Grow. Prepare. Eat: Lessons Learned from a Summer Virtual Food Literacy Program

Food literacy is an important approach to nutrition education that improves diet behaviors and quality (Amin et al., 2019; Vidgen & Gallegos, 2014). The framework includes knowledge of food systems, food safety and freshness, safe gardening and cooking skills, and autonomy in meal preparation (Amin et al., 2018). To date, the most successful in-person food literacy interventions for children are multifaceted and emphasize nutrition education combined with at least one hands-on component such as culinary demonstrations and practice, recipe/food tastings, or school gardens (Allirot et al., 2016; Amin et al., 2018; Bai et al., 2014; Dudley et al., 2015; Hamner & Moore, 2020). Further, programs with the most significant results include substantial parent engagement components (Evans et al., 2006; Bai et al., 2014; Caraher et al., 2010). At the same time that food literacy is on the rise, online learning has become more prevalent in K–12 and higher education (Curtis & Werth, 2015). Current research shows that children’s virtual education must include mechanisms to keep their attention, regular and meaningful interaction with peers, frequent and direct communication with educators, and awareness of parent perspectives and concerns (Archambault et al., 2016; Borup & Graham, 2014; Borup & Stevens, 2016; Elrick et al., 2018; International Association for K–12 Online Learning (INACOL); Kaur et al., 2015; Kwon et al., 2019; Lohse et al., 2012; Lokey-Vega et al., 2018; Murimi et al., 2019). It is unclear how online learning can be utilized most effectively for public health interventions, which often necessitates hands-on and experiential learning (Amin et al., 2018; Dudley et al., 2015; Jarpe-Ratner et al., 2016; Kolb et al., 2001). Specifically, there is a gap in virtual food literacy curricula and suitable student engagement programs. Few virtual food literacy programs exist; even fewer have identified best practices to keep students engaged while providing the necessary hands-on component that appears pivotal to food literacy learning (Garcia et al., 2020). Therefore, this pilot study, informed by the Social Cognitive Theory, aimed to fill the gap in the literature by evaluating a 6-week summer virtual food literacy program on third through fifth graders’ food literacy and dietary knowledge and skills, self-efficacy, and dietary patterns. Study outcomes provided insight into best practices for virtual food literacy education programs to achieve positive changes.

Methods

Instruments

The curricula developed for this study and survey methods used were informed by a theoretical framework that explains how an individual learns, the Social Cognitive Theory (SCT). The SCT posits that learning is affected by a blend of personal, interpersonal, and environmental factors that influence behavior in a dynamic and ongoing process (Bandura, 1991). Lessons emphasized the SCT constructs of self-efficacy and experiential learning and incorporated best practices in nutrition education, such as providing a multifaceted approach to present food literacy content, educational content about all dimensions of food literacy (planning and managing food, selecting food, preparing food, and eating food) and allowed for both child and parental involvement (Allirot et al., 2016; Battjes-Fries et al., 2017; Quinn et al., 2003).The virtual food literacy lessons and post-class hands-on activities were delivered twice weekly for six weeks in the Summer of 2020.

A mixed-methods approach evaluated the program’s effectiveness, barriers, and facilitators of a virtual food literacy curriculum. Quantitative data were collected online via Qualtrics®, using two validated survey instruments: the Harvard School of Public Health Children’s Nutrition Questionnaire (FFQ) (Harvard T.H. Chan School of Public Health Nutrition Department’s File Download Site) and The Tool for Food Literacy Assessment in Children (TFLAC) (Amin et al., 2019). Surveys were completed in the pre and post-six-week program. Qualitative data included transcripts of live lessons, digital portfolio submissions, commentary on those submissions, and the open-ended questions in the expert-reviewed Qualitative Feedback Form (QFF) provided to participants at the end of the program.

The TFLAC contains 40 questions categorized into five domains of food literacy: food systems knowledge, cooking skills, cooking knowledge, nutrition knowledge, and self-efficacy regarding eating. The FFQ collected data from guardians to measure the frequency of each child participant’s consumption of foods and dietary patterns. The QFF assessed changes in participants’ culinary skills, confidence, and self-efficacy in preparing simple meals autonomously and with adult support, changes in fruit and vegetable preferences, and participant engagement/enjoyment of the lesson/activity topics. Data from the post-program were triangulated and validated. This study and all procedures used were approved by the Rutgers University Institutional Review Board under protocol Pro2020001158. Written informed consent was obtained from all parents/caregivers of subjects.

 Sample

Study participants were a convenient sample of parent-child dyads in which the children attended third through fifth grade in central New Jersey (NJ). Recruitment occurred one month before the start of the first lesson via emails to parent groups in community partner organizations and colorful flyers posted within schools. The schools used to recruit study participants were local schools with previous collaborations with the study team. Eligible families were living within NJ at the time of the study, had at least one guardian available to assist during class lessons, had access to an internet-connected device, a kitchen to prepare basic recipes, had space to grow a small plant, and were able to speak and understand English at a proficient level.

Participants were motivated to join the study by gaining food literacy, culinary skills, and nutrition knowledge. Additionally, the study provided a valuable learning opportunity for children during the summer school break and the COVID-19 pandemic. All registrants received a $50 gift card to purchase the program’s materials and food supplies. To discourage attrition, a $50 registration fee was required to register for the program but was returned to subjects who attended at least 80% of classes (live or recordings), uploaded proof of work for at least 80% of activities, and agreed to complete all surveys (Head et al., 2013, Dibsdall et al., 2003). The registration fee was waived for those who could not afford the $50 initial fee; however, no subjects requested their fee be waived.

Data Analysis

The FFQ, completed by guardians, collected data from 101 items to measure the frequency of each child’s usual food intake and dietary patterns. Frequencies of intake for a particular food ranged from zero in the last four weeks to six or more times each day. Questions regarding dietary patterns required a dichotomous yes (1) or no (0) response to consuming different types of food (milk and bread). FFQ responses for individual foods were categorized into the following groups, using USDA Dietary Guidelines for Americans 2020–2025 as a reference: fruit and vegetables, plant-based protein foods, and land animal-based protein foods (Dietary Guidelines for Americans 2020-2025, 2020). From the two dietary pattern questions examined (milk and bread type), the most favorable options, according to USDA Dietary Guidelines for Americans 2020-2025, of lower-fat milk and higher-fiber bread were assessed.

The TFLAC, completed by participants, contained 40 questions categorized into five domains of food literacy: food systems knowledge, cooking skills, cooking knowledge, nutrition knowledge, and self-efficacy regarding eating (Amin et al., 2019). Questions included knowledge-based questions with dichotomous answers and questions measuring behavior with multiple response options. A point value of one was assigned to each correct response or desired attitude or behavior to all questions on the TFLAC. The total possible score range on the TFLAC was 0 to 40, with higher numbers indicating greater food literacy. Interrater reliability was addressed by having two research assistants score each TFLAC and then rescored if discrepancies were found.

Guardians provided sociodemographic characteristics of participants via Qualtrics® before the start of the study to include age, year of school, geographical location within the state, sex, race or ethnicity, members in the household, and participation in the National School Breakfast and Lunch Program.

All quantitative analyses were conducted using the Statistical Package for Social Sciences (SPSS), version 28. Descriptive statistics (means, standard deviations, frequency, and percentages) were performed for all survey variables at baseline and the end of the program, as well as participants’ sociodemographic characteristics at baseline. Percent change in mean responses between the pre-and post-program surveys was also conducted. Due to the small sample size and the scales not being normally distributed, inferential statistics could not be completed.

Qualitative Data and Analysis

Qualitative data analysis identified facilitators and barriers to the success of a virtual food literacy program for 3–5^th graders. The qualitative data included transcripts of live lessons, digital portfolio submissions, and responses to the QFF. The QFF assessed changes in participants’ culinary skills, confidence, and self-efficacy in preparing simple meals autonomously and with adult support, fruit and vegetable preference change, and participant engagement and enjoyment of the lesson/activity topics. Data from the QFF was triangulated and validated, and inductive qualitative analysis was used where the codes emerged from the available data. Sub-themes arose within each larger theme, including Content Knowledge (Plant Science/Growing Food, Composting, and Nutrition), Facilitator and Barriers (Pedagogical Approach, Technology, and Structure), and Empowerment (Confidence and Self-Efficacy). The qualitative data analysis process to determine appropriate themes is visually summarized in Figure 1.

Results

Quantitative

Fifty-four parent-child dyads registered for the lessons; 27 participated in at least 80% of the live lessons and/or digital submissions; and 18 completed the quantitative data surveys (TFLAC & FFQ). Therefore, data reported in this study was from 18 parent-child dyads (n=18). Child participants were primarily ten years old (41%), had completed 4^th grade and were going into 5^th grade in September (48%), residents of central NJ (81%), female (67%), white (41%), and Asian (41%), and members of households with between four to six people (74%). Of the participants, 11% participated in the National School Breakfast and 19% in the National School Lunch Program. For comprehensive participant demographics, see Table 1.

FFQ mean responses indicated many favorable dietary changes pre/post-program (Table 2). Favorable dietary changes noted from the FFQ results include increased post-program fruit and vegetables, plant-based protein, low-fat milk, and high-fiber bread. High-fiber bread had the greatest change in intake, with an increase of 80%, followed by low-fat milk intake, which increased by 23%, and fruit and vegetable intake, which increased by 10%. While plant-based and land animal-based protein intake increased, plant-based protein intake increased more substantially (21% vs. 17%).

There was an increase in the total TFLAC score, approximately 3%, between pre- and post-surveys. TFLAC scores also increased among the following subcategories: food systems knowledge (5%), nutrition knowledge (4%), and self-efficacy regarding eating (5%), reflecting desired knowledge and behavior changes outlined in program goals and objectives. However, there was a slight decrease in the cooking skills and cooking knowledge questions on the TFLAC survey from pre- to post-survey (Table 3).

We hypothesize that the decrease in the cooking domains may be attributed to a few factors. There were very few questions on the TFLAC relating to these two domains, and the sample size was very small. Therefore, one incorrect question could have a large impact on the results. The Food System Knowledge and Nutrition Knowledge domains had more questions related to those areas; thus, a single incorrect response had a different impact. Without inferential statistics, we could not measure the statistical significance of these changes due to the small sample size and the scales not being normally distributed. Second, the scope and sequence of the food literacy lessons may have affected learning and behaviors, as reflected by FFQ and TFLAC survey responses. Lessons related to fruit and vegetables were presented during the first two weeks and referenced throughout the program. The early introduction of those topics and subsequent reinforcement likely led to greater fruit and vegetable intake changes than other foods assessed, such as land animal-based protein, which increased rather than decreased. Lessons related to land animal-based protein were presented later in the program. Therefore, the sequence and timing of lessons and surveys should be considered carefully in future program implementation.

Qualitative

This study’s qualitative data supports other research that finds students enjoy learning about nutrition, culinary, and food systems topics and that the inclusion of regular hands-on activities and opportunities for parental involvement add significantly to student engagement. However, the participant and parental feedback also demonstrate that executing a successful virtual food literacy program will encounter many challenges that in-person programs typically avoid. This is consistent with other findings describing the additional challenges of virtual teaching, such as supporting students fully from afar, the additional time for lesson or activity planning, and the additional training required for instructors to implement virtual teaching successfully (Farmer & West, 2019; Mandernach et al., 2013; McAllister & Graham, 2016). This research adds to the existing knowledge base by elucidating specific facilitators and barriers to virtual food literacy education, which are discussed in more detail below.

 Facilitators

The participants’ prior interest in food literacy supported high engagement and knowledge gains regarding plant science/growing food. Grow. Prepare. Eat. provided many opportunities for the participants to practice their gained skills asynchronously. As the program progressed and skill confidence increased among children, they developed self-efficacy in preparing meals with assistance from guardians. Offering multiple daily class sessions and accessibility via various online and digital platforms (Zoom, WebEx, ClassDojo, Qualtrics) was a facilitator for participation; these technologies allowed for flexible class times, interactions between teachers and students, and data anonymity. Figure 2 outlines technology-specific findings in greater detail.

Barriers

Using assigned code numbers for the anonymity of participants, instead of participant names, inhibited participants’ ability to personalize their interaction with each other or the instructor and conflicted with elementary students’ developmental stage. It also inhibited, rather than encouraged, participant-instructor trust and relationship building. Participants regularly expressed that it made them uncomfortable or sad that we could not know or share their names with their classmates. Knowing someone’s name is part of knowing the individual deeper. Allowing participants to choose their code names would allow them to highlight their personalities and individuality and offer a sense of ownership and uniqueness that only a name, not a number code, can provide (Marrun, 2018; Nieto, 2009; Peterson et al., 2015). Access to technology, acquiring program supplies, and the ability to print out surveys were other barriers that arose, and future programs should consider funding or supply delivery of these items as part of their design.

Participants’ feedback on whether the asynchronous nature of the hands-on activities and the intentional incorporation of parental participation were facilitators or barriers was mixed, depending on familial schedules and parental availability (Figure 2). Additionally, participants communicated that they desired an additional emphasis on culinary skill acquisition and choosing the difficulty of a recipe according to their skill level. Future programs might consider an initial needs assessment to address parental availability, program topics, and recipe complexity.

Conclusions & Applications

This study evaluated the effectiveness of a six-week summer virtual food literacy program on third through fifth graders’ dietary patterns, food literacy knowledge, skills, and self-efficacy. Many studies have developed and analyzed the efficacy of in-person food literacy programs for elementary students. While child-directed food literacy programs have become increasingly popular, few studies describe the development, execution, or rigorous analysis of virtual programs (Allirot et al., 2016; Amin et all, 2018; Bai et al., 2014; Battjes-Fries et al., 2017; Caraher et al., 2010; Dudley et al., 2015; Garcia et al., 2020; Jarpe-Ratner et al., 2016; Muzaffar et al., 2018; Nguyen & Murimi, 2017; Powers et al., 2005; Quinn et al., 2003) This study contributes the literature by examining virtual food literacy programs and their ability to influence on children’s dietary intakes and food literacy. It offers practical recommendations and best practices for future successful hands-on, virtual curricula.

Dietary Intake and Patterns

The observed increases in fruits and vegetables, plant-based protein, lower-fat milk, and higher-fiber bread align with the desired behavior changes outlined in the program’s goals and learning objectives. The Grow. Prepare. Eat. curriculum included educational content and experiential learning activities that encouraged healthy dietary patterns. These improved dietary patterns support the recommendations of the Dietary Guidelines for Americans and provide evidence of a connection between virtual education and dietary change, as well as the power of food literacy in affecting children’s nutrition and self-efficacy. It is important to acknowledge that while the program’s impact on children’s diet is evident, the small sample size limits the statistical robustness and generalizability of the findings. With a limited number of participants, the study lacked the statistical power to perform inference testing and our ability to further support our results.

Food Literacy Knowledge and Behaviors

The TFLAC scores in the food systems knowledge, nutrition knowledge, and self-efficacy regarding eating increased, along with the total TFLAC score, reflecting desired learning outcomes and behavior changes outlined in the curriculum and consistent with other programs that have successfully communicated food literacy messages (Bai et al., 2014; Baranowski, et al., 2003; Battes-Fries et al., 2017; Caraher et al., 2010; Evans et al., 2006; Garcia et al., 2020; Jarpe-Ratner et al., 2016; Nguyen et al., 2017; Powers et al., 2005; Quinn et al., 2003). We provided several hypotheses as to why we did not observe increases in the cooking domains of food literacy and suggest future programs collect more survey data on these factors and ask a similar number of survey questions across all domains to improve the probability of statistically significant and more accurately measure programmatic outcomes across all food literacy domains. In addition, researchers should carefully consider the sequence of lessons in the curriculum, putting the most fundamental food literacy topics first. These domains can be emphasized throughout the program and will likely yield the most significant knowledge and behavior change.

Limitations and Implications for Future Research

Study limitations include a small sample size with missing data and high attrition, which could have contributed to a decrease in TFLAC scores in specific domains, as seen in the analysis and reliability of the quantitative results. There were missing data (participants not completing both surveys) in at least 30% of all surveys completed. Although the data were recoded, and the series mean was used for greater than 30% of the missing data, the results cannot be generalized due to a large amount of missing data. Therefore, expanding recruitment and enrollment in future interventions is recommended to allow for attrition.

Another potential limitation was the length of the program. An extended program (beyond six weeks) would further reinforce critical concepts and likely lead to more significant changes in intake and behavior. Finally, participants were part of a voluntary response sample, thus skewing the results from a random sample with no previous interest in food literacy. Additionally, although the program was advertised throughout New Jersey, all participants had access to the internet, devices, and a well-stocked kitchen, and many participants had consistent adult support. Based on these sample demographics and features that often indicate affluence or abundant resources, results are less generalizable than the broader American or state population.

This pilot study’s primary strength is the vast qualitative data that inform practical recommendations and best practices for virtual food literacy education programs. Any future iterations or programs attempting virtual food literacy education could incorporate the components that facilitated participant learning (technology training, flexible class times (live or recorded), multiple online platforms, parental involvement, educational content paired with experiential learning opportunities), and eventual empowerment while removing and avoiding the pitfalls and barriers that inhibited the fullest degree of participant engagement.

Summary

The results of this pilot study further support research findings that students enjoy learning about nutrition, culinary, and food systems topics and that regular hands-on activities, frequent teacher/student interaction, and opportunities for parental involvement add to student engagement. Our findings also support the connection between food literacy education and healthful dietary patterns. However, the participant and parental feedback regarding difficulties of technology and the occasional lack of availability of a parent or guardian and sufficient time for the activities also demonstrate that the execution of a successful virtual food literacy program will encounter many challenges in-person programs typically avoid. This is consistent with other findings that describe the additional challenges of virtual teachings, such as finding ways to keep students accountable and engaged, supporting students fully from afar so that they can complete all necessary work and assignments, the additional time for lessons or activity planning, and the additional training required for instructors to implement virtual teaching successfully (Farmer & West, 2019; Mandernach et al., n.d.; McAllister & Graham, 2016). This research adds to the existing knowledge about virtual learning and food literacy and provides insight into practices that facilitate student learning. As a result, future programs will likely incorporate the facilitators addressed while avoiding the barriers discussed.

While we cannot establish causal relationships from this study, given the small number of participants and missing data, we can conclude that specific components facilitated or hindered overall participant learning, engagement, and eventual increase in empowerment and confidence in participants’ ability to develop a healthy relationship to food, as defined above. Our quantitative and qualitative results parallel the pilot study goals and outcomes: to evaluate third through fifth graders’ food literacy and dietary knowledge, behaviors, intake, and patterns to develop practical recommendations and best practices for successful hands-on, virtual curricula. This pilot study suggests that a successful virtual food literacy program for third- through fifth-grade students must include frequent and diverse hands-on activities (whether synchronous or asynchronous), regular opportunities for students to get to know their instructors and their fellow students and express individuality, and pre-program training is essential for students, parents, and instructors to maneuver chosen technological platforms with ease and comfort. This pilot study provides immediate and valuable suggestions for techniques to facilitate or hinder student learning and engagement and demonstrates promising trends in healthy food intake and knowledge among youth populations.

Acknowledgments

This research was supported by a grant from the Robert Wood Johnson Foundation (Grant ID 75084) awarded to the New Jersey Healthy Kids Initiative (NJHKI). This Rutgers-led initiative is a partnership between the Institute of Food, Nutrition, and Health and the Child Health Institute of New Jersey that focuses on improving the health of New Jersey children.

Table 1. Participant demographics

Characteristic	Percentage (n=18)
Age
8 years old	26%
9 years old	33%
10 years old	41%
Grade in school
3rd grade	22%
4th grade	30%
5th grade	48%
Region of residence
North Jersey	4%
Central Jersey	81%
South Jersey	7%
Other	7%
Gender
Male	33%
Female	67%
Race or ethnicity
White	41%
Asian	41%
Hispanic, Latino, or Spanish	19%
American Indian or Alaska Native	4%
Middle Eastern or Northern African	4%
Household size
1–3 people	11%
4–6 people	74%
7 or more people	4%
Participated in National School Breakfast Program	11%
Participated in National School Lunch Program	19%

 

 

Table 2. Children’s consumption of Dietary Guidelines food groups at pre- and post-test and % change pre-post intervention (FFQ) (n=18)

Variable	Pre	Post	Pre to Post
	Mean±SD	Mean±SD	% Mean Change
FFQ*
Fruit and Vegetable	2.01±1.35	2.22±1.33	9.46
Plant-Based Protein	1.28±0.92	1.62±1.67	20.99
Land Animal Based Protein	1.21±0.66	1.45±0.97	16.55
Lower Fat Milk	0.17±0.03	0.22±0.03	22.73
Higher Fiber Bread	0.14±0.07	0.71±0.06	80.28

*Reported means correspond to frequency of intake of each food type. Key: 0= 0 times in last 4 weeks, 1= 1-3 times in last 4 weeks, 2= 1 time each week, 3= 2-4 times each week, 4= 5-6 times each week, 5=1 time each day, 6= 2-3 times each day, 7= 4-5 times each day, 8= 6 or more times each day.

 

Table 3. TFLAC scores for food literacy domains at pre-, mid-, and post-test and % mean change post-intervention (n=18)

Variable (Possible Score Range)	Pre	Mid	Post	Pre to Post
	Mean±SD	Mean±SD	Mean±SD	% Mean Change
TFLAC*
Food Systems Knowledge (0–12)	11.12±2.00	11.73±0.90	11.75±0.46	5.36
Cooking Skills (0–3)	2.28±0.93	2.73±0.41	2.13±1.09	-7.04
Cooking Knowledge (0–6)	5.73±0.61	5.95±0.15	5.44±0.62	-5.33
Nutrition Knowledge (0–15)	14.12±0.99	14.73±0.65	14.75±0.46	4.27
Self-Efficacy Regarding Eating (0–4)	3.74±0.53	3.77±0.61	3.94±0.18	5.07
Total Score (0–40)	36.98±3.45	38.91±1.10	38.00±1.95	2.68

*Reported means corresponding to the TFLAC scores in each food literacy domain. A score of 1 was given for correct/positive responses and a score of 0 was given for incorrect/negative responses. Minimum and maximum scores for each domain are provided next to each variable.

Figure 1. Inductive qualitative data coding process

Figure 2. Qualitative data findings

Appendix

File 1. Qualitative Feedback Form (QFF)

Peer-Reviewed Post-Implementation Qualitative Survey Questions: Flesh – Kincaid Grade level 2.2

• Please answer the following questions about the 6-week online food literacy program you have completed:
  1. What was one lesson that stood out to you most? What made it stand out?
  2. What changes would you make to any of the lessons or program as a whole?
• Please answer the following questions based off what you have learned in this summer program about selecting ingredients and preparing healthy snacks and meals:
  1. Describe how you feel after preparing and cooking a nutritious snack or meal by yourself:
  2. Describe how you feel after preparing and cooking a nutritious snack with little help from an adult:
  3. What advice would you give a friend who wants to pick out healthy, fresh ingredients for a meal:

1.  Please describe a new food you tried during this program:
2. Was there anything that surprised you about anything that you tried?