Food Insecurity and its Association with Testing

Food insecurity is a major concern across the educational population. Even with several federal programs combating the issue, there are still students in K-12 schools as well as at college and university who do not have access to the nutrition that they need on a daily basis. This lack of necessary food can lead to detrimental educational effects due to the increase in behavioral issues and the loss of academic achievement. These harmful effects can lead students to underperform in school. In the last several years, a new federal program, the Community Eligibility Provision (CEP), has been implemented in schools in order to support the growing student population that requires assistance meeting their nutritional needs. The focus of this thesis was to examine the CEP effects on public, private, and charter schools that have adopted it in one state in the Northeast. Several t-tests were conducted to determine the association between students’ academic achievement on standardized tests before and after the program was implemented. Additional t-tests were additionally conducted to observe if the difference of academic achievement on the standardized tests was similar to those schools who were either eligible for the CEP and did not enroll and to those schools who are ineligible to enroll in the program. Findings from this study identify a statistically significant association between the CEP and academic achievement in mathematics and add to the current literature in the

iii

ACKNOWLEDGMENTS
The idea, research, and preparation of this thesis would not have been possible without the enormous support from the University of Rhode Island's School of Education faculty and staff.
To Dr. Diane Kern, who was the first person to greet me onto this amazing campus, thank you for your unyielding support. You have encouraged me to keep going, through the highest peaks and deepest trenches, and have helped me get to where I am today. Though this may mark another landmark event in my life, I still have a long way to go before I can begin to show the same level of care and compassion that you have for your students.
To Dr. Jay Fogelman, for the important talks that we have had about what it means to be a teacher. You have challenged me to think deeply about every action that I make with students. Thank you for every question that you had about my thesis, for each one brought me closer to something great. I hope after I graduate you find someone who you can compare Gantt charts with.
To Dr. Shane Tutwiler, thank you for your help with my data. You helped me more than you needed to, and I am endlessly grateful. Thank you for helping make sense of the numbers. I now feel like only a few things you say go over my head.
To Dr. Kathleen Gorman, for greatly helping me with my thesis. Without your help, I would have never learned about the Community Eligibility Provision that is the heart of my research. Thank you for your knowledge and assistance, for without both I would never have gotten this far.
iv To Nicole, who always seems to have her door open when I need it. Thank you for your never ending help even though you had your own chapters to write and your own papers to grade. The level of dedication to your work is astounding, and I only hope that I someday have that same drive that you have. Thank you for your guidance through these last two years.
To Janet, Lisa, and Marsha who have been on my side since the beginning. All of you have been the most important support for me during this process. Thank you for the laughter, kindness, and encouragement. You all saw something special in me, and encouraged others to give me a chance because of it. Thank you all for your unyielding strength and support.
To Dr. Diana Marshall and Josh Smith, for giving me the opportunities to show my dedication to my work. Thank you both for your generosity and kindness. I will always appreciate how you both have helped me.
To my friends who have kept me positive and driven towards my goals. To Alex and Kelsea for keeping me sane the first year of this program. Both of you are amazing teachers and I cannot wait to join your ranks as one. To Annie and Shay for the long lasting friendships, laughter, and talks about changing the world for the better. To Marissa and Liz for the uncontrollable bouts of laughter. Lastly, to Nicole who has shown me the endless amount of love pouring from her heart, and how to actively serve the communities I will be working in. Thank you all.                                    x American households, and low income households under the federal poverty line (Maroto, Snelling, & Linck, 2014;Roselle & Connery, 2016). Students who are faced with either intermittent or prolonged food insecurity can have a wide array of symptoms associated including: irritability, trouble concentrating, lower energy, and higher risk of illness (Roselle & Connery, 2016). Additionally, individuals may experience difficulty cooperating with others and higher probability of self-isolation (Alaimo, 2001). Many studies have examined the effects of food insecurity on cognitive development in younger students (Adrouge & Orlicki, 2013;Alaimo, 2001) as well as the effects on college-age students (Bronton & Goldrick-Rab, 2016;Maroto, Snelling, & Linck, 2014). However, this does not address the levels of secondary education that are meant to prepare students for college. Research addresses that "a child who is hungry struggles to learn, and therefore, it is imperative that educators understand the impact of hunger" (Spies, Morgan, & Matsuura, 2014). This study sought to understand one impact of hunger on students: how it affects their academic achievement.
With the likelihood of being food insecure higher among lower SES students and families (Grutzmacher & Gross, 2011;Maroto, Snelling & Linck, 2014;Roselle & Connery, 2016), one potential path to a food secure future is through a college degree.
However, more research is needed on the degree to which food insecurity hinder students preparing for college in their secondary education, primarily through its effect on standardized tests such as the Partnership for Assessment of Readiness for College and Careers (PARCC) test. Recently, the federal Community Eligibility Provision (CEP) allows schools to serve free breakfast and lunch to all students without parents being required to complete applications, a yearly requirement for free and reduced priced lunches (U.S. Department of Agriculture, 2017). This is important because families may not enroll in these programs due to fear, or be ineligible for these programs due to income levels (Fram, 2014;Gunderson, 2015) There is little research as to how this program can support students in their educational pursuits. As a student who struggled with food security both before and during college and understands to an extent the effect it can have on students, I am interested in understanding the implication of food insecurity as one prepares for college or a career. The purpose of this study was to identify the association between food insecurity and performance on standardized tests and to identify the effects of a specific mitigation program, the Community Eligibility Program (CEP), on students' academic achievement.

Effects of Food Insecurity in K-12 Education
Multiple programs are used to support students with their access to food. The School Lunch Program and the School Breakfast Program are implemented in more than 100,000 schools, reaching 31 million students, and providing free or reduced breakfast and lunch to more than 17 million students (Gunderson, 2015). These programs are assisted by the federal Supplemental Nutrition Assistance Program (SNAP), serving 47 million people with 80 billion dollars, to support families outside of school (Gunderson, 2015). However, these supports are not always adequate for students with prolonged or severe food insecurity; a report from the Urban Institute says that, "wages and benefits together are often insufficient to pay rent, utilities, transportation, and food expenses for a given month, particularly for large families" so students may go without food further into the month when aid runs out (Popkin, Scott, & Galvez, 2016). Breakfast is the meal that is most frequently missed by students with the greatest food insecurity (Grutzmacher & Gross, 2011). Furthermore, one in four children in food-insecure households were ineligible to enroll in food programs because their family income to high (Gunderson, 2015).
Food insecurity can adversely affect student behavior. The absence of necessary food can affect students emotionally, such as becoming more aggressive or depressed (Popkin, Scott, & Galvez, 2016;Roselle, 2016). As the full extent of hunger hits, "students may be more irritable, have difficulty concentrating, have lower energy levels, and get sick more often [...] in addition, children who experience food insecurity may be at higher risk for truancy, behavioral issues, and social difficulties" (Roselle & Connery, 2016). Students as young as seven or eight may even take measures to feed themselves or their families by failing school in order to attend summer classes, stealing, choosing to go to jail, or selling sexual favors (Popkin, Scott, & Galvez, 2016).
Students who are food insecure may also be facing additional hardships that compound their hunger. "Food insecurity is the most frequently reported kind of material hardship and one that often signals the presence of many others, including housing instability, foregone medical care, and loss of essential services like water and heat" (Popkin, Scott, & Galvez, 2016 Students' academic performance can also be affected by their food security status. It has been found that there are "small but significant benefits of food supplementation [can help students] in cognition, academic achievement, and school absence" (Alaimo, 2001, p. 48). The size of this academic impact can fluctuate based on the food's micronutrient content (Adrouge & Orlicki, 2013). As a result of this research, the Healthy, Hunger-Free Kids Act of 2010 (HHKA) changed the requirements for meals served by the National School Lunch Program to provide more balanced meals for students in U.S. schools (Cornish, Askelson, & Golembiewski, 2015).
Students in K-12 education can academically suffer because of food insecurity.
Elementary students who are food insecure, based on socioeconomic status, score lower in mathematics, and as a result, score 16 percent lower on average than those of their food secure peers (Adrouge & Orlicki, 2013). Alaimo (2001) analyzed scores on the Wide Range Achievement Test and Wechsler Intelligence Scale for Children tests, standardized intelligence tests used for both younger and teenage students (Alaimo, 2001). They found that scores were approximately "1.3 to 2.5 points lower (out of a scale of 20) for food insufficient children than for food sufficient children" (Alaimo, 2001, p. 45). In addition, students who had food at school, either through a school or the parent-teacher partnership, had a positive and statistically significant effect on both English and mathematics test scores (Adrouge & Orlicki, 2013). Additionally, there is a strong correlation between food insecurity and students' socioeconomic status; increasing SES by one percentile can increase the math and language test scores by a significant amount (Adrouge & Orlicki, 2013).
As explained above, previous studies have conducted research into food insecurity and its association on different aged students. Existing literature discusses the physical and emotional effects of food insecurity on secondary students (Popkin, 2016;Roselle, 2016). Additionally, studies have explored food insecurity on elementary students standardized tests and on college student's GPA. Research addresses that "a child who is hungry struggles to learn, and therefore, it is imperative that educators understand the impact of hunger" (Spies, Morgan, & Matsuura, 2014).
secondary students academic achievement.

CEP
The CEP (CEP) is a program for schools that serves low income students. This provision allows schools with a population of 40% or more students eligible for free or reduced price lunch to serve free breakfast and lunch to all students without parents being required to complete applications, a yearly requirement for free and reduced The CEP (CEP) is important for students because not all are eligible for free and reduced lunch because "one in four children in food-insecure households were ineligible for any type of food assistance because their income was too high" (Gunderson, 2015). This program allows for all students to receive a breakfast and lunch the CEP and academic achievement as defined by the PARCC standardized test, an area of exploration that will contribute to and expand upon existing research.
The following research questions guide the study:

Participants
In order to determine the relationship between food insecurity and secondary students' standardized test scores, this study analyzed data from each high school enrolled in the CEP. Data was comprised of secondary, de-identified, aggregate achievement scores from the PARCC test including mean scaled scores of the schools and the percent of students that were proficient on the tests. The study was submitted to Institutional Review Board as analysis of secondary data (See Figure 1), since the data was pre-existing, de-identified data. The Institutional Review Board determined this study does not involve human subjects. All protocols were followed. No additional information was requested from the PARCC assessment.

Setting
This study was conducted using aggregated, school-level information from the 9th grade populations of ten different schools across the state of Rhode Island that are enrolled in the Community Eligibility Program as well as six additional schools across Rhode Island that were not enrolled in the CEP: three that were eligible to enroll and three that are ineligible to enroll. This information was taken from the Rhode Island Department of Education website to determine schools' eligibility for involvement in this study. This study was completed by collecting PARCC data of selected schools from the Rhode Island Department of Education (RIDE). This data was stored on the researcher's password-protected computer.

Variables
The dependent variables are aggregate 9 th grade student PARCC scores in English Language Arts /literacy and mathematics scores in both algebra and geometry.
Aggregate scores will be collected from three different time periods aggregate PARCC scores will include mean scaled scores of the students who took the tests and the percent number of students who were proficient for each of the three tests.
There are several confounding variables that will need to be noted as limitations when analyzing the data. Students' base cognition, motivation to complete the test, interest in their education, current socioeconomic status, and interest in continuing their education may all be variables that cannot be accounted for in the scope of this study. These variables may differ in each school due to differences in policy, resources, individual teachers and overall faculty support. Because of CEP basic application requirements, there is limited publicly available information about the differences between schools that could better inform the study. Additionally, because of the nature of the CEP, only schools that have a higher number of lowincome students were examined. While there may be students experiencing food insecurity in more affluent areas, due to the focus of this study in identifying differences in academic achievement within schools enrolled in the CEP, they are not within the scope of this study. Finally, the relatively limited sample size of this study limits the statistical power of any hypothesis tests conducted.

Instruments
The PARCC test is administered to all 9 th graders enrolled in Rhode Island that food insecurity influenced PARCC scores. Additional t-tests will be conducted to identify if the difference in academic scores can be found in other schools across the state as well, either in schools that were eligible for the CEP or schools that were ineligible for the program. It is suggested that further research be conducted as to additional variables that can also associate food insecurity and enrollment in CEP.
Recommendations for further studies can be found in the conclusion.
Importance and Potential Significance of the Study insecurity and the scores on the PARCC. If the results of the study indicate that there is a significant relationship between food insecurity and standardized test scores, it may be important in the context of school funding, enrollment into the CEP, and student education. Students, their families, and school districts may benefit from this study by being enrolled in the CEP in order to make sure that no child is hungry.
Additionally, state or regional policy on food and nutrition may be affected in order to better serve students in their pursuit of a valuable education. and non-parametric methods (Wilcoxen-Rank). The repeated-measures parametric was used because the analyses were comparing different years of the same school-level data, while non-parametric methods were conducted due to the data size. Tests of normality were used to identify if the t-tests were reliable to use. Two analyses, the percent proficient of the ELA test from 2015 to 2017 (See Table 1) and the mean scaled score of the geometry test from 2015 to 2017 (See Table 2) are suggested to have violated the assumption of normality.

Is there an association between school's enrollment in the CEP and student's average achievement on a standardized test?
Several t-tests were conducted to determine whether there is any associations between the adoption of the CEP and academic scores from the PARCC assessment.
Tests were run for each part of the PARCC assessment analyzed (9th grade ELA, algebra 1, and geometry1). The t-tests identified that there was a statistically significant relationship between the CEP and the Geometryassessment and a marginally significant association between the CEP and the Algebra test. The other ttests showed little to no association between the program and the assessments.

and 2016 Comparison
The number of 9 th grade students in 2015 who completed the Algebra test in CEP schools was over 1,300 and increased to more than 1,500 in 2016 (See Table 1).
The 2015 mean number of students in the schools enrolled in the CEP who were proficient on Algebra test scores was 9.2, and increased to 12.11 in 2016 while the mean scores for students who took this assessment remain as 711 throughout both years of taking the tests, with the 2016 score slightly increasing from 711.44 to 711.79 (See Table 2). The percent proficient was marginally significant between the two years (p = 0.063) while the mean scaled score was not significant (p = 0.844) (See Table 3).
Overall eight of the ten CEP schools increased the percent of students that were proficient on their tests and six schools increased their mean scaled scores on the test (See Table 5).
The number of students in 2015 who completed the geometry test in CEP schools was over 1,300 and increased to more than 1,500 in 2016 (See Table 6). The  Table 7). The percent of students proficient was not significant (p = 0.194) whereas the mean scaled score for students was statistically significant (p = .008) (See Table 8). The percent of students proficient on the geometry test increased at six of the schools eligible for the CEP, while two of the three remaining schools' mean scaled scores decreased and one remained the same (See Table 10). During the same time, eight of the ten CEP schools mean scaled scores increased while two decreased (See Table 10).
The number of students in 2015 who completed the ELA test in CEP schools was over 1,200 and increased to more than 1,400 in 2016 (See Table 11). The ELA percent of students that were proficient from schools in 2015 was 18.5, which decreased to 18.3 in 2016 (See Table 12). Similarly, the mean scaled score on the ELA assessment dropped from 716.4 in 2015 to 714.7 in 2016 (See Table 12). Neither the difference in proficiency (p = .995) nor in mean scaled score (p = 0.690) were significant (See Table 13). Overall five of the ten CEP schools increased the percent of students that were proficient on their tests and five schools increased their mean scaled scores on the test (See Table 15).

and 2017 Comparison
The number of students who took the Algebra PARCC test in CEP schools continued to slightly increase in 2017 to almost 1,600 students (See Table 1). The percentage of students in the schools enrolled in the CEP who were proficient on Algebra test in 2015 was 9.2, and to decreases in 2017 to 8.65 (See Table 16). The mean scaled scores for students who took this assessment remain as 711 throughout all three years of taking the tests, with the 2017 score slightly decreasing from 711.44 to 711.58 (See Table 16). Neither the percent proficient (p = .178) nor the mean scaled score (p = 0.684) were significant (See Table 17). Overall, only two CEP schools had a higher percent of students who were proficient on the Algebra test than in 2015, and only three schools had a higher mean scaled score than in 2015 (See Table 5).
The number of students who took the GeometryPARCC test in 2017 decreases from the previous two years from more than 1,000 students to less than 300 (See Table   6). The percent of students proficient in Geometryfor schools in 2015 was 4.41 and increases in 2017 to 12.03 while the mean scaled score for students during 2015 was 707.57, increasing to 718.08 in 2017 (See Table 19) Neither the percent of students who are proficient on the test (p = 0.496) nor the mean scaled score (p=0.345) were significant (See Table 20). Overall, three of the CEP schools increased their percent of students proficient on the test and their mean scaled scores, three decreased these scores, and four did not have data available (See Table 10).
The number of students who took the ELA PARCC test in 2017 increases from the previous two years from more than 1,500 students (See Table 11). The ELA percent proficient from schools was 18.5 in 2015 and decreases in 2017 to 14 (See Table 22). Similarly, the mean scaled score on the ELA assessment dropped from 716.4 in 2015 to 712.4 in 2017 (See Table 22). Neither the percent proficient (p = .223) or the mean scaled scores (p = 0.292) analyzed were significant (See Table 23).
Overall, five of the CEP schools increased their percent of students proficient on the ELA test and four increased their mean scaled score (See Table 15).
Have schools that are not enrolled in the CEP seen differences in their academic achievement?
Additional t-tests were conducted on three schools that were eligible to enroll in the CEP and three schools that were did not meet the requirements in order to identify if there was a similar difference to the test scores as those of CEP schools.
Tests were run for each part of the PARCC assessment analyzed (9th grade ELA, Algebra 1, and Geometry1) comparing 2015 school year to the 2016 and the 2017 school year. The analysis of these six other schools showed little to no difference in percent of students proficient or mean scaled scores during the three-year time span.

and 2016 Comparison
In CEP eligible schools, the mean percent proficient on the Algebra test rose from 5.85 in 2015 to 7.21 in 2016 and during the same time frame the mean scaled score increased slightly from 708.41 to 708.56 (See Table 25). In comparison, the mean percent proficient in schools' ineligible to enroll in the CEP rose from 38.1 in 2015 to 48.9 in 2016 and the mean scaled score increased from 742.3 to 744.3 (See Table 26). Overall the CEP eligible schools decreased their percent proficient and slightly increased their mean scaled scores (See Table 27). In the ineligible schools both percent proficient and mean scaled scores increased (See Table 28).
In CEP eligible schools, the mean percent proficient on the geometry test rose from 4.76 in 2015 to 6.77 in 2016 and during the same time frame the mean scaled score increased slightly from 713.88to 713.97 (See Table 29). The mean percent proficient in school's ineligible to enroll in the CEP fell from 49.5 in 2015 to 48.9 in 2016 and the mean scaled score decreased from 749.2to 748.2 (See Table 30). Overall the CEP eligible schools increased both their percent proficient and mean scaled scores (See Table 31). Overall, the ineligible schools saw a decrease in both percent proficient and mean scaled scores (See Table 32).
The schools that were eligible for the CEP increased their percent of students proficient on the ELA test from 19.1 to 21.7 and increased the mean scaled score from 714.7 to 719 (See Table 33). The schools that were not eligible for the CEP increased their percent of students proficient from 65.5 to 68.4 and increased the mean scaled score from 760.9 to 764.3 (See Table 34). Two of the schools eligible yet not enrolled for the CEP showed a decrease in the number of percent proficient students from 2015 to 2017 (See Table 35) and two of the schools that were not eligible for the CEP slightly increased their percent proficiency (See Table 36).

and 2017 Comparison
In CEP eligible schools, the mean percent proficient on the algeba fell from 5.85 in 2015 to 4.63 in 2016 and during the same time frame the mean scaled score also fell from 708.41 to 708.10 (See Table 37). In comparison, the mean percent proficient in schools ineligible to enroll in the CEP rose from 38.1 in 2015 to 41.9 in 2016 and the mean scaled score decreased from 742.3 to 741.2 (See Table 38). Overall the CEP eligible schools decreased their percent proficient and slightly increased their mean scaled scores (See Table 27). In the ineligible schools, both percent proficient and mean scaled scores increased (See Table 28).
In CEP eligible schools, the mean percent proficient on the geometry test from 4.76 in 2015 to 24.2 in 2017 and during the same time frame the mean scaled score increased slightly from 713.88 to 729.47 (See Table 39). The mean percent proficient in schools ineligible to enroll in the CEP increased from 49.5 in 2015 to 81.6 in 2017 CEP eligible schools increased both their percent proficient and mean scaled scores (See Table 31). and the ineligible schools saw an increase in both percent proficient and mean scaled scores (See Table 32).
The schools that were eligible for the CEP decreased their percent of students proficient on the ELA test from 19.1 to 15.6 and increased the mean scaled score from 714.7 to 716.8 (See Table 41). The schools that were not eligible for the CEP increased their percent of students proficient from 65.5 to 69.5 and increased the mean scaled score from 760.9 to 762.4 (See Table 42). Two of the schools eligible yet not enrolled for the CEP showed a decrease in the number of percent proficient students and the mean scaled scores from 2015 to 2017 (See Table 35) and the schools that were not eligible for the CEP slightly increase overall (See Table 36).

CONCLUSION
This study examined associations between school participation in the CEP and academic achievement as determined by the PARCC assessment.
While there are many studies addressing food insecurity during student's primary education or during their college careers, more research should be done during the important period of student's secondary education (Adrouge & Orlicki, 2013;Alaimo, 2001). Research shows that there is a difference between food secure and food insecure primary students' standardized tests scores (Adrouge & Orlicki, 2013;Alaimo, 2001) as well as college students' GPAs (Maroto, Snelling & Linck, 2014 Is there an association between school's enrollment in the CEP and student's average achievement on a standardized test?

21
One t-test identified that there was a statistically significant connection between the CEP and the GeometryPARCC mean scores of the schools that enrolled in the program (p = .008). Another test also identified that there was a marginally significant association between the CEP and the Algebra PARCC This finding supports the previous research that food security can increase students' academic achievement. The findings in this research show that, overall, the schools that enrolled in the CEP had an increase in their percent of students proficient on mathematics tests and mean scaled scores on these tests.
Have schools that are not enrolled in the CEP seen differences in their academic achievement?
Additional analyses were conducted on data from three schools that were eligible to but did not choose to enroll in the CEP and three schools that were ineligible in order to identify if there was a similar difference to the test scores as those of CEP schools. The analysis of these six other schools showed little to no difference in percent of students proficient or mean scaled scores during the three-year time span. The data identified there were no significant difference in academic achievement on any of the PARCC tests as compared to the 2015 data. One area to note is the large difference between the eligible and ineligible schools in scores.
Schools that were ineligible for the CEP has both higher percent of students proficient and mean scaled scores, ranging between 10 to 60 percent than their eligible counterparts.

22
There were a limited number of schools that could be tested due to the low can also deal with "housing instability, foregone medical care, and loss of essential services like water and heat " (Popkin, Scott, & Galvez, 2016).

Importance of Research and Future Studies
Additional research can be derived from this study into secondary students' food insecurity. Educators and supplemental nutrition program personnel may be interested in this proposed study because it could possibly identify additional areas of support or a higher need for supplemental programming. These findings will be shared, firstly, with the school and districts in the study to assist them in procuring additional resources for their students. The findings will also be shared with programs 23 such as the SNAP, School Lunch Program, and School Breakfast Program in order to support their requests for additional funding or expansions of their programs. Colleges and universities may also be interested in this research because of its implications in their admissions process and as criteria. Further, there may be other local or state organizations that may benefit from this research as well because they also support families' and students' food needs. The study's findings could be published in either an educational or nutritional scholarly publication.
There is an opportunity to complete another study using the new RICAS