































 

 Epidemiology and Society Health Review| ESHR 
Vol. 5, No. 2, 2023, pp. 76-89 ISSN 2656-6052 (online) | 2656-1107 (print) 
 
 

 

      10.26555/eshr.v5i2.6996  

 
 

76 

 
 

  

Research Article  
 
Health Belief Model Application on Food Safety Behavior 
of Bantul Beach Tourism Culinary Food Handlers 
 
Asep Rustiawan1*, Ichtiarini Nurullita Santri1, Yohane Vincent Abero Phiri2 
1 Faculty of Public Health, Universitas Ahmad Dahlan, Yogyakarta, Indonesia 
2 School of Public Health, College of Public Health, Taipei Medical University, Taipei, Taiwan 
 
* Correspondence: asep.rustiawan@ikm.uad.ac.id. Phone: +6285743847969  
 
Received 13 December 2022; Accepted 15 August 2023; Published 05 September 2023 
 

ABSTRACT 

Background: Maintaining food safety is very important to reduce the incidence of 
foodborne diseases. Previous studies have shown that food safety is related to the healthy 
behavior of food handlers, while healthy behavior is related to their beliefs and perceptions. 
The Health Belief Model (HBM) is a theory with a concept of individual reasons for carrying 
out healthy behaviors based on the perception of disease threats and efforts to improve 
behavior when facing threats. This study aims to predict and analyze the effect of variables 
based on the HBM construct on food safety behavior. 
Method: A total of 80 food handlers from all seafood stalls on the Bantul coast were 
interviewed face to face using a printed questionnaire regarding seriousness, vulnerability, 
benefits, barriers, a stimulus to action, and self-efficacy. The food safety behavior of food 
handlers was observed using observation sheets. Data were evaluated using Structural 
Equation Model-Partial Least Square (SEM-PLS) Version 3.0.  
Results: HBM can predict 35.1% of food handlers' food safety behavior variance in the 
Bantul beach tourism culinary area (R2 adjusted = 0.351). Of all the HBM construct variables 
analyzed, two variables had a significant effect, namely the stimulus-to-act variable with a 
path coefficient value (β = 0.305, p = 0.009 <0.05) and the seriousness variable (β = 0.302, 
p = 0.045 <0.05). 
Conclusion: HBM succeeded in predicting the food safety behavior of food handlers in 
tourist culinary delights in Bantul, Yogyakarta, Indonesia. The stimulus-to-action variable 
was the strongest predictor positively affecting food handlers' food safety behavior, followed 
by the seriousness variable. 

Keywords: Bantul beach culinary; Food safety; HBM theory; SEM-PLS 

 

 

https://doi.org/10.26555/eshr.v3i1.3629
https://doi.org/10.26555/eshr.v3i1.3629


 
Rustiawan et al. (Health Belief Model Application on Food Safety Behavior of Bantul Beach Tourism Culinary Food Handlers) 

 
 

 
Vol. 5, No. 2, 2023, pp. 69-89 77     10.26555/eshr.v5i2.6996 

  

INTRODUCTION 

The south coast of Bantul District has become a tourist destination increasingly attracting 
tourist. The results of a survey by researchers at the end of 2020 showed at least 80 culinary 
stalls operating along the Bantul coast. With the growing development of this culinary 
business, the tourist location of the Bantul beach has turned into a busy culinary area on the 
Bantul beach. However, street food is also a cause of severe public health problems.1,2 The 
results of previous studies showed that 40% of food handlers processed food unsafely. 3 Other 
researchers also reported that most of the existing culinary stalls needed more facilities to 
process food healthily, and the behavior of food handlers needed to be more hygienic.4,5 
Healthy behavior is related to beliefs and perceptions of food safety. Rosenstock states that 
individual beliefs influenced by their self-perception of vulnerability, seriousness, benefits, 
barriers, self-efficacy, and external encouragement or stimulus to act will shape behavior 
toward food safety.6 

HBM effectively predicts strong predictors of food safety (FS) behavior.7,8 Nevertheless, based 
on the results of a literature search, research has yet to be conducted in the culinary area of 
the Bantul beach tourism on applying this theory by using SEM analysis to measure the 
influence and predict strong predictors of the HBM construct variable on FS behavior. 
Knowledge related to this is useful in developing food safety improvement programs following 
local conditions to produce culinary processed food for beach tourism that is nutritious and 
safe for consumption. Thus, research to measure the magnitude of the influence and predict 
the determinants of the HBM construct variable on the FS behavior of food handlers in the 
Bantul beach tourism culinary area is very important. 

This study aims to analyze the magnitude of the influence and predict strong predictors of 
perceived vulnerability, seriousness, benefits, barriers, self-efficacy, and stimulus to act on the 
FS behavior of food handlers in the culinary tourism of the Bantul coast. The alternative 
hypothesis is formulated and illustrated in Figure 1 based on HBM Theory.6  

H1: Vulnerability, seriousness, benefits, barriers, self-efficacy, and stimulus-to-act 
variables have a significant positive effect on FS behavior variables. 

 

 

 
 

 

 

 
Figure 1. Preliminary SEM Model: Structural Relations of Exogenous 

Variables and Endogenous Variables 

https://doi.org/10.26555/eshr.v3i1.3629
https://doi.org/10.26555/eshr.v3i1.3629


 
Rustiawan et al. (Health Belief Model Application on Food Safety Behavior of Bantul Beach Tourism Culinary Food Handlers) 

 
 

 
Vol. 5, No. 2, 2023, pp. 69-89 78     10.26555/eshr.v5i2.6996 

  

METHOD 

Research Design and Sampling Techniques 

The research was carried out with a cross-sectional design.9 Exogenous variable data 
(vulnerability, seriousness, benefits, barriers, self-efficacy, and stimulus to act) and 
endogenous variables (food safety behavior) were collected simultaneously. The total 
population of culinary stalls on the Bantul Yogyakarta beach tour is 80 stalls; each shop is 
represented by one food handler as a respondent, so 80 food handlers are obtained as 
respondents.10 Respondents were selected through inclusion and exclusion criteria according 
to the research objectives. 

Questionnaire Design 

The questionnaire developed consists of the characteristics of the respondents and six 
exogenous variables of the HBM construct, namely perceptions of vulnerability, seriousness, 
benefits, barriers, stimulus to action, and self-efficacy, as well as one endogenous variable, 
namely food safety behavior. Exogenous variables are measured with eight statements so 
that 48 reports are obtained. FS behavior variables are measured by observation using the 
Food Safety Score (FSS) form developed by Mudjajanto.11 The questionnaire uses a 5-point 
Likert scale from point 1 (strongly disagree) to point 5 (strongly agree). 

The questionnaire was validated at the Kulonprogo beach tourism culinary location in 
Yogyakarta, involving 32 respondents. Based on the validity test, an invalid statement was 
obtained, which was issued, namely, one report on the stimulus to act. The results of the final 
analysis got 47 valid statements (Corrected Item-Total Correlation test results > 0.20, and 
Cronbach's Alpha results = 0.949 > 0.6(12).12 Furthermore, food handlers' perceptions of 
vulnerability, seriousness, benefits, barriers, stimulus to action, and self-efficacy 
questionnaires that were valid and reliable were used for research. 

Data collection procedure 

Face-to-face interviews collected data to evaluate perceptions of vulnerability, seriousness, 
benefits, barriers, stimulus to action, and self-efficacy. It took between 30-45 minutes to 
interview each respondent. The data of FS behavior was evaluated using non-participatory 
observations made when processing food. The time required is between 45-60 minutes for 
each respondent. This research received ethical approval from Ahmad Dahlan University 
Research Ethics Committee No. 012208112 on August 24, 2022. 

Statistical analysis 

The validity and reliability of the instrument were evaluated by Measurement Model Evaluation 
using SEM analysis with SmartPLS 3.0.13 The influence relationship between perceptions of 
vulnerability, seriousness, benefits, barriers, stimulus to action, and self-efficacy on FS 
behavior was analyzed by Structural Model Analysis using SEM-PLS at a 95% confidence 
level through Goodness of Fit Analysis and Analysis of hypothesis testing. 

 

 

https://doi.org/10.26555/eshr.v3i1.3629
https://doi.org/10.26555/eshr.v3i1.3629


 
Rustiawan et al. (Health Belief Model Application on Food Safety Behavior of Bantul Beach Tourism Culinary Food Handlers) 

 
 

 
Vol. 5, No. 2, 2023, pp. 69-89 79     10.26555/eshr.v5i2.6996 

  

RESULTS 

Characteristics of Respondents 
Respondents in this study were food handlers who made direct contact with food, totaling 80 
people and representing every culinary stall located on the south coast of Bantul District. The 
respondents' characteristics in this study were gender, age, education level, length of work 
experience, and beach location, which were obtained through interviews.  

Table 1. Frequency Distribution of Respondents Based on 
Gender, Age, Education, Length of Work, and Beach Location 

 
Characteristics of Respondents Amount Percentage (%) 
Gender 

Man 0 0% 
Woman 80 100% 

Age 
Adult (25 years-45 years)* 54 67.50% 
Elderly ≥46 years** 26 32.50% 

Education 
Junior High School 45 56% 
SMA/SMK 34 43% 
Undergraduate diploma 1 1% 

Length of work (years) 
Average 9.9  
Minimum 7.7  
Maximum 15.3  

Beach Location 
Parangtritis 19 23.75% 
Depok 26 32.50% 
Samas 2 2.50% 
Goa Cemara  5 6.25% 
Kuwaru 6 7.50% 
Baru 22 27.50% 

     Information:  * Youngest age 25 years 
   ** Oldest 50 years old 

Table 1 shows that all respondents are female. The most age group is the adult group (25-45 
years), as much as 67.5%. The youngest respondent is 25 years old, and the oldest is 50. The 
average length of work as a food handler is 9.9 years, with a junior high school education level 
(56%) and only 1% above senior high school. Most respondents came from Depok Beach 
(32.50%), and the least came from Samas Beach (2.50%). 

Measurement Model Evaluation 

The measurement model undergoes evaluation through validity and reliability tests. The 
validity test assesses construct validity, examining whether the chosen indicators are suitable 
for measuring latent variables. Conversely, the reliability test focuses on assessing the 
consistency of indicators and whether the statement items used as indicators yield consistent 
responses from participants. The validity test includes considerations of Convergent Validity, 
as indicated by the Loading Factor and Average Variance Extracted (AVE) values, as well as 
Discriminant Validity, measured through Cross-Loading and Fornell-Larcker Criterion values. 

https://doi.org/10.26555/eshr.v3i1.3629
https://doi.org/10.26555/eshr.v3i1.3629


 
Rustiawan et al. (Health Belief Model Application on Food Safety Behavior of Bantul Beach Tourism Culinary Food Handlers) 

 
 

 
Vol. 5, No. 2, 2023, pp. 69-89 80     10.26555/eshr.v5i2.6996 

  

The evaluation of the reliability test is presented through Composite Reliability and Cronbach's 
Alpha values. The initial structural model proposed is depicted in Figure 2 below. 

 

Figure 2. Initial structural model of the relationship between vulnerability, seriousness, 
benefits, barriers, stimulus to action, self-efficacy, and the food safety behavior of food 

handler’s culinary tourism in Bantul 
 
Based on Figure 2 above, there are indicator items that have a Loading Factors value of <0.7 
and are then discarded, namely: vulnerability variables (4,6,7) seriousness (3,4,5,6,7); benefit 
variable (5,6); obstacle variables (1,3,4,6,7,8); stimulus variable (1,3); and self-efficacy 
variables (2,6,7,8). Furthermore, the initial structural model evaluated for its Loading Factor 
was tested again, and a second model was produced, as shown in Figure 3. All indicators 
have a Loading Factor value of > 0.7, so it meets the requirements. AVE values for all variables 
> 0.5 so that the model is said to be convergently valid.  
 
Table 2 displays the loading factor values of the manifest indicators for all variables, which are 
consistently above 0.70, ranging from 0.715 to 0.930. Additionally, the AVE values for all 
variables are consistently above 0.50, ranging from 0.566 to 1.000. These results confirm the 
convergent validity of all indicators.14 Table 3 presents the cross-loading values for each 
manifest indicator of every variable, all of which exceed 0.70 and surpass the correlations with 
other latent variables. Similarly, the data in Table 4 reveals Fornell-Larcker Criterion values 
greater than 0.70, indicating stronger distinctions than the correlations with other latent 
variables. This affirms the discriminant validity of all analyzed manifest indicators (14). 
Furthermore, Table 5 demonstrates that all Cronbach's Alpha values exceed 0.70, ranging 
from 0.717 to 1.000, and the Composite Reliability value exceeds 0.70, ranging from 0.798 to 
1.000. These findings confirm the reliability of all analyzed indicators.15 
 

https://doi.org/10.26555/eshr.v3i1.3629
https://doi.org/10.26555/eshr.v3i1.3629


 
Rustiawan et al. (Health Belief Model Application on Food Safety Behavior of Bantul Beach Tourism Culinary Food Handlers) 

 
 

 
Vol. 5, No. 2, 2023, pp. 69-89 81     10.26555/eshr.v5i2.6996 

  

 
Figure 3. The final structural model of the relationship between vulnerability, 
seriousness, benefits, barriers, stimulus to action, self-efficacy, and the food 

safety behavior of food handler’s culinary tourism in Bantul 
 

Structural Model Analysis 
 
Figure 3 illustrates the structural model, which describes the causal relationships between 
latent variables constructed based on established theory. The variables, namely vulnerability, 
seriousness, benefits, obstacles, stimulus to action, and self-efficacy, exert direct influences 
on FS behavior variables. Structural Model Analysis aims to ascertain the relationships 
between latent (exogenous) and endogenous variables. In this study, we specifically 
investigate direct relationships. The subsequent analysis of the structural model will be 
elaborated upon, including discussions on the Goodness of Fit Model and hypothesis testing. 
 
Model Feasibility Analysis (Goodness of Fit) 
 
The Model Feasibility Analysis aims to determine the degree to which empirical data align with 
the theoretically structured model depicting relationships between variables. Various indices 
were employed to assess the model's feasibility, as outlined in Tables 2, 3, 4, and 5. Table 6 
presents the data, indicating an R-squared adjusted value of 0.351 with a p-value of 0.000 (p 
< 0.05). This suggests that the model can account for 35.1% of the variability in FS behavior. 
Chin categorizes R-squared values of 0.67 as substantial (strong) and 0.33 as moderate.16 
Thus, the R-squared value of 0.351 in this study falls into the moderate range, signifying that 
the structural model constructed is sufficiently robust for predicting FS behavioral intentions. 
To validate the overall performance of both the measurement model and the structural model, 
we calculate the Goodness of Fit (GoF) using the following formula: 

https://doi.org/10.26555/eshr.v3i1.3629
https://doi.org/10.26555/eshr.v3i1.3629


 
Rustiawan et al. (Health Belief Model Application on Food Safety Behavior of Bantul Beach Tourism Culinary Food Handlers) 

 
 

 
Vol. 5, No. 2, 2023, pp. 69-89 82     10.26555/eshr.v5i2.6996 

  

GoF =   √  AVE  x  R2    

        =   √  0.703  x  0.351 

        =      0.497 
 

Table 2. Loading Factors and AVE values of the manifest indicators for all HBM variables 

Manifest 
HBM Variables 

Efficacy Obstacle Vulnerability Seriousness Benefit Stimulus Food safety 
behavior (FSB) 

fsb       1.000 
ef1 0.728       

ef3 0.826       

ef4 0.859       

ef5 0.795       

hb2  0.897      

hb5  0.727      

kr1   0.773     

kr2   0.774     

kr5   0.729     

kr8   0.754     

ks1    0.880    

ks2    0.809    

ks3    0.730    

ks8    0.873    

mn1     0.819   

mn2     0.815   

mn3     0.775   

mn4     0.739   

mn7     0.715   

mn8     0.764   

st2      0.744  

st4      0.930  

st5      0.807  

st6      0.910  

st7      0.836  

AVE 0.646 0.667 0.566 0.730 0.596 0.719 1.000 
 
 
 
 
 
 
 

https://doi.org/10.26555/eshr.v3i1.3629
https://doi.org/10.26555/eshr.v3i1.3629


 
Rustiawan et al. (Health Belief Model Application on Food Safety Behavior of Bantul Beach Tourism Culinary Food Handlers) 

 
 

 
Vol. 5, No. 2, 2023, pp. 69-89 83     10.26555/eshr.v5i2.6996 

  

Table 3. Cross Loading Value of the manifest indicators for all HBM variables 

Manifest 
HBM Variables 

Efficacy Obstacle Vulnerability Seriousness Benefit Food safety 
behavior Stimulus 

fsb 0.360 0.439 0.324 0.560 0.522 1.000 0.526 

ef1 0.728 0.433 0.580 0.517 0.529 0.226 0.468 

ef3 0.826 0.383 0.523 0.428 0.494 0.351 0.421 
ef4 0.859 0.551 0.538 0.527 0.483 0.336 0.545 

ef5 0.795 0.436 0.591 0.544 0.608 0.173 0.533 

hb2 0.465 0.897 0.433 0.563 0.440 0.423 0.406 
hb5 0.464 0.727 0.483 0.502 0.602 0.273 0.421 

kr1 0.458 0.438 0.773 0.480 0.438 0.195 0.431 
kr2 0.608 0.438 0.774 0.526 0.529 0.221 0.423 

kr5 0.425 0.388 0.729 0.475 0.483 0.230 0.395 

kr8 0.558 0.497 0.754 0.587 0.465 0.199 0.479 

ks1 0.529 0.652 0.593 0.880 0.621 0.570 0.649 

ks2 0.527 0.474 0.541 0.809 0.633 0.327 0.423 
ks3 0.500 0.332 0.730 0.487 0.414 0.322 0.401 

ks8 0.522 0.506 0.601 0.873 0.722 0.482 0.553 

mn1 0.515 0.517 0.459 0.651 0.819 0.522 0.529 
mn2 0.627 0.545 0.539 0.672 0.815 0.410 0.477 

mn3 0.459 0.460 0.428 0.549 0.775 0.375 0.357 
mn4 0.483 0.408 0.615 0.488 0.739 0.364 0.469 

mn7 0.575 0.456 0.525 0.616 0.715 0.264 0.488 

mn8 0.327 0.414 0.342 0.572 0.764 0.409 0.349 

st2 0.470 0.378 0.421 0.494 0.485 0.292 0.744 
st4 0.574 0.465 0.482 0.617 0.581 0.563 0.930 
st5 0.480 0.440 0.545 0.608 0.459 0.302 0.807 
st6 0.563 0.524 0.498 0.619 0.529 0.530 0.910 
st7 0.449 0.287 0.480 0.440 0.378 0.432 0.836 

Source: SmartPLS Program Data Output Version 3.0 
 

Table 4. Fornell-Larcker Criterion Values of the manifest indicators for all HBM variables 

 Efficacy Obstacle Vulnerability Seriousness Benefit Food safety 
behavior Stimulus 

Efficacy 0.804       

Obstacle 0.560 0.816      

Vulnerability 0.679 0.545 0.752     

Seriousness 0.611 0.649 0.678 0.855    

Benefit 0.636 0.606 0.617 0.766 0.772   

FS behavior 0.360 0.439 0.324 0.560 0.522 1.000  

Stimulus 0.600 0.496 0.564 0.653 0.575 0.526 0.848 
Source: SmartPLS Program Data Output Version 3.0 

 

https://doi.org/10.26555/eshr.v3i1.3629
https://doi.org/10.26555/eshr.v3i1.3629


 
Rustiawan et al. (Health Belief Model Application on Food Safety Behavior of Bantul Beach Tourism Culinary Food Handlers) 

 
 

 
Vol. 5, No. 2, 2023, pp. 69-89 84     10.26555/eshr.v5i2.6996 

  

Table 5. Cronbach's Alpha and Composite Reliability 
Values of the indicators for all HBM variables 

 Cronbach's Alpha Composite Reliability 
Efficacy 0.823 0.879 
Obstacle 0.717 0.798 
Vulnerability 0.813 0.867 
Seriousness 0.821 0.890 
Benefit 0.865 0.898 
FS behavior 1.000 1.000 
Stimulus 0.903 0.927 
Source: SmartPLS Program Data Output Version 3.0 

 
The GoF is considered small when its value falls within the range of 0.00 to 0.25, moderate 
when it falls within the range of 0.25 to 0.36, and large when its value exceeds 0.36. 
Consequently, the value of 0.497 falls into the category of a large GoF. Overall, the combined 
performance of the measurement model and the structural model hypothesized in this study 
aligns well with the empirical data and can be accepted. 
 

Table 6. R Square adjusted value 

 Original 
Sample (O) 

Sample 
Means 

(M) 
Standard Deviation 

(STDEV) 
T Statistics 
(|O/STDEV|) P Values 

FS behavior 0.351 0.400 0.095 3.679 0.000 
Source: SmartPLS Program Data Output Version 3.0 

 
Hypothesis Test Analysis 
Hypothesis testing analysis is conducted once the theoretically developed structural model is 
deemed feasible, indicating its representativeness and alignment with empirical data. In this 
study, hypothesis testing analysis focused on examining the relationships between the 
variables outlined in the structural model presented in Figure 3. Additionally, it involved the 
calculation of statistical t-values and p-values to assess the direct influence between latent 
variables, as detailed in the path coefficient table derived from the SmartPLS bootstrapping 
analysis results, as presented in Table 7. 

Table 7. Summary of the results of hypothesis testing the direct effect of the variables  

 Original 
Sample (O) 

Sample 
Means 

(M) 
T Statistics 
(|O/STDEV|) P Values 

Efficacy -> FS Behavior -0.073 -0.096 0.533 0.297 
Barriers -> FS Behavior 0.107 0.109 0.839 0.201 
Vulnerability -> FS Behavior -0.195 -0.145 1.288 0.099 
Seriousness -> FS Behavior 0.302 0.257 1.695 0.045 
Benefits -> FS Behavior 0.217 0.237 1.235 0.108 
Stimulus -> FS Behavior 0.305 0.320 2.377 0.009 

 

https://doi.org/10.26555/eshr.v3i1.3629
https://doi.org/10.26555/eshr.v3i1.3629


 
Rustiawan et al. (Health Belief Model Application on Food Safety Behavior of Bantul Beach Tourism Culinary Food Handlers) 

 
 

 
Vol. 5, No. 2, 2023, pp. 69-89 85     10.26555/eshr.v5i2.6996 

  

The path coefficient of the structural model studied is shown in Figure 3, and the data in Table 
7 indicates the strength of the direct relationship between variables. Two of the six exogenous 
variables analyzed had a p-value <0.05, meaning a significant effect, namely the seriousness 
variable and the stimulus variable to act. Among the two critical variables, the stimulus variable 
for action has the most considerable path coefficient (β = 0.305, p = 0.009<0.05), followed by 
the seriousness variable (β = 0.302, p = 0.045<0.05). This means that the stimulus variable 
for action is the strongest predictor that has a positive effect on FS behavior from food 
handlers, followed by the seriousness variable, which also positively impacts FS behavior. 
Other variables in this study have path coefficient values that are not significant. Thus, it can 
be concluded that the research hypothesis can be accepted as a whole except for the 
variables of vulnerability, benefits, barriers, and self-efficacy. 

 
DISCUSSION  

This research assessed the FS behavior of food handlers and relevant factors according to 
HBM in the culinary area of Bantul beach tourism. Overall, the results of this study were seen 
from several variables, namely: perceived vulnerability variables, perceived seriousness 
variables, perceived benefits variables, perceived barriers variables, self-efficacy variables, 
and stimulus-to-act variables directly influencing FS behavior variables. Of the various 
variables examined, the seriousness variable and the stimulus variable have a significant 
favorable influence on FS behavior. Food handlers who believe FS behavior can cause severe 
problems if not done correctly are a decisive driving factor in realizing FS behavior. If the belief 
increases, the actualization of FS behavior will also increase. Likewise, the stimulus will be a 
strong driver in learning FS behavior. If the catalyst for action obtained by food handlers 
increases, the actualization of FS behavior will increase. 

The level of seriousness/severity positively affects FS behavior, following Handson et al.17 
state that the perceived seriousness/severity of the foodborne disease is positively related to 
the two measurable dimensions of FS behavior, sanitation and cross-contamination. Those 
who are more aware of the risk of FS will have better food handling practices. An increase in 
perceived vulnerability can result in more remarkable behavior charges when accompanied 
by an increase in perceived seriousness/severity.18 So, this study shows that perceived 
seriousness/severity is essential in maintaining the behavior and attitudes related to FS 
behavior. 

Food handlers' beliefs about the seriousness of FS behavior are influenced by socio-cultural 
background/individual characteristics. As stated by Sjoberg and Wachinger et al., several 
factors, such as knowledge, experience, attitudes, and emotions, can influence personal 
thoughts and judgments regarding the seriousness and acceptance of risk in the FS 
management system.19,20 

It is further said that the level of knowledge of food handlers significantly shapes their 
perceptions of various aspects of proper food handling practices. These perceptions include 
the expected benefits of a commitment to acceptable food handling practices, perceived 
barriers to such commitment, and risks scheduled from food handling malpractices. The 
correlation test results also show that although knowledge positively affects food handlers' 
beliefs about the benefits and risks, knowledge has a weaker negative effect on their beliefs 

https://doi.org/10.26555/eshr.v3i1.3629
https://doi.org/10.26555/eshr.v3i1.3629


 
Rustiawan et al. (Health Belief Model Application on Food Safety Behavior of Bantul Beach Tourism Culinary Food Handlers) 

 
 

 
Vol. 5, No. 2, 2023, pp. 69-89 86     10.26555/eshr.v5i2.6996 

  

about barriers.21 Experts recommend implementing future food safety educational 
interventions through a behavioral change theoretical framework.22,23 Education level is 
essential to understanding and developing appropriate training tools in food safety. Academic 
levels vary from illiterate to primary school or higher studies, so it can help determine the 
approach needed.24 

The stimulus variable (cues to action) positively influences FS behavior; in line with Pie, 25 a 
stimulus to action is defined as a factor that influences individuals to change their behavior 
toward implementing FS behavior. Three elements are measured in cues to action: (1) media 
such as mass media or print media, (2) communication such as communication between 
handlers and stakeholders, (3) food labels on food packaging and posters in the kitchen. 26 
Those who frequently listen to the radio and watch television news about food poisoning are 
likelier to exhibit FS behavior.17 Individuals can also take health-related actions if they receive 
external cues to act, such as educational cues to action or media messages.26 

HBM is effective in predicting strong predictors of FS behavior.7,8 This study showed that HBM 
succeeded in predicting 35.1% of the variance of FS behavior of food handlers in the culinary 
area of Bantul beach tourism. Of all the HBM constructs analyzed, the stimulus to act is the 
most influential variable on FS behavior, followed by the variable of seriousness. Other 
variables have no significant effect. 

In another study, HBM showed 58.9% significance in the practice of food handlers. All HBM 
dimensions significantly contribute to explaining food handlers.21 Among the different HBM 
constructs, beliefs about the expected benefits of proper food handling practices substantially 
contributed to the model's predictability. In contrast, food handlers' perceptions of barriers to 
adequate food handling had the weakest contribution. 

Previous studies have shown disparities in the utilization of the HBM construct within the 
context of food handling. For instance, only one study27 has delved into the role of perceived 
benefits in elucidating food handling practices. Conversely, a limited number of studies17,28,29 
have explored how food handlers' beliefs impact barriers to proper food handling. 
Simultaneously, all prior research has examined the influence of perceived risk on food 
handlers' practices. Food product safety is an absolute demand of consumers. Producers are 
responsible, so applying existing laws and developing new production techniques is critical to 
protect human health.30 

The strength of this study is that it is the first study of specific FS behavior associated with six 
indicators of HBM along the Bantul beach resorts. In addition, this study also uses PLS 
analysis which is rarely used in similar studies. This study also has limitations. Namely, the 
number of samples is relatively small because many places to eat are closed due to the 
COVID-19 pandemic. 

 

 

 

 

https://doi.org/10.26555/eshr.v3i1.3629
https://doi.org/10.26555/eshr.v3i1.3629


 
Rustiawan et al. (Health Belief Model Application on Food Safety Behavior of Bantul Beach Tourism Culinary Food Handlers) 

 
 

 
Vol. 5, No. 2, 2023, pp. 69-89 87     10.26555/eshr.v5i2.6996 

  

CONCLUSION 

The research hypothesis can be accepted as a whole except for the variables of vulnerability, 
benefits, barriers, and self-efficacy predicted 35.1% of the FS behavior variance of food 
handlers in the Bantul beach tourism culinary area. The stimulus for action variable is the 
strongest and most significant predictor with a significant positive effect on FS behavior, 
followed by the seriousness variable, which also has a significant positive impact.  

 

Declarations 

Authors' contribution 

AR designed the study, supervised all data analyses, and drafted the manuscript. INS drafted 
the manuscript. YVAP data analyzed. Each author contributed to the intellectual content of the 
manuscript. This manuscript has been read and approved by the authors.   

Funding Statement 

This project received support from the Research Institute and Community Service at 
Universitas Ahmad Dahlan, Yogyakarta, Indonesia, under grant number PD-156/SP3/LPPM-
UAD/VII/2022. 

Conflict of interest 

There is no conflict of interest in this research.  

REFERENCES 

1. De Sousa CP. The impact of food manufacturing practices on foodborne diseases. Braz 
Arch Biol Technol. 2008;51(4):815–823. Retrieved March 19, 2020, from 
http://www.scielo.br/scielo.php?script=sci_arttext&pid=S1516-89132008000400020 

2. Ghosh M, Wahi S, Kumar M, Ganguli A. Prevalence of enterotoxigenic Staphylococcus 
aureus and Shigella spp. in some raw street vended Indian foods. Int J Environ Health 
Res. 2007;17(2):151–156. doi:10.1080/09603120701219204 

3. Aspiani M, Rustiawan A. Relationship of knowledge, attitudes of food handlers and 
sanitation facilities to food safety in restaurants in the culinary tourism area of Depok 
beach, Bantul Regency in 2019. URECOL-11. Yogyakarta: UNISA; 2020 

4. Kurniaty. Relationship between sanitation hygiene practices and Escherichia coli 
contamination in processed grilled fish at seafood food stalls at Depok Beach, Bantul, 
Yogyakarta. Journal Formil (Scientific Forum). 2017:2(2). 

5. Suryani D. Existence of Grilled Pomfret Fish Germs and Grilled Cutlery. Journal of Public 
Health. 2015; pp:191–196. 

6. Rosenstock IM. Historical Origins of the Health Belief Model. Health Education 
Monographs. 1974;2(4):328–335. doi:10.1177/109019817400200403 

7. Becky JB. Using the Health Belief Model to Determine Differences in University 
Foodservice Employees' Beliefs and Perceptions About Handwashing and Foodborne 
Illness. Manhattan, Kansas; 2013 

8. Asma P, Mohammad-Ali MS, Hossein F, Jalal S. Food Handling Behaviors Based on the 
Health Belief Model in Iranian Women. Kerman, Iran; 2012 

https://doi.org/10.26555/eshr.v3i1.3629
https://doi.org/10.26555/eshr.v3i1.3629
http://www.scielo.br/scielo.php?script=sci_arttext&pid=S1516-89132008000400020
https://sci-hub.tw/10.1080/09603120701219204


 
Rustiawan et al. (Health Belief Model Application on Food Safety Behavior of Bantul Beach Tourism Culinary Food Handlers) 

 
 

 
Vol. 5, No. 2, 2023, pp. 69-89 88     10.26555/eshr.v5i2.6996 

  

9. Notoatmodjo. Health Research Methods. Jakarta: Rineka Cipta. Jakarta; 2012. 
10. Sugiyono. Quantitative, Qualitative and R&D Research Methods. Bandung: Alphabet, 

Bandung; 2019. [In Indonesian] 
11. Mudjajanto ES. Food Safety: Training and Development of Technology and Snack Food 

Safety. Bogor: Department of GMSK, Faculty of Agriculture, IPB and Directorate General 
of Basic Education, Ministry of National Education; 1999. [in Indonesian] 

12. Murti B. Measurement Validity and Reliability. Doctoral Study Program Matriculation, 
Faculty of Medicine, UNS; 2011. [in Indonesian] 

13. Ringle, Christian M, Wende Sven, Becker J-M. SmartPLS 3. Boenningstedt: SmartPLS; 
2015 

14. Ghozali I. Structural equation modeling: An alternative method with Partial Least Squares 
(PLS). Semarang: Diponegoro University Publishing Agency; 2014. 

15. Nunnally JC, Bernstein I. Psychometric theory. In Psychometric theory. 1994;pp:264–265 
16. Chin WW. The Partial Least Squares Approach to Structural Equation Modeling, Modern 

Methods for Business Research. 1998;pp:295-336 
17. Hanson JA, Hughes SM, Liu P. Use of Health Belief Model Variables to Examine Self-

Reported Food Handling Behaviors in a Sample of US Adults Attending a Tailgate Event. 
Journal of Food Protection. 2015;78(12):2177–2183. doi:10.4315/0362-028x.jfp-15-077 

18. Wang M, Huang L, Pan C, Bai L. Adopt proper food-handling intention: An application of 
the health belief model. Food Control. 2021;127. doi:10.1016/j.foodcont.2021.108169 

19. Sjoberg L. Factors in risk perception. Risk Anal. 2000;20():1–11 
20. Wachinger G, Renn O, Begg C, Kuhlicke C. The risk perception paradox—implications for 

governance and communication of natural hazards. Risk Anal. 2013;33():1049–1065 
21. Habiballah MA, Alhelalat JA, Masadeh MA, Al-Ababneh M. Use of the Health Belief Model 

(HBM) to Examine Food Handlers' Practices in Five Stars Hotels at Petra City, Jordan. 
Journal of International Business and Economics. 2020;8(1):32-39. 
doi:10.15640/jibe.v8n1a3 

22. Sivaramalingam B, Young I, Pham MT, Waddell L, Greig J, Mascarenhas M, 
Papadopoulos A. Scoping Review of Research on the Effectiveness of Food-Safety 
Education Interventions Directed at Consumers. Foodborne Pathog. 2015;12():561–570. 

23. Young I, Reimer D, Greig J, Meldrum R, Turgeon P, Waddell L. Explaining Consumer Safe 
Food Handling Through Behavior-Change Theories: A Systematic Review. Foodborne 
Pathog. 2017;14():609–622. 

24. Mishra M, Prakash J. Status of Education and Training among Food Handlers from 
Selected Catering Establishments. International Journal of Food, Nutrition and Dietetics. 
2017;5(2). doi:10.21088/ijfnd.2322.0775.5217.3 

25. Piee NF, Abu R, Rashid AM. The Relationship Between Cues to Action Towards Food 
Safety Behavior among Food Program Students at Vocational School. Middle-East Journal 
of Scientific Research 19 (Innovation Challenges in Multidisciplinary Research & Practice). 
2014;pp:133-137, doi: 10.5829/idosi.mejsr.2014.19.icmrp.20 

26. Kavanaugh M, Fisher K, Quinlan JJ. Use of Focus Groups to Identify Food Safety Risks 
for Older Adults in the US. Foods. 2022;11(37). doi:10.3390/foods11010037 

27. Hanson JA, Hughes SM, Liu P. Use of Health Belief Model Variables to Examine Self-
Reported Food Handling Behaviors in a Sample of US Adults Attending a Tailgate Event. 
Journal of Food Protection. 2015;78(12):2177–2183. doi:10.4315/0362-028x.jfp-15-077 

28. Meysenburg R, Albrecht JA, Litchfield R, Ritter-Gooder PK. Food safety knowledge, 
practices and beliefs of primary food preparers in families with young children. A mixed 
methods study. Appetite. 2014;73():121–131. doi:10.1016/j.appet.2013.10.015 

29. Hanson JA, Benedict JA. Use of the Health Belief Model to Examine Older Adults' Food-
Handling Behaviors. Journal of Nutrition Education and Behavior. 2002;34():S25–S30. 
doi:10.1016/s1499-4046(06)60308-4 

https://doi.org/10.26555/eshr.v3i1.3629
https://doi.org/10.26555/eshr.v3i1.3629
http://dx.doi.org/10.21088/ijfnd.2322.0775.5217.3
https://doi.org/10.3390/foods11010037


 
Rustiawan et al. (Health Belief Model Application on Food Safety Behavior of Bantul Beach Tourism Culinary Food Handlers) 

 
 

 
Vol. 5, No. 2, 2023, pp. 69-89 89     10.26555/eshr.v5i2.6996 

  

30. Clayton DA, Griffith CJ, Price P, Peters AC. Food handlers' beliefs and self-reported 
practices. International Journal of Environmental Health Research. 2002;12(1):25–39. 
doi:10.1080/09603120120110031 

31. Sanlier N, Baser F. The Relationship Among Food Safety Knowledge, Attitude, and 
Behavior of Young Turkish Women. Journal of the American College of Nutrition. 
2019;pp:1–11. doi:10.1080/07315724.2019.1639084 

 

https://doi.org/10.26555/eshr.v3i1.3629
https://doi.org/10.26555/eshr.v3i1.3629

