The Influence of Perceived Travel Risk, Celebrity Endorsement, and Destination Image on Attitude and Revisit Intention of Tourists

This study aimed to explain how attitudes and revisit intentions are affected by celebrity endorsement, perceived risk of travel, and perception of the destination. The study's participants are domestic travelers who have been to Bali. 170 domestic tourists were included in the sample using a purposive sampling technique. Path Analysis utilizing SEM-PLS is the analysis method employed. The findings demonstrated that attitudes and intentions to revisit were negatively and significantly impacted by perceived travel risk. While celebrity endorsement has no impact on the intention to return, it has a positive and large impact on attitude. Additionally, views and inclinations to return are positively and significantly impacted by the final picture. The image of a destination is positively and significantly impacted by attitudes. thanks to celebrity support Impact on the intention to revisit is favorable and considerable. Because of this, it's critical for companies in the tourism industry to focus on among order to foster a positive attitude among visitors and increase their intention to return to Bali, perceived travel risk, celebrity endorsements, and destination image are used which is increasing

This study aimed to explain how attitudes and revisit intentions are affected by celebrity endorsement, perceived risk of travel, and perception of the destination. The study's participants are domestic travelers who have been to Bali. 170 domestic tourists were included in the sample using a purposive sampling technique. Path Analysis utilizing SEM-PLS is the analysis method employed. The findings demonstrated that attitudes and intentions to revisit were negatively and significantly impacted by perceived travel risk. While celebrity endorsement has no impact on the intention to return, it has a positive and large impact on attitude. Additionally, views and inclinations to return are positively and significantly impacted by the final picture. The image of a destination is positively and significantly impacted by attitudes. thanks to celebrity support Impact on the intention to revisit is favorable and considerable. Because of this, it's critical for companies in the tourism industry to focus on among order to foster a positive attitude among visitors and increase their intention to return to Bali, perceived travel risk, celebrity endorsements, and destination image are used which is increasing

INTRODUCTION
The Covid-19 The global epidemic has had a significant influence on the world of tourism. The most popular province in Indonesia to become a tourist destination is Bali Province. Tourism in Bali was destroyed when the Covid-19 virus started to enter Indonesia. Since the beginning of the pandemic until now, Bali has lost 54% of the contribution from the tourism sector (Merdeka, 2021). Business activists in the tourism industry experienced a downturn. Until now the Covid-19 virus still haunts tourists, so many tourists have discouraged traveling (Golets et al., 2020). This intention to travel can be influenced by the attitude of vacationers to a location. The way in which tourists act is a relatively consistent description of the evaluations, feelings, and tendencies of tourists towards a destination. When wanting to travel to a destination, tourist attitudes are an important predictor as a motivator for tourists to travel to a destination.
The existence of the Covid-19 virus will affect the attitude of tourists towards a destination. Information regarding transmission, handling, and the exact impact of this virus is not known with certainty. Rao (2020) stated that when a deadly virus appears that was not known before, it will cause a measure of community stress, dread, and risk so that it will change the attitudes and intentions of tourists to travel ( Bae & Chang, 2021;Sukaatmadja et al., 2021) . Several previous studies have also examined the effect of perceived risk on travel intentions (Sánchez-Cañizares, 2021;Joo et al. 2021;Rather, 2021).
The Indonesian government has made efforts to help restore the condition of the tourism industry, especially in Bali. One of the steps the government has taken is to endorse influencers to promote tourism in Bali (CNN Indonesia, 2020). Celebrity involvement has an influence on affective image felt in generation Y thereby increasing their intention to visit destinations (Chan et al., 2018). Other research also shows that celebrity endorsements have a significant influence on the intention to travel to a destination (Kumari, 2018;Xu and Pratt, 2018). In addition, celebrity endorsements also have a Influence on attitudes was considerable (Vidyanata et al., 2018;Parwati et al., 2021). The government hopes that by taking promotional steps in the form of celebrity endorsements , it can increase the number of tourists visiting Bali.
Until now, Bali is the most popular tourist destination in Indonesia. The good image that Bali has is already attached to the minds of tourists (Utama, 2017). Pereira et al. (2019) and Kim and Stepchenkova (2015) discovered a strong effect of destination image on visitor perceptions. towards destinations. Destination image is also one of the factors that can increase tourist intentions to travel to a destination (Chaulagain, 2019;Ahmad et al ., 2020). If a destination has a good image, tourists will want to return to that destination In 2021, Yang et al. The research goal was to investigate and explain the impact of perceived travel risk, celebrity endorsement, and destination image on attitudes and revisit intention depending on the circumstances of the current circumstance.

LITERATURE REVIEW Research Concept Framework
Perceived travel risk as a result of the coronavirus outbreak that has hit the world can be one of the factors influencing tourists' revisit intention to a destination. This is supported by several studies such as research conducted by ), Neuburger & Egger , (2021, and Sukaatmadja et al. (2021) this study shows a significant positive link between the intention to return and perceived travel danger. It was revealed that perceptions of travel danger have a significant impact on tourists' attitudes towards destinations (Bae & Chang, 2021;Sukaatmadja et al., 2021).
Celebrity endorsement is one of the marketing strategies used to influence tourist behavior. Several studies have shown that celebrity endorsements also affect revisit intention (Osei-Frimpong et al., 2019;Chan et al., 2018). According to Kim and Stepchenkova (2015) and Pereira et al. (2019), celebrity endorsements can greatly alter consumers' perceptions of a product. In addition, celebrity endorsements are also able to strengthen the image of a destination (Malik et al., 2018;Chan et al., 2018).
A strong destination image in a destination will make tourists interested in visiting that destination. Research conducted by Tourist sentiments regarding locations are significantly impacted by destination image, according to Kim and Stepchenkova (2015) and Pereira et al. (2019). Additionally, studies by Chan et al. (2018), Liu et al. (2018), and Ahmad et al. (2021) demonstrate how the perception of a place affects not only the tourists' subjective perceptions but also how they rate their following travel experiences and how they plan to travel in the future. Therefore, attitudes will influence intentions to return, while attitudes will also effect perceived travel risk, celebrity endorsement, and destination image (Darmawan and Iriani, 2021;Nurliasari et al., 2021;Sukaatmadja et al., 2021); so that in research it analyzes the model as shown in Figure

Research Hypothesis
The research hypotheses that can be put together based on the conceptual framework that is now in place are as follows. The effect of perceived travel risk on the revisit intention perceived travel risk can reduce tourists' revisit intention (Carvalho, 2022). This is also in line with the results of research conducted by several other researchers, such as: According to Hasan et al. (2018), ), Hanafiah et al. (2021), Sukaatamdja et al. (2022), and Sengel et al. (2022, the thought of returning has a negative and significant effect on how risky traveling is viewed. The ensuing supposition can be derived from some of the findings of current empirical studies: arranged.

H1 : Perceived travel risk has a negative effect on the revisit intention The influence of celebrity endorsement on revisit intention
According to the research results from Bauer et al. (2022), stated that celebrity endorsement had a favorable and strong impact on the inclination to revisit. In other words, tourists are more likely to want to return to a place if a celebrity endorses it well. The findings are consistent with those presented by a number of scholars, including Gong (2021); Foroudi et al. (2021); The following hypotheses can be tested using data from some of the available empirical studies: developed.

H2 : Celebrity endorsement has a positive effect on the revisit intention Effect of destination image on the revisit intention
Destination image is the perception of tourists about a destination, both in terms of positive and negative images (Abbasi et al., 2021). Destination images that include cognitive destination images and effective destination images can increase tourists' revisit intention (Hu and Shen, 2021). Furthermore, there are other studies which show that destination image can Destination image is the perception of tourists about a destination, both in terms of positive and negative images (Abbasi et al., 2021). Destination images that include cognitive destination images and effective destination images can increase tourists' revisit intention ( 2022), there are numerous other researchers.; , stated that the better the destination image of a tourist spot, the higher the intention of tourists to visit again. Based on the empirical research, the theory can be constructed as follows.

H3 : Destination image has a positive effect on revisit intention The effect of perceived travel risk on tourist attitudes
According to Rather's research (2021), states that perceived travel risk can make tourists' attitudes negative. The perception of a travel risk is losses that a tourist might experience when visiting a destination, such as the risk of illness, risk of discomfort, security risk. Perceived travel risk is also a burden on tourists' minds so that it can have a negative impact on their attitudes. All of this resulted in a negative attitude of tourists. Other research also states that tourists' perceptions of travel risk are negatively and significantly impacted (Hasan et al., 2018). The same thing was also shown by other researchers such as: Rahmafitria et al. (2021); Isaac et al. (2021); Sujood et al. (2022); Lai et al. (2022), which demonstrates the study's findings that unfavorable tourist attitudes increase in proportion to perceived travel danger. The following theories can be supported by many empirical research that are now available developed .

H4 : Perceived travel risk has a negative effect on tourist attitudes
The influence of celebrity endorsements on tourist attitudes Celebrity endorsements used to promote a destination can have an impact on tourist attitudes. The better the celebrity endorsement used, the positive attitude of tourists will (Parwati et al., 2021) Increase. Other studies have confirmed the findings of this study, including Centeno and Wang (2020) and Khare et al. (2021), which demonstrate that celebrity endorsements have a favorable and significant impact on opinions. The following assumptions can be ordered in accordance with available empirical studies.  2021), shows the results that destination image has a favorable and important impact on the views of tourists. The following assumptions can be ordered in accordance with available empirical studies. H6 : Destination image has a positive effect on tourist attitudes The influence of celebrity endorsement on destination image Celebrity endorsement is a person who is given the task of conveying information about a destination so that the destination has a good image . Celebrity endorsements had a good and considerable impact on destination image, according to the findings of multiple subsequent studies (Yang, 2018;Zhang et al., 2020;Li et al., 2021). The following assumptions can be ordered in accordance with available empirical studies.

H7 : Celebrity endorsement has a positive effect on destination image The influence of tourist attitudes on the revisit intention
According to research by Hasan et al. (2018), stated that the attitude of tourists can increase the revisit intention. Tourist attitudes reflect tourists' feelings based on views of a tourism product and destination features, towards a vacation location and destination tourism services. The tendency of visitors to behave in a closed manner in reaction to particular stimuli or items is known as their attitude. All of this motivates visitors to have higher intentions of returning. According to additional research, visitor attitudes have a favorable and considerable impact on the propensity to return (Wang et al., 2022).  2022), which demonstrates the findings of research that the higher a tourist's intention to return to a location, the more positive or better their attitude toward the location. Based The following hypotheses can be established based on several existing empirical investigations.

METHODOLOGY
Given the severity of the issue, this study employs a quantitative methodology, specifically looking at how attitudes and intentions to travel again are affected by perceptions of travel risk, celebrity endorsements, and destination image. The only participants in this study were domestic travelers who had been to Bali. According to the formula (5 -10 x n), the sample size for this study was calculated using 17 indicators, which resulted in a sample size of 170 respondents.
Using a five-level Likert scale, from strongly disagree = 1, disagree = 2, quite agree = 3, agree = 4, and strongly agree = 5, perceptions of domestic tourists who have visited Bali are used to measure variables such as perceived travel risk, celebrity endorsements, destination image, attitudes, and intention to return. This study makes use of primary data, or information that respondents-in this case, tourists who have visited Bali-provided in response to surveys. Data collection methods involved handing out questionnaires to visitors both inperson and online (through a Google form).
To be able to measure what they wished to measure and to know the consistency of the replies provided by the respondents, the validity and reliability of the deployed instruments were assessed. The Product Moment correlation technique from Pearson, with a minimum limit of r = 0.3, is used to test the validity of the instrument (Sugiyono, 2018: 150). By determining the Cronbach's Alpha reliability coefficient with a minimum limit of Alpha coefficient > 0.6, instrument reliability testing was conducted (Sekaran, 2003: 312). The validity test findings demonstrate that all variable indicator results are valid because r count > 0.3, and the reliability test results demonstrate that all variables are reliable since Cronbach Alpha values are all greater than 0.6. In addition, hypothesis testing was completed utilizing inferential analysis using the SPSS 24.0 computer program and the Structural Equation Modeling (SEM) method of partial least squares (PLS).

Characteristics of Respondents
In this study, the respondents' characteristics were determined by their gender, age, status, final educational attainment, employment, and In this study, the respondents' characteristics were determined by their gender, age, status, final educational attainment, employment, and income. Table 1 displays the breakdown of the traits of the research participants. 100.00 Table 1 shows the characteristics of responses by the sex of travelers. Between 52.35% and 47.65% of responders were male and female, respectively. In terms of age, respondents between the ages of 21 and 30 made up the majority (44.71%). According to status, tourists made up the majority of the study's respondents, with 98 having an unmarried status (a ratio of 57.65%) and 72 having a married status. According to their most recent education, travelers with bachelor's degrees made up the majority of research participants-93 persons, or 54.71 percent-of the total respondents.
According to their type of employment, respondents to this study were predominately tourists with jobs as private employees, numbering a total of 54, with a percentage of 31.76 percent. According to their total income, respondents to this study were predominately tourists, numbering 80, with a percentage of 47.06 percent. The least were respondents, numbering as few as 7, with a percentage of Rp. 2.5 million.

PLS SEM Analysis Results
Testing of the structural model or inner model was carried out to see the relationship between the constructs, the significance value, and the R-square of the research model. Testing the inner model is carried out using the R-square of the dependent latent variable with the same interpretation as the regression and the construct model's Q-square of predictive significance, which assesses how effectively the model and its parameter estimations reproduce the observed values. The model is more predictively relevant if the Q-square value is greater than zero; alternatively, if the Q-Testing of the structural model or inner model was carried out to see the relationship between the constructs, the significance value, and the R-square of the research model. Testing the inner model is carried out using the R-square of the dependent latent variable with the same interpretation as the regression and the construct model's Q-square of predictive significance, which assesses how effectively the model and its parameter estimations reproduce the observed values. The model is more predictively relevant if the Q-square value is greater than zero; alternatively, if the Q-square value is lower than zero. If the data is distribution-free, the Partial Least Square (PLS) predictive approach's structural model is assessed using R-square for the dependent construct and Q-square for predictive relevance.

R-square
The goodness of fit of the model is assessed using the R-square for endogenous latent variables. The R-square measures how well the model and parameter estimates replicate the observed values in structural models. The worth of the R-square's magnitude ranges from 0 to R2 to 1. R2 values that are coming closer to 1 show that the model is improving (significantly). The indicators for each variable can be thought of as tools for measuring latent variables if the test results on the inner model are significant. Table 2 displays the findings of the R-square test.  Table 2, For the final destination's picture, the R-square value for the celebrity endorsement variable is calculated. 0.580. This shows that celebrity endorsement has a contribution of 58.00 % to explain the destination image variable. 58.0% R-square (R2) is a reasonable number for a model. i.e., the celebrity endorsement variable can account for 58.00% of the destination image variable's variance, with the remaining 42.00% of the variance being explained by variables outside the model.. R-square values for celebrity endorsements, destination perception, and traveler attitude of 62.50%. This shows that perceived travel risk, celebrity endorsements, Destination image has an influence of 62.50 % on tourist attitudes towards Bali destinations. Attitude R-square (R²) value 62.50% is a good model , meaning that The factors celebrity endorsements, perceived travel danger, and destination image are having the capacity to account for 62.50% of the variation in the attitude variable, with the remainder 37.50% being explained by factors outside the model. R's square for perceived risk also travel risk, celebrity endorsements, destination image , and attitude towards revisit intention by 79.50%. This shows that perceived travel risk, celebrity endorsements, destination image , and attitudes have an influence of 79.50 % on the revisit intention tourists to Bali. The R-square value (R²) of revisit intention is 79.50%, including a very good model , meaning that the variable celebrity endorsements, destination image, perceived travel danger, and attitudes can all influence account for 79.50% of the revisit intention variable, leaving 21.50% of the explanation to variables outside the model.

Q-square
The Q-square value, a goodness of fit model test, can also be used to perform inner model testing. If the Q-square value is larger than 0, the model's predictive relevance is stronger; if it is lower, the model's predictive relevance is less. Following are the steps to calculate Q-square: The equations above result in a Q-square value of 0.968. As a result, The model warrants being described as having relevant predictive value since it has predictive relevance value. The Quadrangle value is greater than 0. According to the test's findings, the structural model is rated as very good. In other words, exogenous construct variations can account for 96.8% of the endogenous constructs, but 3.2% of the endogenous constructs are still not fully explained by the model.

Hypothesis Testing Results
Using the path coefficients menu, hypothesis testing is performed to calculate the likelihood of a data set. A research hypothesis is considered to be supported by rule of thumb if the coefficient or direction of the variable relationship indicated by the original sample value agrees with the hypothesis and the t-statistic value is greater than 1.645 (one tailed) or greater than 1.960 (two tailed) or the p-value is less than 0.05, which is considered to be significant. This indicates that one latent variable significantly affects another hidden variable.

Results of the Direct Effect Test
Using Partial Least Square (PLS), five direct effect hypotheses are evaluated. On each path of impact between the variables, the t-test (t-test) is used to conduct this test. The bootstrap approach is used in PLS to statistically test each postulated association on the sample. The goal of bootstrap testing is to reduce the issue of anomalous research data. Table 3 and Figure 2 below present the findings of the direct effect test on the structural equation analysis. This indicates that the intention to revisit is negatively impacted by perceived travel risk. The p-value of 0.010, which is less than 0.05, shows that perceived travel risk significantly influences the intention to return. As a result, hypothesis 1 (H1), according to which perceived travel risk has a detrimental and significant impact on the intention to return, is accepted. 2. Determining the regression coefficient value of 0.120 by testing the hypothesis regarding the impact of celebrity endorsements on revisit intention. This indicates that the celebrity endorsement has a favorable impact on the intention to revisit. The celebrity endorsements' p-value of 0.205, which is higher than 0.05, suggests that they have no discernible impact on consumers' intentions to revisit. Consequently, the second hypothesis (H2), which claims that celebrity endorsements have a favorable and large impact on the desire to revisit, is true rejected. 3. A regression coefficient value of 0.474 is obtained when looking into the potential influence of destination image on revisit intention. This indicates that the perception of the destination influences the intention to return. The p-value of 0.000, which is less than 0.05, shows that the perception of the destination significantly influences the intention to return. Thus, hypothesis 3 (H3), according to which the intention to return is positively and significantly influenced by the destination image, is approved. 4. A regression coefficient value of -0.148 is obtained when the hypothesis regarding the influence of perceived travel risk on attitudes is tested. This indicates that views are positively impacted by perceived travel risk. The pvalue of 0.011, which is less than 0.05, shows that attitudes are significantly impacted by perceived travel risk. Thus, hypothesis 4 (H4), which claims that accepted attitudes are positively and significantly influenced by perceived travel danger.Testing the hypothesis on the influence of celebrity endorsements on attitudes produces a regression coefficient value of 0.365. This means that celebrity endorsements have a positive effect on attitudes. The p-value of 0.003 which is smaller than 0.05 indicates that celebrity endorsements have a significant effect on attitudes. Thus, hypothesis 5 (H5) which states that celebrity endorsements have a positive and significant effect on attitude is accepted. 5. A regression coefficient value of 0.392 is obtained when the effect of destination image on attitudes is tested. This indicates that sentiments are positively impacted by the destination's reputation. Having a substantial impact on attitudes is the destination image, as shown by the p-value of 0.002, which is less than 0.05. As a result, hypothesis 6 (H6), which claims that attitude is positively and significantly influenced by destination picture, is accepted.Testing the hypothesis on the effect of celebrity endorsements on the destination image again produces a regression coefficient value of 0.762. This means that celebrity endorsement has a positive effect on the image of the destination. The p-value of 0.000 which is smaller than 0.05 indicates that celebrity endorsement has a significant effect on destination image. Thus, hypothesis 7 (H7) which states that celebrity endorsement has a positive and significant effect on destination image is accepted. 6. A regression coefficient value of 0.297 is obtained when the hypothesis regarding the influence of attitudes on revisit intention is tested. This indicates that the intention to revisit is positively impacted by attitude. The p-value of 0.003, which is less than 0.05, shows that attitude significantly influences the intention to revisit. As a result, hypothesis 8 (H8) is accepted, which implies that attitude has a favorable and significant impact on revisit intention.

Research Implications
According to the analysis and discussion, there may be a connection between the latent variables celebrity endorsements, destination image, attitudes, and intent to return in this study on perceived travel risk study.This study can provide management of businesses involved in the tourism industry with ideas. The idea being debated is that the business consistently creates plans to raise celebrity endorsements, destination image and reduce perceived travel risk , as well as create a good or positive attitude for tourists who have visited Bali so as to be able to boost the likelihood that domestic visitors will return to Bali.

CONCLUSION AND RECOMMENDATION
The following conclusions can be taken from the study, analysis, and interpretation outlined in the previous chapter, as well as from a number of theories and the findings of earlier studies. 1. Perceived travel risk has a negative and significant effect on the revisit intention domestic travelers' propensity to return to Bali increases in proportion to their perception of the risk of their trip Bali decreases. 2. Celebrity endorsement has a positive and not significant effect on the revisit intention for domestic tourists to Bali, meaning that the better the celebrity endorsement felt by tourists, it is not able to increase the revisit intention tourists to Bali. 3. Destination image has a positive and significant effect on the revisit intention domestic tourists to Bali, meaning that the better the image of the destination perceived by tourists, the higher the intention of tourists to return to Bali. 4. Perceived travel risk has a negative and considerable impact on attitudes, which means that the less favorably foreign visitors view Bali, the higher the perceived travel danger they feel. 5. Celebrity endorsement has a positive and significant impact on attitudes, which means that the more favorable the perception of Bali among domestic tourists is influenced by celebrity endorsement. 6. The opinion of a destination has a favorable and considerable impact on attitudes; hence, the more positively local visitors view Bali, the more positively foreign tourists view Bali. 7. Celebrity endorsement has a positive and significant effect on the image of the destination, meaning that the better the celebrity endorsement felt by domestic tourists, the better the image of the Bali destination. 8. Attitude has a favorable and considerable impact on domestic tourists' intention to return to Bali, which means that the better their attitude about Bali, the higher their want to travel there.

FURTHER STUDY
This research still has limitations, so it is necessary to conduct research related to the topic of The Influence of Brand Image, Social Media Advertisement, and Word of Mouth Toward Customer Attraction in order to perfect this research and add insight for readers.
Zhang, Y. , Kim, E. and Xing, Z. (2021), Image congruence between sports event and host city and its impact on attitude and behavior intention, International Journal of Sports Marketing and Sponsorship , 22 (1), pp. 67-86.
Zhou, Q. , Pu, Y. and Su, C. (2022), The mediating roles of memorable tourism experiences and destination image in the correlation between cultural heritage rejuvenation experience quality and revisiting intention, Asia Pacific Journal of Marketing and Logistics , https ://doi.org/10.1108/APJML-11-2021-0829