Recipient Organization
RUTGERS, THE STATE UNIVERSITY OF NEW JERSEY
3 RUTGERS PLZA
NEW BRUNSWICK,NJ 08901-8559
Performing Department
Nutritional Sciences
Non Technical Summary
For almost 50 years, The Expanded Food and Nutrition Education Program (EFNEP) has worked to improve the overall health and well-being among limited-resource population. NC3169, a multi-state research project, developed a new evaluation tool for EFNEP to update and improve its impact assessment. However, there is a need for EFNEP to also re-examine its "approach" to evaluating program impacts. The major focus of this project is to examine the feasibility of a new evaluation design (the retrospective-pretest) to measure program effectiveness and to compare it with the current design (the traditional pretest-posttest) among limited-resource population in nutrition education setting. If this new design is comparable to the assessment methods currently used (i.e., the traditional pretest-posttest method), it may increase EFNEP educational time and reduce labor costs.
Animal Health Component
100%
Research Effort Categories
Basic
0%
Applied
100%
Developmental
0%
Goals / Objectives
To complete and expand testing of newly developed EFNEP Food and Physical Activity Questionnaire (FPAQ) a. Complete further reliability testing b. Conduct Spanish reliability and validity testing c. Develop Retrospective pre/post questionnaire and examine feasibility of design on EFNEP impact reporting. d. Complete validity and reliability testing of infant feeding questions
Project Methods
This project is designed to address sections a and c of the second NC3169 goal.In year 1, while similar activities are simultaneously taking place at other stations for the other question sets (e.g., food safety, shopping, dietary intake, etc.), sensitivity to change assessment will be completed on the survey's PA questions to establish the FPAQ' s reliability.To test the sensitivity of the PAquestions in the FPAQ, a quasi-experimental study will be implemented. A convenience sample of NJ EFNEP participants will be recruited at their first classes, as the experiment group. During the same period, another group of low-income parents from the same geographic areas will be recruited for the comparison group. To the best of the researchers' abilities, the groups will be matched for age, gender, and educational levels. To be eligible for study inclusion, participants must: (1) be low-income (income<= 185% of the current Poverty Guidelines); (2) have at least one child less than 19 years old; (3) be at least 18 years old; (4) be able to read, write, and speak English; (5) have not previously attended EFNEP classes.At baseline (prior to the first EFNEP class attended), both the experiment and comparison groups will be asked to wear the accelerometers (GT3X model, ActiGraph, LLC, Pensacola, Florida) for 7 consecutive days. Participants from both groups will receive instructions on wearing the device above the right hip for >10 hours/day and removing the device during water-based activities (e.g., swimming, showering). Seven days later, they will return the accelerometers to the researchers and completed the 3 PA questions in the FPAQ to self-report their PA levels in the past week. The experimental group will then receive a 6-10-week EFNEP nutrition education as the intervention while the comparison group will receive no education.For post-testing, one week before the end of the intervention, both the experiment and comparison group will be asked to wear the accelerometers again for 7 days, following the same protocol described above. After 7 days, the 3 PA questions will be readministered and the accelerometers returned.Participants' whose data were not recorded for a minimum of 10 hours for 7 days will be excluded.The mean changes measured by FPAQ will be calculated by subtracting the baseline responses from FPAQ from the post-intervention measures, as well as the mean changes measured by accelerometer (within-group comparison). Spearman correlations coefficients and will be used to examine the changes measured by the FPAQ and the accelerometers. Meanwhile, standardized response mean,the response index, and the Guyatt's responsiveness statisticwill be calculated as another way to reflect sensitivity to change, as mentioned in the literature.2 Both t-tests and effect size will be used to examine the mean changes in PA measured by the FPAQ and the accelerometers between the experiment and comparison group(between-group comparison).The responsiveness/sensitivity to change of the 20-item FPAQ will also be assessed using nationalpretest and posttest data for fiscal year 2018. Shapiro-Wilk tests will be used to test the normality of the 20-items from FPAQ at both pre-test and post-test. Paired t-tests or nonparametric Wilcoxon signed-rank test will be used to examine if there are significant differences between pre-test and post-test, as well as standardized response mean.Aim 2, to identify the existence of response shift bias (i.e., a change in the way respondents interpret and respond to a question pre-and post- program evaluation, due to a change in their understanding of the topic).12,13 in different domains of FPAQ, will be addressed in years 2-3. A convenience sample of EFNEP participants will be recruited from at least four states. Equal representation from different regions, as well as different races/ethnicities, will be addressed. The eligibility criteria for inclusion will be the same as Aim 1.Participants will be recruited at the beginning of the EFNEP class series and asked to respond to the 20-item FPAQ (pretest). In a small sub-sample objective measures for PA will be incorporated to validate participants' responses. All participants will then receive a 6-10 week EFNEP program and complete the FPAQ at the conclusion the program (posttest). Participants will also respond to a retrospective version of the FPAQ (i.e., think back and report on their behaviors at the beginning at the EFNEP program) at the same time, and later to discuss any differences in the responses they gave between their pre-test responses, and their "think back" responses on their retrospective pre-tests. While this protocol increases the response burden for these participants, it is the protocol that is employed in this type of work, and participants receive subject payment for participation, which compensates them for the extra time spent completing survey measurees.Mean differencesfor each item from the FPAQ will be calculated for each test combination, including traditional pretest and retrospective pretest, traditional pretest and posttest, and retrospective pretest and posttest. Analysis of variance (ANOVA) will be used to assess the mean differences for each item in FPAQ at pretest, posttest, and retrospective pretest. This will allow for the comparison of change scores from pretest to posttest scores and change scores from retrospective pretest to posttest scores. Meanwhile, a significant difference between pretest and retrospective pretest indicate the existence of response-shift bias.Project Aims 3 and 4 aims will be addressed during years 3-5, likely at the same time, using the same EFNEP classes. Aim 3 will assess if measures of program effectiveness differ when traditional pre-post assessment protocol is used vs. if either a combined retrospective pretest with the posttest or a separate retrospective pretest from the posttest is used, and if so, which is superior. Project Aim 4 will assess if class educational time is increased and/or attrition reduced, according to which method is used.To accomplish Aim 3, a quasi-experiment design will be employed using a convenience sample of 3 EFNEP groups. Results from a comparison group (traditional pre-post design) will be compared to results from two experimental groups. One experimental group will employ "the Post & Then" design, in which the retrospective pretest questions are adjacent to the posttest questions in the same form. The other experimental group will employ the "Post-Then" design, in which 2 separate forms, one that addresses pre-intervention behaviors, and the other that addresses post-intervention behaviors. There will be approximately the same number of participants per group. Participants recruited from the same place will be assigned to the same group to avoid cross-talk. Aim 4 will be addressed at the same time, by evaluating the average length of time spent in education per class and class attrition rates for all three conditions.ANOVAwill be used to assess the differences among the two experiment groups (retrospective pretest) and the comparison group (traditional pretest-posttest)and independent t-test will be used to examine the mean differences between the two experimental groups for all variables.