Food Protection Trends - September/October 2020 - 298

FIGURE 1. Conceptual research model using the social cognitive theory.

yielding 555 independent ethnic restaurants. Only Chinese
and Mexican restaurants were selected, because they are
among the most popular ethnic restaurants in the U.S.
(3, 45). There were 119 independent Chinese restaurants
and 176 independent Mexican restaurants in the refined
list. Restaurants were at first identified as either Chinese
or Mexican by name. Further cross-checking was done
on the website of each restaurant to ensure that it was an
independent Chinese or Mexican restaurant. To ensure
consistency, Chinese or Mexican restaurants that operate
in hotels, grocery stores, or clubs were excluded. Any
independent Chinese or Mexican restaurant from which
employees were selected for the focus group interviews or the
pilot study were removed from the database prior to selecting
the final sample. The sample estimated to achieve a 95%
confidence interval based on the number of constructs under
investigation was 138 food handlers, according to results
obtained by use of G*Power (Version 3); however, in order to
check for the measurement scales' validity, 250 food handlers
were targeted.
Survey design
It has been recommended that a focus group have six to
eight participants, but not more than ten (40, 55). For the
purpose of designing the questionnaire, three focus group
interviews with food handlers from independent Chinese
and Mexican restaurants were planned, with the goal of
having a minimum of six participants in each group. Because
of employees' time constraints and difficulty of recruitment,
conducting three focus groups was not feasible. Therefore,
one focus group interview was conducted with seven food
handlers from an independent Mexican restaurant. In
addition, four focus group interviews, each with two or three
participants, were conducted, for a total of ten food handlers
from independent Mexican and Chinese restaurants. All
interviews followed a questioning route with open-ended
questions and other probe questions if required. The

298

Food Protection Trends September/October

questioning route was prepared by the researcher based on
previous research (1, 12, 19, 37, 53, 59, 73)
All interviews were audiorecorded, transcribed, and
coded and with use of a thematic codes list that included
both pre-established and free codes. To ensure reliability and intercoder agreement (20), another researcher
was asked to independently transcribe and code the
recordings. Coding themes were then examined, and any
disagreement was resolved. The coded data was analyzed
using the procedures of NVivo 12 Plus for Windows (Version 12). Analysis of the interviews' responses showed
nine main themes: self-efficacy, self-regulation, environmental factors, outcome expectations, behavioral intentions, food safety behaviors, cultural background, attitude,
and knowledge.
The results of the interviews were used to generate an
item pool for the measurement scales in the questionnaire.
The most frequently discussed thematic items identified
from the transcripts, along with other items adapted
from the literature, were developed into 55 statements
in a Likert format on a scale of 1 to 5. For example, selfefficacy included ten items to measure the participants'
level of confidence about their ability to perform the three
investigated behaviors. Self-efficacy items were phrased
using language similar to that used by the interviewees,
such as the scale item "constantly prepare food in a sanitary
manner, even when I am busy with other tasks." Similarly, the
sub-themes of the environmental factors were used to build
up the items in the environmental determinants scale. For
instance, the sub-theme time constraints was developed
into the scale item "I get sufficient time to work in a hygienic
and safe food way." The sub-themes of outcome expectations
regarding the advantages and disadvantages of performing
proper food safety behaviors were also used in the outcome
expectations scale. For example, reducing the risk of
foodborne illnesses, avoiding lawsuits, and maintaining a
good reputation were sub-themes that were incorporated

Food Protection Trends - September/October 2020

Table of Contents for the Digital Edition of Food Protection Trends - September/October 2020