Food Protection Trends - January/February 2023 - 30

8.9% of the variance in posttest compliance (R2 = 0.089,
P < 0.000). Posttest knowledge was a significant predictor of
compliance (B = 0.313; P < 0.000), but intention was not.
DISCUSSION
Effect of food safety trainings on compliance
In this case study, training resulted in apparent increases
in compliance (i.e., decreases in noncompliance) for criteria
associated with the training topics altogether and specifically
for the lessons on GMPs 1 and food defense. (Fig. 1).
However, with a one-tailed comparison P-value of 0.08 and
sample size of eleven processing plants, the significance of
this result was marginal, and more audit samples would be
needed to provide more confidence in this interpretation of
the data. This uncertain relationship between training and
compliance has been reported by other researchers. A gap
between training and behavior has been reported (1, 3).
However, Nieto-Montenegro et al. (11) reported a reduction
in rates of noncompliance for several food safety practices
in their food safety training assessment study. The results
from the present study could be influenced by the lower
response rate (responding plants might have lower pretest
noncompliance than nonresponding plants), respondent
bias (reporting less noncompliance than actually present), or
high starting levels of compliance. Each plant that completed
supervisor audits did not administer the same type of
training to all employees, so comparisons could not be made
for the effect of each training on measured compliance within
those individual facilities.
Effect of food safety trainings on knowledge
There was a significant increase in knowledge indicated
by the aggregate data. This association between a training
intervention and an increase in knowledge is consistent with
findings from other training assessment studies, although
those researchers did not reference learning theories in
their work (9, 11, 13). In the comparison between pre- and
posttest training knowledge scores, group training was
generally more effective for generating knowledge gains. The
two group training options (direct instruction and social and
situated learning) resulted in significant increases in knowledge
as indicated by the aggregate data, but the programmed
instruction did not. Because both of the interventions that
had significant effects were performed in group settings
rather than an individual setting, group training appeared to
be more effective than individual training. However, because
the experimental design did not directly allow for this comparison,
definitive claims comparing group and individual
training could not be made. Increased presence of management
during training may help convey food safety values of
management and the importance of training and allow for
discussion between employees and management that could
resolve any questions or confusion. Occupational safety leadership
had a direct positive effect on safety performance and
30
Food Protection Trends January/February
indicates that the relationship is mediated by occupational
safety culture (16). Another explanation for the knowledge
score differences across training delivery methods could be
that the knowledge increase scores were biased by the pretest
knowledge scores, whereas the posttest knowledge scores
were very similar for all training delivery methods (83.33 to
85.00) compared with the pretest knowledge scores (75.00
to 80.00).
The computer-based training events performed in a group
setting utilizing direct instruction was more effective than
the group discussion-based virtual tour utilizing social and
situated learning training in terms of increasing knowledge
and intentions. This difference could be due to the fact
that the virtual tour training is a new technology that plant
operators were unfamiliar with and so might not have been
utilized to its fullest potential. Future experiments should
compare trainings that are designed on the same platform
with different learning theories to determine whether
differences in effectiveness are due to the technological
platform or the learning theory. The differences also could
be due to the fact that the direct instruction module was
very structured, whereas deviation and variation was
possible in the social and situated learning module. The
direct instruction training option was the most effective in
terms of knowledge transfer. However, this training option
was not designed for this study specifically, and the plant
employees administering the training decided to take what
was given to them and deliver the training themselves. Thus,
these results should be interpreted cautiously. One outside
factor that might have affected the performance of the social
and situated learning training option is the familiarity of the
trainers and employees with the training medium. Virtual
reality is still a novel concept, so trainers may not have been
comfortable leading the training, despite having received
instructions via a webinar.
Effect of food safety trainings on intentions
For the aggregate data, there was a significant increase in
food safety intentions after training, but when comparing the
three training deliveries, there was only a significant increase
for the direct instruction training. Training that incorporates
the direct instruction learning theory may have a stronger
impact on increasing employee intentions than do the other
two learning theories. However, because the distributions for
the pre- and posttest intentions were so negatively skewed,
we question whether our survey questions led to social
positive desirability bias. Thus, it was unclear how accurate
these comparisons were.
Mediators of intentions and compliance
There was an overall significant predictive effect of pre- and
posttest employee attitudes, perceived norms, and personal
agency on employee food safety pre- and posttest intentions
for the aggregate data, which was expected based on the
Food Protection Trends - January/February 2023

Table of Contents for the Digital Edition of Food Protection Trends - January/February 2023
