food insecurity rate of 20%–50% of college students, which is higher than the
overall food insecurity rate of 14.6% in the United States (Freudenberg et al., 2019). Data
from four surveys of over 30,000 2- and 4-year college students from 121 colleges in 26
states also indicated that 57%–67% of college students experienced some level of
food insecurity, and for community college students, 11%–38% experienced very low
food security with hunger (Broton & Goldrick-Rab, 2018). In 2016, due to the high cost
of college and higher debt, 56% of community college students and 25% of 4-year
college students were food insecure (Que & Baldridge, 2019). Research in 12 states with
eight community colleges and 26 4-year colleges also showed that 48% of student
respondents had experienced food insecurity in the 30 days before the survey; and ethnic
students, as well as first-generation students, were more likely to be food insecure, and
56% of foodinsecure students had a paid job, with 38% of them working at least 20 hours
per week, and 75% of food-insecure students were receiving financial aid (Cady et al.,
2016).
The highest rates of food insecurity have been reported among California
community college students (Goldrick-Rab et al., 2017). Research has found that 58% of
California community college students were food insecure (O’Neill, 2018). Out of the
California State University’s 27,805 students who completed the California State
University Phase 2 survey, 41.6% were food insecure, despite working an average of 13.8
hours per week (California State University, 2019).
More recent research in 2020 suggested that out of the 469 colleges who
responded, the colleges believed up to 80% of students were food insecure, and 86% of
the colleges believed food insecurity was related to the reason students were not
completing their degrees (Goldrick-Rab, 2020). Additionally, part-time students had a
food insecurity rate 11% higher than full-time students, and the longer students were in
college, the higher the rate of food insecurity (Baker-Smith et al., 2020). The Hope
Center conducted its fifth annual Real College Survey in the Fall 2019 semester at 227
colleges across the United States and had 167,000 student responses. The results showed,
in the previous 30 days, 39% of students experienced food insecurity (Baker-Smith et al.,
2020). Results from surveys from 2015–2019, which includes 411 colleges and over
330,000 students, also showed that 33%–48% of 4-year college students and 42%–56%
of 2-year college students were food insecure (Baker-Smith et al., 2020). In 2019, the
Government Accountability Office (GAO) conducted a review of 31studies that were
completed on food insecurity in college students and determined that 2 million students
would have qualified for the SNAP but did not receive benefits in 2016. Federal student
aid is not enough to pay for all college expenses, creating more low-income students, and
low-income students are more likely to have an additional risk factor for being food
insecure, such as being a single parent or a first-generation student (U.S. GAO, 2019).
Consequences of Food Insecurity
Experiencing food insecurity leads to a multitude of problems, including academic
failure, because it is hard to focus on education when people are physically hungry (Meza
et al., 2019). First-year students have done worse academically if they were food
insecure, regardless of their background, and they had more stress and anxiety, less
money for everything, less time because of possibly working, and inadequate access to
required school materials as well as computer and internet access to complete the
assigned work (Woerden et al., 2019). Based on a study of California students, the
average GPA for food-insecure students was 3.17 compared to 3.32 for food secure
students (California State University, 2019).
Despite these links between food insecurity and academic performance, it is often
not recognized as a reason for lack of academic success. Students have felt that it was the
college’s responsibility to help ensure students had enough food to eat, and they felt
frustrated and angry at their college for not supporting their need for food, which resulted
in their lack of mental energy and lack of focus on school as well as their physical pain of
hunger (Meza et al., 2019). For students to be successful in college, they need to have
their basic physiological needs for food, water, rest, and health are met before they can
focus on school or other priorities higher up on Maslow’s hierarchy, which is why
colleges should provide support for students’ basic needs. Short-term and long-term
solutions, such as increasing Pell grants, creating more on and off-campus jobs for
students, providing meal plan subsidies, and teaching food education must be identified to
prevent food insecurity in college students (El Zein et al., 2019).
The rising cost of college contributes to almost half of food-insecure college
students that have student debt (Shipley & Christopher, 2018). Most students who
experience food insecurity work at least part-time jobs, and 75% receive financial aid of
some sort, but the cost of college has increased 439% between the 1980s and 2010, and
consumer paychecks have not kept pace with that level of inflation thus contributing to
food insecurity (Shipley & Christopher, 2018). Five trends that contribute to food
insecurity among college students include the growing number of students who are
financially challenged, the increasing cost of college, Pell grants pay for less of a
percentage of the educational cost (60% of community college costs compared to 100%
in the past), more difficult to pay for college by working, and colleges have less money
for student support (Freudenberg et al., 2019). As a result of the increasing costs of
college, students are challenged to have enough money to pay for college and pay for
their basic living expenses (Nazmi et al., 2019). About 26% of community college
students who were food insecure earned incomes that were 85% higher than the poverty
line (Goldrick-Rab et al., 2017). Students typically have a limited budget that must go
toward housing, tuition, other living expenses, and food, thus leaving limited funds for
adequate food, and first-year college students are often in charge of their own budget for
the first time, which creates additional chances of becoming food insecure due to lack of
experience managing money and shopping for food.
The most significant predictor of being food insecure in college is struggling for
adequate nutrition before college and students’ mental health. If students did not have
adequate nutrition before college, they likely came from a food-insecure home, which
increases the likelihood of being food insecure in college. However, even if college
students were not food insecure before college, they can still be food insecure as college
students, as evidenced by the prevalence of food insecurity in all U.S. households being
11.1% in 2015 compared to 43.5% of U.S. college students being food insecure, as
identified in a systematic review of food insecurity in U.S. college students (Nazmi et al.,
2019). A study done on University of California students found 57% of students were not
food insecure when they lived at home as a child (Downing, 2017). In other words,
32.4% of U.S. college students were food insecure in college, but not in childhood;
therefore, just because students have enough food to eat before college does not mean
they will have enough food to eat when in college. College students are adults; therefore,
they are expected to support themselves with limited or no support. This illustrates the
need to determine the profile of a food-insecure college student rather than assuming if a
student were not food insecure in youth, they would not be food insecure as an adult
college student.
Henry (2017) studied the impacts of food insecurity on college students at the
University of Northern Texas and found food insecurity was associated with poor grades,
poor concentration, lack of energy, and limited participation in outside-of-class activities,
such as exercise and school functions. With students having poor grades because of food
insecurity, they are less likely to meet their educational goals and more likely to have
limited social experiences, which negatively impacts the overall collegiate experience and
negatively impacts the mental health of students. Regardless, if the consequences of food
insecurity are a nutrient deficiency or stress resulting from lack of money, food insecurity
can affect students’ ability to continue schooling (Cady et al., 2016).
Bruening et al. (2017) conducted a systematic review of studies conducted on
college students and food insecurity and found that 32.9% of college students experienced
food insecurity. Students who are food insecure have a hard time concentrating on school
(Ilieva et al., 2019). Not only does food insecurity impact academic success, but it also
increases rates of depression and decreases the chance of eating breakfast, consuming
home-cooked meals, and receiving food from parents (Bruening et al., 2016).
Additionally, students in the ten University of California campuses who were food
insecure had higher BMI levels and worse health because of fewer days of adequate
sleep, lack of exercise, and low fruit and vegetable intake, which all lead to a high BMI
and poor health (Martinez et al., 2019). The impacts of food insecurity clearly go beyond
academic success. Experiencing food insecurity during college and the poor health
behaviors and dietary habits learned during this time can last for a lifetime and increase
the chance of chronic diseases and poor nutritional status. El Zein et al. (2019) found that
students who had GPAs under 3.0 were twice as likely to be food insecure, showing that
food-insecure students may perform worse in school due to the impacts of lack of food,
which may result in anxiety, lack of sleep, fatigue, weakness, and disordered eating;
therefore, policies and programs must be put in place to screen for students at risk for
food insecurity to promote a healthier environment. Food insecurity is associated with
obesity, diabetes, depression, poor nutrient intake, and poor self-reported health, as well
as poor academic outcomes, such as lower GPAs and higher drop-out rates (Freudenberg
et al., 2019). In rural Nova Scotia, Frank (2018) found that 38% of students were food
insecure, and associations were revealed between food insecurity and living
arrangements, money for school and food, poor overall health, poor mental health, high
stress, and poor academic performance. These outcomes affect multiple areas of student
lives and can have long-term consequences. Students who were food insecure had lower
academic success and had worse health outcomes (Bruening et al., 2017; Maynard et al.,
2018; O’Neill & Maguire, 2017). Since lifestyle habits are established when young, being
food insecure can lead to continued poor lifestyle behaviors, which sets students up for a
lifetime of poor health outcomes, both mentally and physically. Meza et al. (2019)
investigated how food insecurity impacts students at a large public university in
California by conducting interviews with food-insecure students. Food insecurity affected
student relationships; gave feelings of hopelessness and undeserving of help, frustration
with the college for not helping enough, and impacted their academic performance
because they were hungry and mentally unfocused. Besides being less likely to complete
courses and degrees, there are also lasting physical and mental effects of food insecurity
in college students (Que & Baldridge, 2019). Given the high rate of food insecurity in
California Community College students, food-insecure students must be identified so
resources can be given to food-insecure students.
Supporting Student Needs
The amount of financial aid debt, exogenous financial shock, and use of food
assistance programs were related to food insecurity; therefore, basic skills training and
education are needed to help minimize food insecurity in college students living
offcampus (Knol et al., 2018). Colleges should identify resources to help support
students' need for access to adequate food intake. Sometimes resources are available, but
students are not aware of the resources or are too embarrassed to access the resources.
Sharing information through social networks and education campaigns is vital to increase
awareness, acceptance, and utilization of available food resources for students (Ortiz et
al., 2019). Students at UCLA reported their food needs were not met from the UCLA
food environment, and they wanted more financial and food literacy life skills, which
could help address food insecurity (Watson et al., 2017). Teaching students how to cook
foods at home and how to shop smartly can help decrease food insecurity. Both shortterm
and long-term solutions must be identified to help support students’ basic needs to
increase the chances of student learning and student success (Nazmi et al., 2019). I
conducted an assessment with this study to determine the prevalence of food insecurity at
HCC. Policy changes should happen at local, state, and federal levels to include college
students in the list of eligibility criteria for services (O’Neill & Maguire, 2017). Some
colleges have set up food pantries, outreach and education programs, and campus food
vouchers with the help of student advocates who conduct outreach and education to
students on food insecurity (Freudenberg et al., 2019). Woerden et al. (2019) calculated
the lost revenue from students dropping out of school was 50 million dollars per year;
therefore, figuring out strategies to support food insecure students will increase the
finances of colleges since food-insecure students are more likely to drop out of school.
Fisher and Crawford (2020) described a school that went from a school in crisis to
a distinguished school because of their focus on students’ physiological needs, which
showed students school staff cared about them personally, and then the staff was able to
focus on each of the other steps in Maslow’s Hierarchy. This study was not performed at
a higher education institution, but the outcomes are likely to be the same. Using
Maslow’s Hierarchy of Needs as a supporting theoretical framework, Tichy (2017)
studied a charter school to help determine how schools can help support at-risk students
so students can progress up Maslow’s Hierarchy of Needs. Until society focuses on how
to grow students and offer protective factors, Maslow believed students could not
progress up his Hierarchy (Tichy, 2017). Therefore, supporting students’ basic needs can
lead to improved educational outcomes.
A USDA-funded health promotion study, Get FRUVED, consisted of data from
participants from eight U.S. universities in eight states. The data indicated that to improve
the academic performance and health of students, colleges should screen for food
insecurity and implement programs to encourage a healthy college experience (El Zein et
al., 2019). To make college more affordable for students and to equalize the college
experience for all, a comprehensive assessment of the costs and funding of college should
be conducted along with initiating federal efforts to combat food insecurity in college
students (Nazmi et al., 2019).
Students attending community colleges typically have greater support needs than
students attending other types of schools; however, community colleges generally offer
fewer food assistance resources to students (Goldrick-Rab, 2018). Students attending
community colleges have not been studied in-depth, possibly because the cost of
attending a community college is cheaper; however, community college students are
highly food insecure, even though they receive financial aid and tend to work at least
part-time (Broton & Goldrick-Rab, 2018). Instead of considering supporting students’
basic needs a social support initiative, administrators should consider the effort a push to
increase student success (Goldrick-Rab, 2018). One strategy community colleges have
implemented to increase student success is the guided pathway structured model of
offering specific program pathways for students to help with motivation and real-world
applicability (Mann, 2018). Food is considered a level 1 physiological need, which must
be met before students can move to the next level of safety of having a secure job, and
then to level 3 of love and belonging, which can include the desire for a better job or
higher level of knowledge; this is where education comes in on Maslow’s Hierarchy of
Needs, and level 4 is the esteem category, which is where accomplishing a goal of
completing a degree fits (Schulte, 2018).
There are many actions schools can take to provide support for students by doing
simple things, such as offering a listing of currently available resources and providing
referrals for students. Prior to 2011, there were only two studies done on food insecurity
in college students; therefore, colleges were not aware of the basic needs of their students
(Goldrick-Rab et al., 2017). With the cost of education and the costs for basic needs
increasing more than ever, food insecurity is a more significant issue than ever before and
must be dealt with to further student success and to support the Precautionary Principle
that says health professionals must take action when there is enough compelling evidence
of the health benefit of taking action (Freudenberg et al., 2019). According to the Hope
Center report, colleges must change their thinking about basic needs security and begin
engaging with the community and private organizations, as well as governmental
programs to help prevent students from experiencing basic needs insecurity so that
students can focus on learning, which can ultimately increase completion rates
(BakerSmith et al., 2020). College leaders and community members must take
responsibility for ensuring that lack of food is not a reason students do not complete
college (Que & Baldridge, 2019). Cady et al. (2016) suggested colleges have food
pantries, food recovery programs, meal donations, help increase access to benefits, such
as SNAP, and create community gardens and farmers’ markets on campus.
USDA Six-Item Short Form Survey Tool
The USDA Six-Item Short Form Survey is a standardized survey instrument used
to identify and measure food insecurity in households. This tool is widely accepted and
utilized in food insecurity studies. According to a systematic review of food insecurity in
U.S. college students, there were eight studies that met inclusion criteria and were
included in the study; six of those eight studies used the USDA Six-Item Short Form
Food Security Survey (Nazmi et al., 2019). The advantage of using this Short Form is
students are more likely to understand the questions and complete the survey. The
downside of using this survey tool is that the survey instrument has not been validated for
use with college students. However, this is the most valid and commonly used tool that
currently exists.
First-Generation Students
A first-generation college student is one who is the first person in his family to
attend college or is a student whose biological parents have not completed a 4-year
degree (Center for First-Generation Student Success, 2017). Given the variance in
definitions, when comparing data on first-generation students, it is important to
understand how first-generation students were defined in each study. An additional
challenge for first-generation students is the potential lack of cultural capacity for
attending college (Center for First-Generation Student Success, 2017). Being a
firstgeneration college student does not mean a student is poor, just as being poor does
not automatically mean a student is a first-generation student; however, many poor
students are first-generation college students (Goward, 2018). This is an important
distinction since most studies classify all first-generation students in the same way. If
firstgeneration students have adequate support and resources, they are more likely to
graduate even though they are first-generation students. First-generation students are less
likely to graduate, and if poverty is also an issue for first-generation students, the success
rate is even lower (Goward, 2018). The highest rate of food insecurity was 65.9% of
Black students who were also first-generation students (California State University,
2019).
Among community college students in the U.S., 29% are first-generation college
students (Center for First-Generation Student Success, 2020). Camelo and Elliott (2019)
studied academic achievement and food security at a public university in the U.S. and
found first-generation college students, Black students, and Hispanic students were more
likely to be food insecure and have lower GPAs. Since community colleges have a greater
number of minority students and first-generation students, community colleges are more
likely to have a high number of food-insecure students. Many community college
students do not complete their programs of study or earn a degree which is a problem for
society, the economy, and employers in need of a trained workforce. Davidson and
Morrell (2020) found 25% of their study participants at a northeast university to be food
insecure, with first-generation students and those receiving financial aid most likely to be
food insecure. These are students who have a high need for support.
Single Parents
A single parent is someone who is raising at least one child without a partner. In
community colleges in the U.S., 15% of all students are single parents (Center for First-
Generation Student Success, 2020). Spaid and Gillet-Karam (2018) conducted a
mixedmethods study on community college students in Maryland to investigate food
insecurity and the relationship to concentration levels and energy as determinants of
student success. Women, single parents, and minority women were more likely to
experience food insecurity. According to Spaid and Gillet-Karam (2018), multiple studies
have found women, especially minority women and single parents, were more likely to be
food insecure.
Since there was no national survey or effort to determine the level of basic needs
insecurity in college students, the Hope Center coordinated a survey of community
college students to figure out students’ basic needs security. The results were included in
a Report titled Hungry to Learn and had results from 33,000 students from 70 different
community colleges throughout 24 states. The results of those surveys showed almost
67% of these 33,000 racially and economically diverse community college students were
food insecure, with 33% of the students being highly food insecure, predominantly
minority students, and 63% of single parents were food insecure (Goldrick-Rab et al.,
2017). The Hope Center survey results found a 16% higher rate of food insecurity in
parenting students and a 23% higher rate of food insecurity in students who had been in
foster care (Baker-Smith et al., 2020).
Ethnicity and Gender
Of all Hispanic undergraduate students in the U.S., 56% are enrolled in
community colleges, and 44% of Black students are enrolled in community colleges, plus
community college students are more likely to be low-income (Ilieva et al., 2019). Some
geographic areas have community colleges with even higher percentages of minority
enrollment. Martinez et al. (2018) studied the prevalence of food insecurity and
associated risk factors; age, ethnicity, childhood food insecurity, and receiving financial
aid, and found students who were food insecure had higher rates of difficulty
concentrating, lower academic performance, and had a lack of money to purchase food,
especially healthful food. Ilieva et al. (2019) surveyed minority students at a Hispanic
Serving Community College in New York and found 74% of students had very low food
security, and 26% had low food security; however, very few students used food
assistance, and only 20% of students even knew there were resources on campus to help
with food insecurity. As a result of being food insecure, students were not as able to focus
on school and were less likely to complete their degrees (Ilieva et al., 2019). According to
Payne-Sturges et al. (2018), African Americans who were housing insecure were more
likely to be food insecure, increasing concerns about student success. Broton et al., (2018)
found students who were food insecure as children had a 40% chance of being food
insecure as college students, and 35% of racial/ethnic minorities experienced food
insecurity.
Black households experienced a 21.2% rate of food insecurity, and Hispanic
households had a 16.2% rate, while the national average of household food insecurity was
11.1%. Students of color had higher rates of food insecurity (Bruening et al., 2017). Men
of color are more likely to experience food insecurity, especially if they have children,
because oftentimes, the men will ensure their kids have enough to eat before they will eat
(Vasquez et al., 2019). This practice can hinder academic success for men of color. Public
colleges show a graduation rate of 37.5% for Black men, regardless of their income level
(Goward, 2018). According to the Hope Center’s Real College survey in 2019, 54% of
Black students, 47% of Hispanic or Latinx students, and 60% of Indigenous students were
food insecure in comparison to 36% of white students; with females, non-binary, and
transgender people experiencing more food insecurity at much higher rates than males;
and gay, lesbian, or bisexuals had a higher rate of food insecurity than heterosexuals
(Baker-Smith et al., 2020).
In households with children headed by a single woman, 27.8% of those
households were food insecure in 2015 compared to 15.9% of households with children
headed by a single man and compared to the national average of 11.1% of households
experiencing food insecurity (USDA ERS, 2019). In addition, women living alone
experienced a food insecurity rate of 14.2% compared to 12.5% of men living alone.
Food insecurity in college students will follow the same pattern.
Racial and sexual undergraduate minority students experienced food insecurity at
a higher rate, as did undergraduate students whose families were lacking in money and
students who did not have secure housing (Willis, 2019). A national basic needs security
survey was given to college students at 123 two- and 4-year colleges in the U.S. Students
at 2-year colleges had higher rates of food insecurity as did African Americans, LGBTQ
students, veterans, former foster youth, and convicted criminals (Goldrick-Rab et al.,
2017). Because being poor impacts food security, student success is also affected.
Goward (2018) suggests including social class as one of the diversity initiatives at
colleges since poverty can impact many variables associated with student success.
Non-Traditional Students
A non-traditional student is a student outside of the typical 18-24-year-old age
fresh out of high school. According to the National Center for Education Statistics (n.d.),
a non-traditional student meets any of the following criteria: is enrolled in college on a
part-time basis, does not start college immediately after high school, works full-time, is
self-supportive financially, has at least one child, is a single parent, or does not have a
high school diploma. Non-traditional students tend to be ethnic minorities, women, and
students with parents who have less education than parents of traditional students
(National Center for Education Statistics, n.d.). One characteristic of a non-traditional
student is one who works while attending school. In community colleges throughout the
US, 62% of all full-time students work, and 72% of all part-time students work while
attending school (Center for First-Generation Student Success, 2020). A study of college
students and their basic needs insecurity was conducted and found one-third of the
students who were food insecure worked and received financial aid (Goldrick-Rab et al.,
2017). When students must work while going to school, they will have less time to focus
on school, thus could be less successful academically. Students who attend community
colleges are more likely to be racially and economically disadvantaged and are more
likely to need financial assistance. The results of a national survey of college students
revealed students over age 25 were more likely to experience food insecurity
(GoldrickRab et al., 2017).
In the fall of 2017, Spaid and Gillet-Karam (2018) surveyed over 200 Maryland
Community College students on food insecurity and student success and found more
women than men experienced food insecurity; especially minority women, women over
20, and women who received Pell Grants, plus they found more of the food insecure
students were single parents. Receiving Pell Grants or federal financial aid is not the only
indicator of food insecurity for anyone, nor is receiving a grant a complete solution to
food insecurity. Many food-insecure college students receive financial aid but do not
receive public assistance, such as SNAP or cash aid (Broton et al., 2018).
Summary and Conclusions
Food is a basic physiological need, and in the absence of this basic need, it
becomes difficult to focus on meeting higher levels of needs. Those of color, women,
single parents, first-generation college students, and non-traditional students are more
likely to be food insecure. Food insecurity creates a roadblock for community college
students to be able to complete their degrees (Goldrick-Rab et al., 2017). Lack of food is
a problem that must be addressed before food insecure students will be able to focus their
energy on obtaining a higher education degree. And before food insecurity on college
students can be addressed, the prevalence of food insecurity and the profile of a
foodinsecure student must be identified. Figuring out the profile of a food-insecure
student will enable colleges to provide referrals and resources to students at risk of being
food insecure, and colleges can advocate for policy changes regarding the affordability of
college (Payne-Sturges et al., 2018).
Community College students in the Central Valley of California must be surveyed
to figure out the profile of a food-insecure student. Once surveys are analyzed, a profile
of those most likely to be food insecure will be determined, which will address the gap in
knowledge, and then the college can create a network of support for these students. With
greater support, food-insecure students will have the resources they need to prevent
hunger. If fewer students are food insecure, more students will be able to achieve a higher
level of needs, including focusing on a higher education degree.
Chapter 3: Research Method
The purpose of this quantitative study was to determine the profile of a
foodinsecure college student in the Central Valley of California and to determine whether
relationships existed between food insecurity and ethnicity, gender, being a
firstgeneration college student, being a non-traditional student, and being a single parent
in students at HCC in the Central Valley of California. In this chapter, I describe the
method used to collect data, the design and approach of the study, the survey instrument
used, method of sampling, target population, statistical tests, and the variables that were
studied. Self-reported rates of food insecurity were used and measured from the results of
the USDA Six-Item Short Form (USDA ERS, 2012). The results of this study provide
data for community colleges to be aware of the prevalence of food insecurity and to know
students at risk for food insecurity. Currently, there is limited data on the profile of food
insecure community college students in the Central Valley.
Research Design and Rationale
The research design of this study was a quantitative correlational design. A
quantitative design was chosen because I tested for relationships between variables in my
hypotheses, and the resulting data were analyzed statistically (Burkholder et al., 2016).
As part of quantitative research, correlational designs are commonly used when
identifying relationships between variables (Burkholder et al., 2016). The dependent
ordinal variable was food insecurity, and the independent categorical variables were
ethnicity, gender, being a first-generation college student, being a non-traditional student,
and being a single parent. There were not any time or resource constraints with this
quantitative correlational research design.
By figuring out the characteristics of students who are likely to be food insecure,
colleges will have data that can be used to proactively identify students who may be at
risk for food insecurity. Colleges can use this data to target at-risk students by providing
needed support for students’ basic needs. By supporting students’ basic needs, colleges
will ultimately be supporting student success because students can focus on school
instead of where they will get their next meal. The results of this research can advance
knowledge in the field of education and student success, as well as advance public health
knowledge of the impacts of food insecurity going beyond physical and mental health
issues.
Methodology
Population
The study was conducted with community college students at HCC, found in a
large agriculture city in the Central Valley of California. HCC has a student population of
approximately 40,000 students, of which 10,000 are high school dual enrollment students.
Dual enrollment students are students who attend college and high school at the same
time. Since dual enrollment students are minors and more likely to be supported by their
parents, they were excluded from this study. The target population consisted of the
remaining 30,000 students at HCC.
Sampling and Sampling Procedures
The target population of this study included 30,000 non-dual enrollment HCC
students. A total population sampling technique was utilized to survey the large student
population at HCC. Total population sampling is a sampling technique used when a
research design calls for selecting an entire group that has something in common
(Burkholder et al., 2016).
Students at HCC received an email explaining the purpose and benefit of the
study, and they were asked to complete the anonymous survey. Student names were not
collected in this research. The survey was conducted at one point in time and did not
involve an experiment, intervention, or follow-up. The survey questions consisted of
demographic questions and questions on food security. To determine food insecurity
levels, the USDA Six-Item Short Form questions were asked (USDA ERS, 2012). Some
questions had multiple answers to choose from, and some questions had dichotomous
yes/no answers. The survey was sent to all 30,000 non-dual enrollment students at HCC.
A G*Power analysis states the number of participants needed to make sure survey
results are genuinely due to the results and not due to a sample size being too small (Faul
et al., 2007). I conducted a priori G*Power analysis for a two-tailed regression with a
power of 0.8, an alpha of 0.05, and a medium effect size, which indicated a needed
sample size of 250 participants minimum. This study used all 318 valid student responses
from the spring 2020 semester, even though the Priori G*Power analysis said only 250
respondents were required.
Procedures for Recruitment, Participation, and Data Collection
An anonymous survey was emailed to students from the student email listserv at
HCC. Students received an email request asking them to complete the survey. The
emailed survey began with an informed consent statement which students had to agree to
before they were allowed to begin the survey. The following demographic data were
asked on the survey: ethnicity, gender, first-generation college student, non-traditional
student, and single-parent status. Students were also asked six questions on their level of
food insecurity. The answers to the survey questions were collected electronically
through SurveyMonkey’s confidential system. Once students completed the survey, they
received a notification they completed their expectations for the study, and no further
work was needed. After 1 week, the 250 required responses had not been received, so
students received a second email asking them to complete the survey. There was no
intervention, experiment, or treatment included in this study, nor were archival data used
in this study.
Instrumentation and Operationalization of Constructs
In a review of 31 studies, the U.S. GAO (2019) found that being a first-generation
student increased the likelihood of being food insecure, as did being low income, being a
single parent, receiving SNAP, having a disability, being a former foster youth, and being
homeless. Women, minority women, and women over 20 years of age, even though these
women were more likely to receive a Pell Grant, were most likely to be food insecure in
the 16 community colleges that were studied (Spaid & Gillet-Karam, 2018). The results
of the Real College Survey of 33,000 students at 411 colleges across the U.S. showed that
racial/ethnic minorities are more likely to be food and housing insecure (Baker-Smith et
al., 2020; Broton et al., 2018; Bruening et al., 2017; Spaid & Gillet-Karam, 2018; Willis,
2019). Non-traditional students, students who are over 25 years of age, as well as women
and minorities, are more likely to receive a Pell Grant, and receiving a Pell Grant makes it
more likely students will be food insecure (Goldrick-Rab et al., 2017; Spaid &
GilletKaram, 2018). As a result of ethnicity, gender, being a first-generation college
student, being a non-traditional student, and being a single parent being risk factors for
being food insecure in other studies, they are the variables that were evaluated in this
study.
The standardized USDA Six-Item Short Form Survey instrument was created by
the National Center for Health Statistics and Abt Associates, Inc. The USDA survey was
utilized to assess the level of food insecurity in HCC students (USDA ERS, 2012). This
tool is a subset of the USDA 18-Item Form that was created by the USDA in 1995 and
last revised in 2012. The Six-Item Short Form is considered an acceptable substitute for
the 18-item form and has the advantage of lowered respondent burden (USDA ERS,
2012). The limitations of this tool include being less precise than the 18-item tool and
does not identify the worst levels of food insecurity (USDA ERS, 2012). However, in the
general population, the short form was shown to correctly identify 99% of food insecurity
in households without children and 97.7% in households with children (Blumberg et al.,
1999). The USDA publishes this survey tool on its website and allows for free use
without permission needed. This form is a highly utilized tool to assess food insecurity in
college students because of its simplicity. There have been six other studies on food
insecurity in college students that have used the USDA Six-Item Short Form Food
Security Survey and tested the reliability and validity of the tool (Nazmi et al., 2019). The
survey questions and operational measures are listed below in Table 1.
Table 1
Description of Operations Measures for Questions
Dependent
Variables
USDA Six-Item Food Security
Survey Questions
Response Category Response Code Type of Variable
Food
insecurity
Q1: The food that I bought just
didn’t last, and I didn’t have money
to get more.” Was that often,
sometimes, or never true for you in
the last 30 days?
-Often true
-Sometimes true
-Never true
-DK
Yes = 1
Yes = 1
No = 0
No = 0
Ordinal with three
groups
Q2: I couldn’t afford to eat
balanced meals.” Was that often,
sometimes, or never true for you in
the last 30 days?
-Often true
-Sometimes true
-Never true
-DK
Yes = 1
Yes = 1
No = 0
No = 0
Q3: In the last 30 days, did you
ever cut the size of your meals or
skip meals because there was not
enough money for food?
-Yes
-No
Yes = 1
No = 0
Q3A: If yes to Q3 above, in the ____ Days >3 = 1
last 30 days, how many days did
this happen?
-DK
<3 = 0
No = 0
Q4: In the last 30 days, did you
ever eat less than you felt you
should because there was not
enough money for food?
-Yes
-No
-DK
Yes = 1
No = 0
No = 0
Q5: In the last 30 days, were you
ever hungry but did not eat because
there was not enough money for
food?
-Yes
-No
-DK
Yes = 1
No = 0
DK = 0
Ethnicity What is your ethnicity? -Asian
-Black or African
American
-Hispanic/Latino -
Native American or
other Pacific
Islander
-White
-Unknown
1
2
3
4
5
6
Categorical
(Nominal)
Gender What is your gender Male
Female
Nonbinary
1
2
3
Categorical
(Nominal)
Firstgeneration
College
Student
Do either of your parents have a
4year college degree?
Yes No 1
0
Categorical
(Dichotomous)
Non-
traditional
Student
What category best describes your
age?
18-24 25
or older
1 – Yes
0 - No
Ordinal
Single Parent Are you raising a child without a
partner?
Yes No 1
0
Categorical
(Dichotomous)
USDA Six-Item Short Form Scoring
Each variable was scored based on answers to the survey. Responses of often or
sometimes on questions 1 and 2 and yes responses on questions 3, 4, and 5 were coded as
yes and were assigned a value of 1. All other answers were recorded as no and were
assigned a value of 0. On Q3A, responses of three days or more were coded as yes and
were assigned a value of 1, and fewer than three days were coded as no and were
assigned a value of 0. The total of yes responses to the six questions provided a raw score
on the USDA Six-Item Short Form Survey Scoring Scale (USDA ERS, 2012). Food
security status was based on the scoring listed in Table 2. The scoring results of the
USDA Six-Item Short Form Survey can be considered an ordinal variable with three
groups or a binary variable with two groups. The groups were based on the USDA
scoring results for an ordinal variable with high, low, or very low food security groups. If
the binary variable option is chosen, the two groups are food insecure and food secure.
This study uses the option of using three groups.
Table 2
USDA Six-Item Short Form Survey Scoring Using an Ordinal Variable With Three
Groups
Raw Score Food Security Status Group Food Security Definitions*
0-1 High or marginal food Group 1 High food security: no reported security (raw score 1
indications of food-access problems may be considered or limitations
marginal food security.)
Marginal food security: one or two
reported indications -typically of
anxiety over food sufficiency or
shortage of food in the house. Little
or no indication of changes in diets
or food intake
2-4 Low food security Group 2 Low food security: reports of
reduced quality, variety, or
desirability of diet. Little or no
indication of reduced food intake
5-6 Very low food security Group 3 Very low food security: reports of
multiple indications of disrupted
eating patterns and reduced food
intake
SPSS version 27 was the software used to conduct statistical analyses of the data
once the results from the student surveys had been received. I screened the data to ensure
all survey questions had been answered, and then the results were congregated and set up
correctly for statistical analyses.
Research Question and Hypothesis
Research Question: Is food insecurity related to ethnicity, gender, being a first-generation
college student, being a non-traditional student, or being a single parent in HCC students?
H0: There is no relationship between food insecurity and ethnicity, gender, being a
first-generation college student, being a non-traditional student, or being a single parent
in HCC students.
Ha: There is a relationship between food insecurity and ethnicity, gender, being a
first-generation college student, being a non-traditional student, or being a single parent
in HCC students.
Statistical Tests
An ordinal logistic regression test was performed to assess the relationship
between one ordinal dependent variable and multiple independent categorical variables
(Frankfort-Nachmias & Leon-Guerrero, 2018). Ordinal logistic regression tests are used
to predict the impact independent variables have on an ordinal dependent variable and to
determine any interactions between the independent variables (Frankfort-Nachmias &
Leon-Guerrero, 2018). Ordinal logistic regression tests are multiple binomial logistic
regression tests run at the same time using cumulative logits with each regression test
predicting the probability of being categorized in each category, with each category
having an equal odds ratio; referred to as a proportional odds model (Udo &
Akwukwuma, 2019).
Using SPSS version 27, an ordinal logistic regression test was performed to
determine any relationships between food insecurity and ethnicity, gender, being a
firstgeneration college student, being a non-traditional student, and being a single parent.
The dependent ordinal variable was food insecurity. The independent categorical
variables were ethnicity, gender, being a first-generation college student, being a non-
traditional student, and being a single parent. The test explained the relative contribution
of each independent categorical variable in relation to the variance. There were no time or
resource limitations for this study.
Assumptions Made When Running an Ordinal Logistic Regression Test
According to Udo and Akwukwuma (2019), to determine if data fits an ordinal
logistic regression test, four assumptions are made; two of which the design of the study
must meet:
1. The dependent variable is an ordinal variable.
2. At least one of the independent variables must be continuous or categorical.
Two assumptions that SPSS can test for:
3. No multicollinearity can exist between two or more independent variables,
meaning the variables cannot be highly correlated.
4. There should be proportional odds that each of the independent variables has the
same effect. The GENLIN procedure was used to test for the assumption of
proportional odds.
Threats to Validity
External
External validity occurs when a dependent variable is measured under natural and
normal conditions (García-Pérez, 2012). The external threats to validity included
volunteer and self-selection bias; perhaps students who fill out surveys have something in
common that could influence the outcome of the study. The results of this study are only
generalizable to other community college students in the Central Valley of California.
Community college students in other areas of California or the U.S. may have different
outcomes due to having a different student population. Students at 4-year universities
may have different results as well.
Internal
Internal validity of a study is obtained when researchers control for confounding
factors and are confident in the accuracy of their conclusions (García-Pérez, 2012). One
threat to internal validity in this study is the history effect. As a result of the COVID
pandemic, many people have lost jobs and may be more likely to be food insecure
compared to before the pandemic. Many of the jobs lost during the pandemic were
service industry jobs, and many college students were employed in the service industry
(Owens et al., 2020).
Construct and Statistical Conclusion Validity
Construct validity is an ongoing process that researchers engage in to assess the
validity of a measurement instrument that is used to measure a construct, such as a survey
to measure food insecurity (Burkholder et al., 2016) Content validity, criterion validity,
convergent validity, and divergent validity are included in the assessment of construct
validity, which is why construct validity cannot be determined in one study (Burkholder
et al., 2016). To ensure construct validity, researchers should use clearly established
definitions and measurement types (García-Pérez, 2012).
The primary threats to construct validity are the history effect of the COVID
pandemic, and the USDA Six-Item Form has not been validated in college students
specifically, despite being widely used with college students. However, the USDA
SixItem Short Form is considered a good alternative to the USDA 18-Item Form.
Statistical conclusion validity means study conclusions were based on proper data
analysis that holds up to scrutiny (García-Pérez, 2012). The threat to statistical conclusion
validity could have been if too few students completed surveys and results were provided
as if there was statistical significance.
Ethical Procedures
Human Participants
HCC students were asked to complete an anonymous survey that was emailed to
students at their college emails. The Walden IRB and the IRB at HCC approved the
student survey (Walden IRB approval number 05-05-21-0739144). There were no ethical
concerns with conducting this study because the study consisted only of an anonymous
survey at one point in time. There were no ethical concerns related to data collection
because the surveys were anonymous. The study site is where I work; however, there
were not any conflicts of interest or potential for gain from doing this anonymous student
survey.
Treatment of Data
All information obtained in the survey was anonymous; therefore, there was not
any confidential data to protect. All data obtained from this study will be stored on a
password-protected computer for at least 5 years and then will be permanently deleted.
Results from this study were shared with the dissertation committee and the IRB of
Walden and HCC.
Summary
This chapter describes the methods and research design that were used in this
study. The research study consisted of a quantitative correlational design using a total
population sampling technique to conduct a survey that identified the prevalence of food
insecurity and any relationship between the independent ordinal variables and food
insecurity. An ordinal logistic regression test was performed to evaluate any relationship
that existed between the dependent ordinal variable food insecurity and the independent
categorical variables of ethnicity, gender, being a first-generation college student, being a
non-traditional student, and being a single parent. This research contributed to the
knowledge of the prevalence of food insecurity in community college students and also
gave needed data for community college administrators to identify students at risk for
food insecurity.
Chapter 4: Results
The purpose of this quantitative study was to determine the profile of a
foodinsecure college student in the Central Valley of California and determine any
relationships between food insecurity in HCC students and ethnicity, gender, being a
first-generation college student, being a non-traditional student, and being a single parent.
Data collection methods, a statistical analysis of the data, survey results, and key findings
are explained in this chapter.
Data Collection
Participants were not surveyed until after approval was received from the Walden
University Institutional Review Board and HCC Institutional Review Board. On May 6,
2021, using the college student listserv, an email request was sent to 30,000 HCC
students asking them to complete a survey on whether they had experienced food
insecurity via a link to the survey on SurveyMonkey. Students had 2 weeks to complete
the survey and were reminded to complete the survey after 1 week. At the end of the
survey period, 438 students had responded, with 120 incomplete responses, leaving 318
valid survey responses. There were no discrepancies in data collection from the original
plan. The G*Power analysis indicated a need for 250 student responses, but instead of
choosing 250 of the 318 valid responses, all 318 responses were used in the survey. No
covariates existed in the study. No treatment or intervention was included in the study.
Descriptive and Demographic Statistics
The entire HCC student body was included in the request to complete the survey,
so no bias existed in selecting the participants. There were 318 HCC students who
completed the survey. All students were current students at HCC, a community college in
the Central Valley of California. The ethnic breakdown of the student respondents was 18
(6%) Asians, 21 (7%) Blacks, 183 (58%) Hispanic/Latino, 10 (3%) Native Americans,
and 86 (26%) Whites. A comparison of internal documents at HCC shows the ethnic
breakdown of this study was representative of the average actual ethnic breakdown of the
college as a whole.
Of the 318 students who responded, 79 (25%) were males, and 239 (75%) were
females. Females were represented at a higher rate in this study compared to the national
average of 57%; however, females reply to surveys more often than males (Goldrick-Rab
et al., 2017). A large-scale food insecurity study that was conducted on 33,000 students at
70 community colleges in 24 states also had a high percentage of female respondents
compared to males, with 72% of their respondents being female (Goldrick-Rab et al.,
2017). Females make up 60% of the total student body at HCC. Females represent more
than half of the student population in colleges across the United States (Marcus, 2017).
My results showed that there were 259 (81%) students who were considered
firstgeneration college students since neither of their parents had a 4-year degree, and
there were 59 (19%) students who were not first-generation college students. Of all
students at HCC, 69% of students were first-generation students. Perhaps a higher
percentage of first-generation students completed my survey because they actually
experienced food insecurity and wanted their voices to be heard.
Of the 318 students, 156 (49%) were non-traditional students based on their age
being 25 years and older, and 162 (51%) were traditional students since they were under
25 years of age. Of all HCC students, 32% were non-traditional students. Older students
may see the value in participating in surveys, so they might be more likely to complete
surveys. Plus, older people may have more responsibilities and expenses, thus making
them more likely to experience food insecurity, which may make them more likely to
answer a food insecurity survey.
There were 36 (11%) students who responded that they were single parents,
meaning they were raising at least one child without a partner, and 282 (89%) students
who were not single parents. Raising a child/children requires more money; therefore,
some of the single parent's resources must be spent on the needs of the child/children,
leaving less money to support the needs of the student parent.
Of the 318 students who responded to my survey, 189 (59%) experienced food
insecurity according to the USDA survey criteria, meaning they did not have enough food
or enough healthy food to eat. The HOPE Food Insecurity Survey of community colleges
in 24 states also found 59% of students from schools in the West were food insecure
(Baker-Smith et al., 2020). One potential factor not analyzed was whether students were
part-time or full-time students. Only 30% of HCC students were full-time students, which
could influence the level of food insecurity of students, especially considering The Hope
Center Survey found that part-time students and students who took longer to graduate
were more likely to be food insecure (Baker-Smith et al., 2020).
Data Analysis
Research Question and Hypothesis
Research Question: Is food insecurity related to ethnicity, gender, being a
firstgeneration college student, being a non-traditional student, or being a single parent in
HCC students?
H0: There is no relationship between food insecurity and ethnicity, gender, being a
first-generation college student, being a non-traditional student, or being a single parent
in HCC students.
Ha: There is a relationship between food insecurity and ethnicity, gender, being a
first-generation college student, being a non-traditional student, or being a single parent
in HCC students.
Ordinal Logistic Regression Assumptions
An ordinal logistic regression test was performed to determine any relationship
between food insecurity and ethnicity, gender, being a first-generation college student,
being a non-traditional student, and being a single parent. Food insecurity was the
dependent ordinal variable. Ethnicity, gender, being a first-generation college student,
being a non-traditional student, and being a single parent were the independent
categorical variables. In order for an ordinal logistic regression analysis to be performed,
four assumptions must be met to ensure valid results (Udo & Akwukwuma, 2019):
1. The dependent variable must be an ordinal variable.
2. At least one of the independent variables must be continuous or categorical.
3. No multicollinearity can exist between two or more independent variables.
4. There should be proportional odds that each of the independent variables has
the same effect.
For the first assumption, my dependent variable was an ordinal variable with three
groups. The three groups were high/marginal food security, low food security, and very
low food security. For the second assumption, my data included five independent
categorical variables: ethnicity, gender, being a first-generation college student, being a
non-traditional student, and being a single parent. For the third assumption, all my
independent variables were categorical, so they could not be collinear. Finally, for the
fourth assumption, the Genlin procedure was used to test for the assumption of
proportional odds. The p value must be greater than 0.5. The main assumption of
proportional odds was not met by my data as assessed by the Test of Parallel Lines; X2(8)
= 19.185, p=.014. My data violated this main assumption of the ordinal logistic
regression model by having non-proportional odds. Consequently, the ordinal logistic
regression analysis would not provide valid data, so it could not be used in my study.
According to the seminal work of Bender and Grouven (1998), when ordinal data violates
the assumption of non-proportional odds, separate binary logistic regression tests can be
run, but doing so is not as efficient as conducting an ordinal logistic regression test.
However, the USDA states the U.S. Household Food Security Survey Six-Item Short
Form survey is reliable and valid if the dependent variable is binary rather than ordinal
(USDA ERS, 2012). When reporting the survey results by using a binary analysis, the
categories of low food security (raw score of 2-4) and very low food security (raw score
of 5-6) should be combined as one group of food insecure (raw score of 2-6) and the
second group should be food secure (raw score 0-1; USDA ERS, 2012).
Additional Statistical Test of Hypothesis
Since my data failed the main ordinal logistic regression assumption by having
non-proportional odds, an ordinal logistic regression analysis could not be used with my
data due to a lack of validity. And since the USDA allowed for reporting survey results in
a binary format instead of ordinal, I transformed the data to match the requirements of the
USDA by collapsing the three groups into two groups. The two groups were food
insecure and food secure. According to the USDA ERS (2012), “For some reporting
purposes, the food security status of households with raw score 0-1 is described as food
secure, and the two categories ‘low food security’ and ‘very low food security’ in
combination are referred to as food insecure” (p. 4). With having two groups for the
dependent variable instead of three, a binary logistic regression analysis was the
appropriate statistical test to use. I used the transformed data that was compliant with the
USDA standardized survey requirements and performed a new data analysis using a
binary logistic regression analysis. The binary logistic regression was performed to
answer the research question below.
Data Transformation
The data was transformed to coincide with the USDA requirements for binary
data. The scoring results of the USDA Six-Item Short Form Survey can be considered a
binary variable with two groups, with the groups being based on the USDA scoring
results for food insecure and food secure. Group 1 included those with a raw score of 0–
1, who were considered food secure, and Group 2 scored between 2–6 and were
considered food insecure.
Binary Logistic Regression Analysis
A binary logistic regression predicts the chance that an observation or result will
fall in one of two categories of a dependent variable based on independent variables that
are continuous or categorical (Frankfort-Nachmias & Leon-Guerrero, 2018). After
transforming the data to include two groups for the dependent variable in the binary
logistic regression, the two groups were food insecure and food secure.
According to Udo and Akwukwuma (2019), For a binary logistic regression to be
valid, there following seven assumptions must be met:
1. There must be a dichotomous dependent variable. My dependent variable
consisted of two groups, food secure and food insecure.
2. At least one independent variable must be measured on a nominal or
continuous scale. All of my independent variables were categorical.
3. All nominal independent variables must be mutually exclusive, and the
categories of the dependent variable should be independent. My study had five
independent variables, and they were all mutually exclusive. Ethnicity,
gender, being a first-generation college student, being a non-traditional
student, and being a single parent were all mutually exclusive, so they could
not be in two categories.
4. There must be at least 15 cases per independent variable. My study had 318
responses in each of the five categories of independent variables.
5. Assumption of linearity. The continuous independent variable and the logit
transformation of the dependent variable should have a linear relationship.
This is not applicable in this study since there were no continuous independent
variables. All of my independent variables were categorical.
6. No multicollinearity can exist. This did not apply to my data because, in order
to have a correlation, there must be continuous variables, and my data did not
have a continuous variable.
7. No significant outliers can exist. Since all variables are categorical, no outliers
existed.
The first four assumptions were met by my choice of study design and the types of
measurements made. The main assumption of binary logistic regression analysis is the
assumption of linearity, which states for every one-unit increase in continuous
independent variables, the log odds (logit) of the dependent variable will increase by a
consistent amount (Udo & Akwukwuma, 2019). Since my data only included nominal
independent variables, the assumption of linearity does not apply to my data. The
BoxTidwell procedure is not applicable to my data since there were not any continuous
independent variables in my data. There were not any significant outliers or unusual
points in the data. There was not any multicollinearity because there were not any
significant correlations between independent variables (Udo & Akwukwuma, 2019).
Results
The results of the binary logistic regression data analysis are described in this
section. The demographic makeup of the 318 students who completed the survey is listed
below in Table 3.
Table 3
Categorical Variable Coding
Variable Frequency %
Ethnicity Asian 18 6%
Black 21 7%
Hispanic/Latino 183 58%
Native American 10 3%
White 86 27%
Gender Male 79 25%
Female 239 75%
First Generation College Student Yes 259 81%
No 59 19%
Non-traditional Student Yes 162 51%
No 156 49%
Single Parent Yes 36 11%
No 282 89%
According to the Classification Table Prediction, without considering any
independent categorical variables, 59% of the time, it can be correctly assumed students
were food insecure. There were 189 food-insecure students and 129 secure food students
in my study. The Omnibus Test of Model Coefficients shows how well the model
predicts categories when there is no independent variable (Frankfort-Nachmias &
LeonGuerrero, 2018). If the Omnibus Test results show statistical significance, the p
value will be <.05, which represents a poor-fitting model for the data. The Omnibus Test
was a good model for my data since the data was not statistically significant, as indicated
by a p value of .296. The binomial logistic regression model was not statistically
significant. x2(8) = 9.579, p =.296. The Nagelkerke R Square value explains the amount
of variability of the dependent variable, which can be explained by the model (Frankfort-
Nachmias & Leon-Guerrero, 2018). The model (Nagelkerke R2) showed 4% of the
variance in food insecurity was explained by the five independent variables, a minimal
effect size.
Of the five predictor variables, none were statistically significant, as shown in
Table 4. When there were no significant predictors, I reran the binary logistic regression
analysis using a stepwise enter of the variables to see if there were any significant
predictors. The results were unchanged. Since none of the predictor variables had
statistically significant results, the null hypothesis was accepted. There was no
relationship between the dependent variable of food insecurity and the independent
variables.
Table 4
Variables in the Equation
Variable B SE Wald df Sig. Exp(B)
Ethnicity-Asian -.375 .543 .476 1 .490 .687
Ethnicity-Black -.087 .498 .030 1 .862 .917
Hispanic/Latino -.195 .285 .469 1 .493 .823
Native American -1.037 .846 1.503 1 .220 .355
White 1.439 1.329 1.172 1 .279 4.216
Gender .151 .275 .300 1 .584 1.163
First-generation student .411 .310 1.757 1 .185 1.509
Non-traditional student .414 .241 2.948 1 .086 1.512
Single parent -.195 .401 .235 1 .628 .823
Although each of the categorical independent variables, ethnicity, gender, being a
first-generation college student, being a non-traditional student, and being a single parent,
were found to be predictors for food insecurity in other colleges, this was not true for
HCC students in the Central Valley of California (Baker-Smith et al., 2020; Broton et al.,
2018; Bruening et al., 2017; Goldrick-Rab et al., 2017; Spaid & Gillet-Karam, 2018; U.S.
GAO, 2019). Perhaps my study did not find statistical significance between food
insecurity and the independent variables because there is something unique about the
student population or environment in the Central Valley of California, such as having
only 30% of students as full-time students, having a high percentage of minority students
or because of having 81% of student respondents who were first-generation students.
Additionally, the issue of food insecurity is so complex that a profile of a food-insecure
student is difficult to determine.
Sensitivity refers to obtaining true positive results, meaning results that were
predicted correctly by the model (Frankfort-Nachmias & Leon-Guerrero, 2018). The
positive predictive value refers to the percentage of cases that were accurately predicted
by the model. The positive predictive value for my data was 100 x (17 ÷ (20+17)) =
45.9%, which means 45.9% of food insecure students were correctly predicted.
Specificity refers to obtaining true negative results, meaning results where the model
accurately predicted that the characteristics would not be present (Frankfort-Nachmias &
Leon-Guerrero, 2018). The negative predictive value refers to the percent of accurately
predicted cases that did not have the observed characteristic compared to the total cases
that were predicted not to have the characteristic. The negative predictive value for my
data was 100 x (169 ÷ (169 + 112)) = 60%, meaning that 60% of the time, the model
correctly predicted cases of students not having enough food to eat.
Summary
My data failed the non-proportional odds assumption of the ordinal logistic
regression analysis, thus disallowing the use of an ordinal logistic regression test. Since
the USDA allowed for classifying the survey results into two groups, food insecure and
food secure, a binary logistic regression test with a stepwise approach was used to
determine any relationships between food security and ethnicity, gender, being a
firstgeneration college student, being a non-traditional student, and being a single parent.
There was no statistical significance found between the independent variables of
ethnicity, gender, being a first-generation college student, being a non-traditional student,
and being a single parent, and food insecurity. Therefore, the null hypothesis was
accepted. There was no relationship between food insecurity and ethnicity, gender, being
a first-generation college student, being a non-traditional student, or being a single parent
in HCC Students. However, the study results identified 59% of the student population as
food insecure. Additionally, 75% of respondents were females, and 81% were
firstgeneration college students. Discussion, conclusions, recommendations, and social
change impacts are described in Chapter 5.
Chapter 5: Discussion, Conclusions, and Recommendations
This quantitative study was conducted to determine the profile of a food-insecure
community college student and to determine any relationships that existed between food
insecurity and ethnicity, gender, being a first-generation college student, being a
nontraditional student, and being a single parent at HCC in the Central Valley of
California. Since my data violated the proportional odds assumption in the originally
planned ordinal logistic regression analysis, the ordinal logistic regression analysis could
not be performed. The correction for a failure of the proportional odds assumption is to
perform a multiple binary logistic regression analysis. If there is a violation with the
nonproportional odds assumption, separate binary logistic regressions tests can be
performed, even though doing so is not as efficient (Bender & Grouven, 1998). The
USDA stated that their survey is reliable and valid if participants are separated into two
groups instead of three. Therefore, the options were to do multiple binary logistic
regression tests or perform a binary logistic regression. Since the USDA allows the
survey results to be categorized into two groups, I chose a binary logistic regression
analysis.
A binary logistic regression was performed to determine the effects of ethnicity,
gender, being a first-generation college student, being a non-traditional student, and being
a single parent on the likelihood the students were food secure. The null hypothesis was
accepted since there were no relationships between food insecurity and the independent
categorical variables of ethnicity, gender, being a first-generation college student, being a
non-traditional student, and being a single parent. To verify the lack of relationships, a
stepwise binary logistic regression analysis was also conducted. Ethnicity, gender, being
a first-generation college student, being a non-traditional student, and being a single
parent were not predictors of food insecurity in HCC students. Although no statistical
significance was found, it was valuable to know that 59% of HCC students were food
insecure, 81% of all students were first-generation college students, and 75% of
respondents were females. The study results will be shared with community college
administrators so they will have the information needed to develop and target support
services for those students.
Interpretation of the Findings
Out of the 318 students who responded to the survey, 59% were food insecure.
This result is consistent with other research in the field showing California Community
College students have the highest rates of food insecurity (Goldrick-Rab et al., 2017). For
example, Que and Baldridge (2019) found that 56% of community college students were
food insecure.
Of the five categorical independent variables, gender, ethnicity, first-generation
student, non-traditional student, and single parent, they were not significant predictors of
food insecurity. Some other studies on food insecurity found each of the variables to be
predictors, but none of the studies reviewed in this paper found all variables to be
predictors. Additionally, each study had a different makeup of independent variables, and
some had different definitions of some of the variables. Future studies should be
conducted on a consistent set of independent variables to strengthen the understanding of
predictor variables.
Even though there were not any significant predictors of food insecurity,
identifying a high prevalence (59%) of students who were food insecure is important to
know in order to recognize the prevalence of food insecurity and the importance of taking
action to provide support for students. Considering that Maslow’s hierarchy of needs says
people must have their physiological need of food met before they can progress to
higherlevel needs, such as achievement, having 59% of HCC students as food insecure
means they were stuck in the physiological needs level and could not progress to
academic achievement until their basic need of food is met (Kretchmar, 2019). Having so
many food-insecure students could have a significant negative impact on collegiate
student success, providing additional motivation to college leaders to identify and help
support food-insecure students.
Knowing there were so many food-insecure students in the community college in
the Central Valley should trigger additional research to determine why there was such a
high number of food-insecure students in community college students in the Central
Valley of California. Research should also be done to identify what can be done to
provide support. Addressing food insecurity could improve student success as well as
support students’ need for food.
Realizing 81% of all students were first-generation students is useful to
understand when developing resources for student support and success because
firstgeneration students may have needs beyond other students. A college education is
associated with greater financial security (College Board Communications Office, 2017).
Therefore, having such a high percentage of first-generation college students could mean
these students come from families that make less money, thus contributing to the overall
high rate of food insecurity at HCC. Further research would need to be conducted to
determine if this is valid.
With 75% of all HCC student respondents being female, and with Goldrick-Rab et
al. (2017) having 72% of their 33,000 student respondents being females, additional
research should be done to evaluate if females are more likely to respond to surveys in
general or if females are more likely to respond to surveys specifically on food insecurity,
and why.
There may have been other predictors of food insecurity in this population;
however, they were not examined. For example, income and employment status may have
been predictors of food insecurity, but income and employment status were not studied.
Other variables, such as being a Pell Grant recipient, were also not addressed but should
be in future studies. It is important to note I looked at commonly evaluated demographic
data that students do not have control over, meaning that I chose variables that are used in
many studies, and they are variables people cannot change; ethnicity, gender, being a
first-generation college student, being a non-traditional student, and being a single parent.
These variables were found to be predictors in other studies; however, not every study
found all of these variables as predictors in every study. Additionally, those studies were
conducted in multiple types and sizes of colleges throughout the U.S. This study was only
conducted on the very specific population of community college students in the Central
Valley of California. Future studies should be conducted specifically in California due to
the high cost of living in California, which could make it more likely for students to be
food insecure.
HCC students persevere and still obtain an education despite being food insecure
and despite HCC offering a food pantry and referrals for help. There may be
consequences to being food insecure that should be identified, and there may be reasons
why HCC students are more likely to be food insecure. Food insecurity is a complex
multifactorial problem that makes identifying precise predictors hard to determine.
Limitations of the Study
A standardized survey instrument was used for this study, so the survey could not
be customized for the study population. The survey was conducted at a Central Valley
Community College; therefore, the results only apply to HCC. The nature of the study
was correlational, so causation was not determined. Students were sent an email request
to complete an online survey. Some students do not read emails from the college. There is
an inherent survey completion bias in those that are willing to take their time to complete
a survey. Potentially, students impacted by food insecurity may have been more likely to
complete the survey than students who were not impacted. Having 75% of females
complete this survey could be reflective of this potential limitation. The survey was only
given in English, which could have been difficult for students for whom English is a
second language. This survey was completed during the global COVID-19 pandemic,
which could have influenced students’ level of food insecurity. The survey was conducted
on students from a community college in the Central Valley of California; therefore, the
results are only applicable and generalizable to community college students in the Central
Valley.
Recommendations
There was no advantage to using the ordinal three-group option of categorizing
food security. In fact, there was a disadvantage. By considering the USDA survey results
as an ordinal variable with three levels of groups of food insecurity, students who fall in
the middle category of low food insecurity are at a disadvantage because they would be
classified separately, therefore, may not be seen as having as much of a need as compared
to students classified as high food insecurity; therefore, students in the middle category of
low food insecurity may not get as much help as students in the higher level of food
insecurity. Students who are food insecure at any level deserve to have help. From a
social justice and public health perspective, no level of food insecurity is acceptable;
therefore, all food-insecure students should receive needed support. To ensure all
foodinsecure students are recognized as needing help, I recommend conducting future
studies using the USDA allowed option of having two groups instead of three so as to
identify food insecure versus food-secure students.
By conducting this study, I hope to have created a template that others will use in
future studies on food insecurity. Future studies can look at the same predictors in the
same location to see if the results are duplicated, and they can examine other predictors in
the same location. Additional locations should also be studied. Studying first-generation
students in greater depth can generate a greater understanding of the specific needs of this
population. I also recommend considering other independent variables, such as insurance,
employment status, income, family support, homelessness, being a Pell Grant recipient,
receiving SNAP, low income, GPA, the poverty rate of the community, cost of living, the
percentage of the population with at least a bachelor's degree, and consequences of food
insecurity. According to Goldrick-Rab (2018), community college students have a higher
need for support than other students; therefore, studying multiple areas of basic student
needs may be useful. Adding a qualitative component to the research can provide
additional information to help understand the complex issue of food insecurity and the
resulting consequences. The issue of food insecurity is complex, making a profile of a
food-insecure student hard to define. The relationship between variables should also be
studied. For example, the relationship between gender and being a single parent should be
studied to determine if there are multiple factors that determine the likelihood of being a
food-insecure student. Students’ past histories should be examined as well. Further
research into college students in the Central Valley of California should be conducted to
determine if there are additional factors specific to this population that could increase
their rate of food insecurity. The size of the college, rural versus urban settings, and
poverty rates of the surrounding population should also be studied. Lastly, additional
studies of the types mentioned above should be conducted that utilize a stratified
sampling technique in order to ensure there are equal proportions of students representing
each variable.
Implications for Social Change
Identifying that 59% of students were food insecure sent a strong message of the
need for community college students in the Central Valley to have greater access to
adequate food so their basic needs can be met and help improve student success. This
study contributed to social change by identifying the high percentage of students who do
not have enough food to eat, thus promoting the need for college leaders to develop
programs and resources to support students’ basic need of having an adequate amount of
food. Community colleges must strive to provide needed assistance to their students in
order to further student success and support students’ basic needs. From a public health
and human rights perspective, all food-insecure students need help and support, not just
those who have are severely food insecure. This is why the USDA survey results should
be considered two groups instead of three groups when further studied so as to capture all
food-insecure students. Further study should be done to identify potential causes for food
insecurity and ramifications from being food insecure. Knowing 81% of all students were
first-generation college students shows the importance of studying this specific
population to help understand what special needs they may have in order to be successful
in college. The results of this study should motivate college leaders to take action and
create support for food-insecure students and first-generation students. Some ways
students can be supported include establishing a food pantry, providing cafeteria meal
vouchers, referral to SNAP and the WIC program, providing help with applying for a Pell
Grant and scholarships, offering cooking, shopping, and budgeting classes, and providing
referrals to foodbanks.
Conclusion
Using Maslow’s Hierarchy of Needs as the theoretical framework, I conducted a
quantitative correlational study on Community College students in the Central Valley of
California. The purpose of the study was to determine the relationship between food
insecurity and ethnicity, gender, being a first-generation college student, non-traditional
student, and being a single parent. A binary logistic regression analysis was performed on
the data. Results found 59% of students were food insecure, 81% of all students were
first-generation college students, and 75% of all students were females. Knowing the
prevalence of food insecurity is essential information for college leaders to have when
developing support services and designing outreach services for students.
This study on food insecurity in community college students in the Central Valley
was helpful to determine the prevalence of food insecurity in this population and to
identify that 81% of all food-insecure students were first-generation. The categorical
independent variables were not significant predictors of food insecurity in this population,
so the null hypothesis was accepted. The non-significant findings may be a reflection of
the complex nature of determining a single profile of a food-insecure student and why a
multitude of studies have had varying results with what was significant for identifying a
food-insecure student. Knowing 59% of students were food insecure is critical
information to be aware of in order for college leaders to have the needed knowledge to
realize the importance of providing basic needs support and outreach targeted to food-
insecure students. Student success efforts should include basic needs support for food-
insecure students and first-generation college students. Having the majority of community
college students experience food insecurity is not an acceptable human rights situation
and can have long-lasting impacts that should be further studied.
Addressing basic food needs in college students not only benefits students but benefits a
college’s student success rate, as well as positively impacts communities at large.
