Food Protection Trends - November/December 2020 - 427

FIGURE 2. USDA FNS regional offices: MARO, Mid-Atlantic; MWRO, Midwest; MPRO,
Mountain Plains; NERO, Northeast; SERO, Southeast; SWRO, Southwest; WRO, Western.

Each school district was provided with an introduction
letter, questionnaires, and business reply envelopes to
facilitate return of the questionnaires. The introduction letter,
addressed to school nutrition directors, explained the purpose
of the study and requested distribution of the questionnaire
among employees. All school nutrition employees, except the
director, were invited to participate. Following Dillman's (9)
recommendation, a follow-up postcard was sent to prompt
nonrespondents to complete the questionnaire. Postcards were
sent by email to those with email addresses and by postal mail
to those with no email address listed on the school district's
website. A second reminder postcard was sent to those with
email addresses 2 weeks after the first reminder.
Data analysis
Returned questionnaires were coded, and the data were
processed and analyzed with the Statistical Package for
Social Sciences (SPSS) version 18.0 (IBM, Armonk, NY).
Descriptive statistics were analyzed for frequencies, means,
and standard deviations. Cronbach's alpha was used to measure
internal consistency with a threshold of 0.70 (8).
For the direct measures of attitude, perceived behavioral
control, subjective norm, and behavioral intention, principal
axis factor analysis with varimax rotation was conducted on
the constructs independently. Using a minimum eigenvalue
of 1.0, one factor was extracted for each of the three measures.
Further analysis enabled comparison of absolute intenders
(those who absolutely intended to perform the behavior) to
other intenders that were not absolute. To assess differences

between absolute intenders (behavioral belief score of 7.0 on
a 7-point scale) and those who did not intend to perform the
behavior, a series of t-tests were conducted for each behavior.
RESULTS
A total of 3,850 surveys were sent to participating school
districts, and 580 surveys were returned. Of the 580 returned
surveys, 172 were excluded because of patterned responses
or missing data, resulting in 408 usable surveys, for a usable
response rate of 10.6%. According to Dillman (9), the number
of completed responses obtained (408) still achieved a 95%
confidence level with a ±5% sampling error. The response
rate is also comparable to that in previous research conducted
by Sneed and Patten (26), Sauer et al. (23), and Grisamore
and Roberts (12), who obtained response rates of 7, 9, and
14%, respectively. The representability of the findings was
improved given the standardization and homogeneity of
practices and procedures found across schools, regardless
of location or size. For example, school nutrition programs
practices were homogenous because of national compliance
with the standards published in the Child Nutrition and WIC
Reauthorization legislation (27).
Of the 163 districts that were invited to participate, 43
districts were represented in the final sample. All states that
were invited to participate were included in the final sample;
thus, all USDA FNS regions were represented. Most of the
returned surveys were from large (41%) and medium (36%)
school districts and 18% were from small districts; no completed
surveys were returned from the mega school districts.

November/December Food Protection Trends

427
Food Protection Trends - November/December 2020

Table of Contents for the Digital Edition of Food Protection Trends - November/December 2020
