School Gardens and Vegetable Consumption

Research conducted by: Michelle M. Ratcliffe, Kathleen A. Merrigan, Beatrice L. Rogers, and Jeanne P. Goldberg

Case study prepared by: Robert F. Houser and Georgette Baghdady

Overview
School garden programs are gaining popularity because of their numerous benefits for children: outdoor exercise, social skills, connecting with nature, environmental stewardship, active learning, experiential science education, higher academic achievement, and transformed attitudes and habits related to fruits and vegetables.  By integrating the regular science class with gardening activities in which students plant, nurture, harvest, prepare, and consume produce grown in the schoolyard, studies are showing that garden-based learning can improve children’s consumption of fruits and vegetables.

This study investigated the impact of participating in a school garden program on the ability to identify, willingness to taste, preference for, and consumption of vegetables.  Subjects were 320 sixth-grade students aged 11 to 13 years at two intervention schools and one control school.  At the intervention schools, garden-based learning activities were incorporated into the regular science class for a period of four months.  The control school did not include a garden program as part of its science class.  Two questionnaires – Garden Vegetable Frequency Questionnaire and taste test – assessed the outcome variables using vegetables typically grown in school gardens that were also ethnically and culturally appropriate for the study population.  The Garden Vegetable Frequency Questionnaire assessed the types of vegetables consumed the day before as well as usual consumption frequency.  The taste test involved tasting five raw vegetables (carrots, string beans, snow peas, broccoli, and Swiss chard).  Both questionnaires were administered at the outset and end of the study.  Change scores (posttest minus pretest) were compared between the garden (intervention) group and the control group.

Questions to Answer
Do hands-on school garden programs increase vegetable consumption in children?  What are some of the potential sources of bias in research studies?

Design Issues
This study used a “quasi-experimental” design, which differs from an experiment in that the students were selected and assigned to the intervention group and control group by a method other than random assignment.  With this type of design, there is a greater chance that the intervention and control groups might differ at the outset of the study in ways that could bias the results of the study.  Since the study population was middle-school students living in low-income, urban communities, the results of the study cannot be generalized to other settings.  The study did not measure the actual amounts of vegetables that students consumed, so no conclusions can be drawn about number or size of servings.

Descriptions of Variables

Variable Description
School garden program group

Garden (intervention) and Control groups: Whether or not a student experiences hands-on gardening activities at school
Consumption of vegetables at school Assessed by the taste test, it measures whether or not a student ate each of five specific vegetables at school
Consumption of vegetables at home

Assessed by the taste test, it measures whether or not a student also ate each of the five specific vegetables at home


References

Ratcliffe, M. M., Merrigan, K. A., Rogers, B. L., Goldberg, J. P. (2011). The effects of school garden experiences on middle school-aged students' knowledge, attitudes, and behaviors associated with vegetable consumption. Health Promotion Practice, 12, 36-43.

Links

Ratcliffe et al. article
The benefits of school gardens
First Lady Michelle Obama Hosts White House Garden Spring 2011 Planting


Exercises

Please read the Ratcliffe et al. article before performing the exercises.

  1. What is the independent variable?  What are its levels?

  2. Is the independent variable a between-subjects variable or a within-subjects variable?  Please explain.

  3. Why did the researchers pretest the vegetable consumption component of the Garden Vegetable Frequency Questionnaire against a 24-hour dietary recall?

  4. The authors did not report student demographics for the intervention and control groups separately.  Suppose the percentage of English-Language Learners were dramatically lower or higher in the control group than in the intervention group.  What might the effect be on the results?

  5. The dictionary definition of “attrition” is “a reduction or decrease in numbers, size, or strength.”  In research studies, the term “attrition” conventionally refers to “loss to follow-up,” which means that a subject completed the baseline assessment and actively participated in the study for a period of time, but did not complete the study because, for example, the subject dropped out or moved from the study site.
    1. How might attrition affect the power of a study?
    2. How might attrition bias the results of a study?

  6. In this study, attrition is referring to incomplete data due to a variety of reasons, not just loss to follow-up, which is an unconventional use of the term.  Based on the authors’ unconventional definition of attrition:
    1. Which group – garden or control – had greater attrition in completing both pre- and post-Garden Vegetable Frequency Questionnaires?
    2. Which group had greater attrition in completing both pre- and post-taste tests?

  7. One reason the authors gave for attrition – based on their unconventional definition of the term – was lack of parental consent.  Is lack of parental consent a form of selection bias?  Please explain.

  8. In Table 2, what does “M Change” refer to?

  9. Using the means, standard deviations, and sample sizes in Table 2 and statistical significance based on p<0.05:
    1. Conduct an independent samples t test comparing the change in consumption of vegetables at school between the garden group and the control group.
    2. Test the null hypothesis that there is no change in consumption of vegetables at school in the garden group.
    3. Test the null hypothesis that there is no change in consumption of vegetables at school in the control group.
    4. Interpret your findings and report the p values.

  10. Using the means, standard deviations, and sample sizes in Table 2 and statistical significance based on p<0.05:
    1. Conduct an independent samples t test comparing the change in consumption of vegetables at home between the garden group and the control group.
    2. Test the null hypothesis that there is no change in consumption of vegetables at home in the garden group.
    3. Test the null hypothesis that there is no change in consumption of vegetables at home in the control group.
    4. Interpret your findings and report the p values.

  11. In a hypothetical program designed to reduce saturated fat intake of children by encouraging them to eat more vegetables, the saturated fat intake in both the intervention group and the control group increased.  However, the saturated fat intake in the intervention group did not increase as much as it did in the control group.  Was the program a failure?  Why or why not?