What are the Factors affecting Tourist Behavior based on the Perception of Risk? Romanian and Serbian Tourists’ Perspective in the Aftermath of the recent Floods and Wildfires in Greece

Abstract

During the summer season 2018 and 2019, natural hazards (namely, floods and wildfires) have occurred at some of the most popular tourist destinations in Greece, endangering tourists’ safety and vacation quality. These have influenced tourists’ perception of Greece as a “potentially risky destination”, causing a decrease in the number of tourists willing to visit this country. However, the current study assumes that some tourists will show more courageous travel behavior in the aftermath of natural hazards, while others will remain cautious when deciding to travel to risky destinations. Therefore, the questionnaire on a sample group of 431 respondents from Serbia and Romania was conducted to explore the factors influencing such differences. The study aims to explore whether tourists’ individual characteristics influence tourism worries and tourist behavior based on perceived risks. The study also intends to analyze the moderating role of tourists’ nationality, considering the relationship between personality and tourist behavior based on the perceived risk. The findings provide evidence that not only tourists’ personality but also sociodemographic characteristics influence tourism worries and tourist behavior based on the perceived risk. Additionally, the study is the first to explore and confirm the role of nationality in tourist behavior based on the perceived risk, as well as the moderation role of nationality in regression between tourist’s personality and behavior based on the perception of risk. Theoretical and practical implications are discussed in the paper.

1. Introduction

The effects of natural hazards are felt every year in the entire human society, and their impact is increasing [1]. Besides the loss of life and economic damages, natural hazards have constantly forced the human society to develop adaptive strategies to improve the theory and practice of risk mitigation and management [2,3,4,5,6]. Rising the level of resilience is based on the accumulation and investigation of seemingly chaotic events that take place year after year [7] in an increasingly technological and digitalized but also in a much more communicative and sensitive society [8,9,10,11].

From global to local scale, the impact of natural hazards is a complex function that mainly includes the magnitude and frequency of the hazard and the vulnerability of the affected communities. The new “era of catastrophes” [12] is shaped by new and concatenated risks [13,14,15] in our unprecedented dynamic and interconnected world. The palette of the current crisis generated by COVID-19 pandemic is completed daily with new challenges whose complexity has never been encountered by humanity: the economic costs, as well as the heavy losses of human lives [16,17,18,19] are overlapped with major dysfunctions of health systems [20,21], transnational labor migration crisis [22,23], social stigma and racism [24,25], or geopolitical reassessments [26,27]. Even certain beneficial environmental impacts of the quasi-generalized outbreak of the 2020 spring have already been recorded [28,29,30], the governments will be faced with potential intersections and collisions of the current pandemic with other regional and global crises, generated especially by climatic and geopolitical stressors [31,32,33]. Besides the stated challenges caused by pandemics, tourism certainly is one of the most hit economic fields. This is especially due to measures that the majority of countries applied to fight pandemics such as restricted mobility and social distancing [34].

Although the current crisis will change the world we know today [35] and the economic impact will be felt in the next years, few people will give up tourism activities. Alongside the aviation and hospitality industries, the tourism sector is suffering major consequences caused by the COVID-19 pandemic [36]. Some early projections from UNWTO (World Tourism Organization) for 2020 suggest international arrivals could decline by 20–30% relative to 2019. This sector, recognized as one of the most vulnerable to risks [37] must adopt rapid, flexible, structural, and functional measures to reduce the impact of the 2020 outbreak and to switch the actual state into a transformational opportunity [38]. In the present threatening mosaic of hazards, tourism organizations and business will be forced to change their offer and to include as many details as possible about the destination places. These improvements should also take into account all the particularities of potential tourists, their behavior, and pretensions. At the same time, it is certain that significant changes will occur in post pandemic travel planning [39] and tourists’ behavior [40], especially those regarding the way in which they will evaluate and perceive their exposure to the different risks induced by natural hazards. People will become more aware and will think twice during the complex process of choosing their tourist destinations [41]. Moreover, the global study of the effects of natural disasters on international tourism [42] provides evidence that the occurrence of different types of natural hazards changes tourist flows to varying degrees, depending on the type of hazard. In general, the impacts are negative, resulting in reduced tourist arrivals following an event. The study also indicates that disasters such as floods and storms have smaller and shorter-term impacts compared to some larger disasters such as volcano eruptions. They also discuss that it is still not clear whether the negative effect is due to the possible destructive effects on infrastructure or to the negative image of the destination caused by these occurrences.

In this context, the studies regarding the knowledge of risk perception associated with natural hazards are becoming more and more important. Additionally, as stated by Wong and Yeh [43], not all tourists react the same in case they feel that risks are too high. They state that some of them will delay, postpone, or avoid risky behavior, while others will be ready to travel to risky destinations. Thus, the main aim of the current study is to analyze who are the tourists who will show more courageous behavior and who are those who will be more cautious when deciding whether to travel to risky destinations. The study context in which this is investigated is wildfires and floods, which have hit Greece in 2018 and 2019. A classic destination for summer vacation, the Greek territory is also susceptible to various natural hazards due to the morphotectonic, topographic, and climatic conditions. Sudden manifestations of these hazards, such as extreme meteorological events and induced hazards (e.g., floods and wildfires) can generate human losses and economic damages, but also severe disruptions of social activities. Additionally, to their local impact, these events are felt internationally in the current state of the strong sensitivity of people towards extreme events due to their broad media coverage and accessibility to online information.

The main assumption of the study is that tourist behavior based on the perceived risk of natural hazards is related to tourists’ individual characteristics and nationality. In light of this, the main purpose of the study is to explore whether tourists’ individual characteristics (personality and sociodemographic characteristics) influence tourism worries and tourists’ behavior based on the perceived risk. This study also indents to analyze the moderating role of tourists’ nationality (Romanian and Serbian tourists), considering the relationship between personality and tourists’ behavior based on the perceived risk. Previous research on the relationship between sociodemographics and personality has been rare concerning tourist behavior in the face of the perceived risk, whereas a focus on the moderating role of nationality has not been explored yet. Thus, the current study aims at contributing to a better understanding of such complex relationships while answering the call opened by Kovačić et al. [44] for additional research needed on this topic.

2. Theoretical Background and Hypotheses

2.1. Tourists Personality and Risk Perception

Tourist personality is largely related not only to the way tourists perceive the risk but also the way they behave and respond to it [44,45,46,47,48,49]. Tourists’ risk-seeking and novelty-seeking also inspired some of the earliest tourist typologies such as those of Cohen [45] and Plog [50]. Cohen [45] based his typology on the belief that novelty-seeking travelers will be more ready to visit risky destinations. Plog [50] differentiated three types of tourists (allocentric, psychocentric, and midcentric) based on the level of their anxiety about travel and readiness to visit novel and risky places. According to Plog, allocentric tourists perceive destinations as less risky, and consequently, they are more ready to visit such destinations. They prefer to travel to less exploredplaces, feeling less anxiety about travel compared to others [51].

Novelty-seeking and sensation-seeking traits are often connected with tourists who are less sensitive to risk and more ready to visit risky destinations [46,47,48,52,53,54]. For instance, in the Lepp and Gibson’s study, risk factors at tourism destinations were perceived as less risky by novelty-seekers [46], while both novelty-seekers and sensation-seekers were more ready to visit risky destinations. Reisinger and Mavondo [48] also argued that novelty-seekers will be less anxious about travel, more ready to visit an unfamiliar environment, and will perceive it as less risky compared to tourists seeking familiarity. In connection to this, travel anxiety and travel worry are factors that are largely connected with risk perception and desire to visit risky destinations [55]. Lepp and Gibson’s (2008) study [53] showed that sensation-seekers are more ready to visit risky destinations. Pizam et al. [47] also suggested that risky behavior and sensation-seeking are positively correlated. Contrary to these findings, the study by Kovačić et al. [44] have found no significantrelationshipbetweensensation-seekingandtourists’behaviorbased on the perceived travel risk. Thus, although sensation-seeking is related to risk perception and intention to visit risky destination in the majority of previous studies, Kapuściński and Richards [49] claim that relationship between sensation-seeking and risk perception in tourism is still underexplored area, so further research should help in better understanding of these concepts. Based on these findings, two hypotheses can be drawn:

Hypothesis 1 (H1): 

Novelty-seeking will positively influence Courageous behavior, but negatively Cautious behavior.

Hypothesis 2 (H2): 

Sensation-seeking will positively influence Courageous behavior, but negatively Cautious behavior.

When it comes to other personality traits, outside the tourism field, Cooper et al. [56] revealed that neurotic individuals and extraverts are more ready to engage in risky behaviors. Furthermore, the study by Weller and Tikir [57] explored risk-taking at specific areas (health, recreation, social, and ethical) and tourist personality measured by HEXACO (Extraversion, Emotionality, Agreeableness, Consciousness, and Openness to experience, and Honesty-Humility). This study revealed that in the all explored area, neurotic individuals perceive more risk and are more ready to take a risk. Open individuals are more ready to take risks related to recreation or social activities, whereas those with lower honesty/humility are more ready to engage in risks related to health and ethics, but tend to perceive things and situations as less risky. However, honesty-humility did not show effect on perception or risk engagement in the field of recreation or social activities, which are closely linked to tourism and travelling. Conscientiousness also showed to negatively influence risk perception in all explored areas. In the study of Lee et al. [58], extraverts showed to be more ready to take risk, while agreeableness was correlated with risk perception.

Generally, personality traits and risk-taking were rarely explored in the tourism field. Some of them were studies by Valencia and Crouch [59] who explored self-confidence and by Reisinger and Mavondo [48] who explored extroversion, activity, confidence, helpfulness, and adventurousness in relationship with travel risk perception. Those studies revealed that individuals who are confident, extroverted, and adventurous tend to show greater intention to engage in risk. The study by Kovačić et al. [44] explored the role of personality traits and tourism worries on the tourist behavior based on the perceived risk of affected destinations, on the case study of natural hazards in Greece. The mentioned study revealed that Novelty-seeking negatively affects Courageous behavior, Extraversion and Agreeableness positively affect Conscious behavior, while Conscientiousness positively affects Cautious behavior. Their model also showed that Tourism worries mediate the relationship between certain personality traits (such as Honesty-Humility, Neuroticism, and Openness to experience) and Cautious behavior. Based on the analyzed literature, six hypotheses can be drawn:

Hypothesis 3 (H3): 

Extraversion is positively related to Courageous behavior.

Hypothesis 4 (H4): 

Neuroticism is negatively related to Courageous behavior, but positively to Cautious behavior and Tourism worries.

Hypothesis 5 (H5): 

Conscientiousness is negatively related to Courageous behavior, but positively to Cautious behavior and Tourism worries.

Hypothesis 6 (H6): 

Agreeableness is positively related to Conscious behavior.

Hypothesis 7 (H7): 

Openness to experience is positively associated with Courageous behavior, but negatively to Cautious behavior.

Hypothesis 8 (H8): 

Higher Honesty-Humility is positively related to Cautious behavior.

2.2. Sociodemographic Characteristics and Risk Perception

Travel risk could be differently perceived by tourists of different sociodemographic characteristics (gender, age, education, marital status, nationality, culture, etc.) and they can also differ in their reactions to the perceived travel risk. Many previous studies indicate that people of different sociodemographic characteristics differ in their risk perceptions and risky intentions [46,48,60,61,62,63,64].

The current study represents the extension of the study by Kovačić et al. [44] aimed to test whether risk perception and influence of tourist personality on risk perception and related behavior depends on tourists’ nationality. For this purpose, Romanian and Serbian tourists, as some of the most dominant tourists’ groups visiting Greece, were compared. Some previous studies have shown that different nationalities could have different perception of risk, but can also differ in their desire to visit risky destinations [48,65,66,67,68,69,70,71].

In tourism field, Reisinger and Mavondo [48] found that people from different countries differ regarding travel anxiety and readiness to take risks. The authors claim that risk perception and related behavior is connected with uncertainty avoidances, which is more pronounced in certain cultures. This was also confirmed by the study of Kozak et al. [60]. Similarly, Fuchs and Reichel [72] revealed significant differences in the overall tourism risk perception among tourists of various nationalities. The study by Arias-Febles [73] also indicates that tourists of different nationality differ in travel risk perception. In addition, study by Simpson and Siguaw [61] revealed significant country and ethnic differences in different types of the perceived risk, while Slovic et al. [66], for instance, showed that people from different cultures differ in their perception of health risks. Moreover, the study by Fourie et al. [74] showed that levels of safety and security in tourists’ origin country can influence their destination risk perception. Specifically, they have found that tourists from stable countries prefer traveling to countries with the same conditions, while tourists from unstable countries are more tolerant with insecurity at the destination country.

An interesting research was also conducted by Kozak et al. [60], which involved international tourists from 14 countries, showing that there are differences in risk perception and decision to travel based on tourist nationality. Taking into account cultural model by Hofstede [75], the authors concluded that uncertainty avoidance (cultural dimension) affects differences in risk perception. Additionally, the authors consider that tourists’ personality should be researched in order to support and complement these the findings. The study by Seabra et al. [71] provided the first research-based segmentation concerning tourists’ perception of risks. Their results indicate that tourists coming from different countries belong to different groups based on risk perceptions regarding international travel. Based on the previous findings, we can assume that:

Hypothesis 9 (H9): 

Romanian and Serbian tourists will differ in their behavior based on the perceived risk.

Hypothesis 10 (H10): 

Nationality will moderate relationship between personality traits and behavior based on the perceived risk.

When it comes to gender differences, previous research shows that women are more likely to avoid travel risk and are more concerned about it. Regarding purchasing tourists’ services, the study of Maciejewski [76] has found that women had a higher perception of risk. Mattila et al. [62] found males and females differ in health risk-taking at travel destinations, specifically confirming that male tourists are more ready to engage with such behaviors. The studies of Granger et al. [77] and Granger and Hayne [78] show that women are more worried and afraid by natural hazards than males, which was also confirmed in the study of Armaş [63] about risk perception of seismic activity in Bucharest. In addition, studies by Lepp and Gibson [46] and Staats et al. [79] indicate that women report more risk aversion than men. Kim and Seo [80], throughout three studies they have conducted, also found that men consistently displayed a greater preference for the choice of high-risk travel activities, adventurous destination choices, and adventurous appeals in travel advertising. Some previous studies have also found that women may be more susceptible to anxiety than men [81], more worried about travel risks than are men (e.g., [46]), but also more ready to confess when they feel fear than man [82]. Based on this, the following hypothesis can be drawn:

Hypothesis 11 (H11): 

Women will score higher on tourism worries and will show more Cautious behavior than Man.

Kolb’s [83] claimed that age, but also the level of information respondents have, can affect risk perception and related behavior. His study reveals that older individuals may have more tendencies towards risk-taking than younger individuals, depending on their learning experiences. Gibson and Yannakis [84] report that younger individuals perceive more risk than older individuals. The study by Simpson and Siguaw [61] indicates that younger respondents showed much more concern about various risks than older age groups. The study by Williams and Baláž [64] highlights that the 18–45 years age group of travelers were more likely to avoid travelling to risky destinations, especially when risk is related to the weather. They were generally less confident about their general competences to deal with risks and solving travel problems. On the other hand, travelers aged 35–65 years, tend to be more willing to take risks, more likely to tolerate uncertainty, and less likely to be deterred by particular hazards. Sonmez and Graefe’s [85] also revealed that elderly people are more likely to go on a planned trip even if the risky situation happens, while similarly, Kozak et al. [60] found that all tourists who wish not to change their destination due to the existing risks are males and elderly. Based on this, the following hypothesis can be drawn:

Hypothesis 12 (H12): 

Younger respondents will show more tourism worries and will show more Cautious behavior and less Courageous behavior.

Kolb’s [83] argued that individuals with higher education may be more aware of risks and possible consequences because they obtain more information and related experience. However, although they might be more aware of the risk, more educated individuals tend to show more willingness to visit destinations affected by natural hazards in the past. Williams and Baláž [64] argue that this is because highly educated people are more able to manage risks, having wider experience, which diminishes worries and fears about uncertainty. Armaş [63] also indicates that the education level is closely connected to the awareness of the natural hazards. Her study revealed that 63% of highly educated people pay no attention to this matter. According to this study, those with an average education have the greatest tendency to avoid risks. Wang et al. [86] have conducted a cluster analyses, which identified three tourists’ groups regarding risk perception attitudes and safety behaviors: vulnerable, cautious, and adventurous tourists. In their study, cautious tourists appeared to be less educated, while adventurous travelers were better educated individuals. Contrary to this, in terms of consumer purchasing decisions, the study of Maciejewski [87] showed that people with basic education, more often take risk when purchasing tourist services.

In the study by Thapa et al. [88], the majority of Cautious travelers have finished high school, while the smallest number were postgraduates. On the other hand, the majority of Courageous travelers were highly educated. In addition, a recent study [89] showed that individuals with a higher level of subjective knowledge of a destination had a higher perception of destination cognitive and affective image, which positively influenced their decision to visit a country considered risky. Based on this, we can assume:

Hypothesis 13 (H13): 

Less-educated respondents will show more tourism worries and will show more Cautious behavior and less Courageous behavior.

3. Materials and Methods

3.1. 2018 and 2019 Floods and Wildfires Events in Greece

Located in Southern Europe in a dynamic morphotectonic assemblage [90] with a Mediterranean climate, the Aegean and Continental Greece is one of the most affected regions by natural hazards in Europe. Considered a characteristic case of a multi-hazard environment [91], this territory has a long history of devastating volcano eruptions [92,93], tsunamis [94,95], and earthquakes [96,97]. Although geological hazards and those induced by them have a random frequency, those induced by climate are manifested every year, either as a permanent stressor such as heatwaves, droughts, heavy rains, and storms [98,99,100] or as sudden events: floods, wildfires, and landslides [101,102,103,104,105].

The summers of 2018 and 2019 were characterized by one of the most destructive wildfires and floods in the last decades. In June 2018, after intense precipitation (more than 110 mm in 24 h) parts of Attica, Thessaly, and Central Macedonia were hit by “Storm Nefeli” when 150 buildings were affected [106]. In July 2018, Attica region was struck by several fast-moving fires and 126 people have died and 1500 buildings and vehicles were destroyed [107].

In July 2019, the island of Evia was hit by wildfires and people from five villages were evacuated [108]. Severe damages were reported on 13–14 July 2019 in western Greece: flood water and debris blocked the roads, 7 people were killed, and over 100 were injured by violent winds in the north of the country [109]. From this synthetic list of events, the fires from July 2018 had probably the strongest impact on local people and tourists. “apocalyptic” pictures with fires surrounding beaches were rapidly spread in newspapers and social media [108]. Their effects were felt far beyond those unfortunate places.

3.2. Participants

The sample consisted of a total of 431 participants, 224 of which were Serbian and 207 Romanian tourists. The sample was convenient, based on the respondents’ willingness to participate (with condition they are over 18 years old and from Serbia or Romania). Moreover, a snowball convenience sample technique was also applied, as respondents were asked to share the questionnaire with their friends from Serbia and Romania. Researchers tended to collect similar number of respondents from Serbia and Romania, in order to enable reliable comparison between the groups, and still the satisfactory sample size to perform the required statistical analysis. The majority of Romanian tourists come from North-East Romania where Iaşi, Bacău, and Suceava International Airports provide a good connection with charters to Greece in summers. Overall, there were more female (67.1%) than male (32.9%) respondents in the sample. Although the questionnaire was sent to the similar number of men and women, women have shown higher responsive rate, which is quite usual for quantitative studies [110].

The average age of the sample was 31.91 (18–64 years, Std = 10.110). The majority of respondents were highly educated (76.3%). Majority of respondents tended to go on vacation once or several times a year. However, there was also a significant number of those who have traveled abroad only several times until the present, meaning they were not frequent travelers (

Table 1. Sociodemographic characteristics and travel frequency of respondents.

Gender (%)		Age
Male	32.9	Average age: 31.91 (interval of 18–64 years, Std = 10.110)
Female	67.1
Education		Frequency of travelling
		I have never travelled abroad	3.5
Secondary school	20.9	I have traveled several times till now	30.2
Higher school/college	2.8	I usually travel once a year	23.9
Faculty/master/PhD	76.3	I usually travel several times a year	40.6
		I usually travel once a month	1.9

).

More than half of the respondents (52.9%) had visited Greece in the last few years, while almost all respondents (94.9%) were aware of the floods and fires that occurred in Greece in the summer of 2018 and 2019. Respondents mainly claimed (58.9%) that these events will not affect their future choice of Greece as a travel destination, while 36.7% said they would travel to nonaffected places. Only 4.4% of respondents said they would avoid Greece because of the risk of natural hazards.

3.3. Instruments

The survey was used to collect the study data. The first part of the survey included sociodemographic characteristics of respondents and their travel frequency.

In the second part, respondents were asked if they have visited Greece in the last few years and are they aware about the natural hazards that happened there during summer 2018 and 2019. Respondents were then evaluated for the effect of hazards on their intention to visit Greece in the future (1—It will not influence my intention, 2—I would choose nonaffected destinations, 3—I will not visit Greece because it was affected by natural hazards). Additionally, they have rated on the scale from 1 to 5 (1—is not likely at all, 2—it is unlikely, 3—I’m not sure, 4—it’s likely that I would, 5—it’s very likely) how likely it would be for them to change the decision to visit a destination in case some natural disasters (floods and fires) took place there (in the recent past). Respondents were also asked to estimate on a scale of 1 (very calm) to 5 (very worried) how much they would be concerned about your personal safety in case a certain natural disaster occurs at the destination in which they are spending their vacation.

In the last three parts of the survey, respondents’ answers were measured on a 5-point Likert scale (1—I strongly disagree; 5—I strongly agree). The third part of the questionnaire measured personality traits of respondents using MINI IPIP-6 scale [111]. The study also included Sensation seeking (8-item scale [112]) and Novelty seeking (9-item scale based on [113]).

The fourth part of the survey referred to respondents’ tourism worries on the 8-item scale developed by Larsen et al. [55]. Finally, the fifth part of the questionnaire included a 13-item scale developed by Thapa et al. [88] measuring tourist behavior based on the perceived risk.

3.4. Procedure

The online survey was conducted using Google Forms, for Serbian tourists, and a specialized site [114], for Romanian tourists. The survey in Romania was translated to Romanian, while a survey in Serbia was conducted in Serbian. The questionnaire was in two phases: 1) the first phase from July until September 2018 and 2) the second phase from September to December 2019. The survey was distributed with the help of travel agencies in both countries. The agencies sent the questionnaire via email to more than 1200 travel agencies’ clients in Serbia and more than 1000 travel agencies’ clients in Romania (two travel agencies sent emails to their customers with kind requests to complete the survey). Additionally, the survey was distributed via social networks (Facebook and Instagram) with a snowball convenience sample technique. The survey was anonymous and participation was voluntary.

3.5. Statistical Analysis

Data analysis was done in SPSS (Statistical Package for Social Sciences) 23. The tests applied to analyzed the data were linear regression analysis, ANOVA test, independent sample t-test, and Pearson test of correlations.

4. Results

4.1. Descriptive Statistics

Descriptive statistics for analyzed variables is presented in

Table 2. Descriptive statistics and scale reliability for all analyzed variables.

Dimension	Mean	Std.	α
Cautious behavior (8 items)	2.85	0.830	0.791
Courageous behavior (4 items)	3.16	0.677	0.713
Tourism worries (8 items)	2.10	0.985	0.857
Sensation seeking (8 items)	2.99	0.830	0.798
Novelty seeking (9 items)	3.91	0.720	0.848
Conscientiousness (4 items)	3.94	0.759	0.750
Extraversion (4 items)	3.25	0.865	0.760
Openness to experience (4 items)	3.80	0.747	0.720
Agreeableness (4 items)	3.90	0.777	0.706
Neuroticism (4 items)	2.78	0.786	0.738
Honesty–Humility (4 items)	2.54	0.844	0.754

. It can be seen that Cronbach’s alpha coefficient for all variables is above recommended 0.07, indicating that the reliability of instruments is satisfactory [115].

It is interesting to note that respondents show higher scores on Courageous compared to Conscious behavior. Additionally, they show quite low scores on Tourism worries. Moreover, the respondents rated how likely it is to change the decision to visit a destination in case some natural disasters (floods and fires) took place there (in the recent past). The mean value of the answers to this question is 3.28, with standard deviation of 1.078. When being asked how concerned they would be about their safety in case a certain natural disaster occurs at the destination in which they are spending their vacation, the average value of the answers is 3.79, with standard deviation of 1.048. Further analysis showed that 63.4% of participants claim that they would be worried and very worried if natural hazards took place at the destination on which they are spending their vacation.

4.2. Exploratory Factor Analysis (EFA) for Tourist Behavior Based on Perceived Risk

The scale used in the paper for measuring tourist behavior based on the perceived risk is the scale developed by Thapa et al. [88]. In their paper, the authors extracted three types of tourists according to their tourist behavior based on the perceived risk: Cautious, Courageous, and Conscious travelers. To check the factor structure, the exploratory factor analysis (EFA) was performed, with principal component analysis. Representativeness was good (KMO = 0.784) and Bartlett’s sphericity test was significant (χ2(55) = 1054.53, p < 0.001), which confirmed that the data are suitable for the analysis. Due to the communalities lower than 0.3, item 5 “The possibility of natural hazards in Greece discourages me from traveling there (I5)” was excluded from further analysis. Based on the parallel analysis, instead of three factors extracted in the study of Thapa et al. [88] being Conscious, Courageous, and Cautious, two dimensions (Cautious and Courageous) were extracted with 46.07% of variance explained. Varimax rotation was applied since the extracted components were not correlated (

Table 3. Rotated component matrix—factor structure of reactions to the perceived travel risk.

Dimension	Cautious A = 0.791	Courageous α = 0.713
Security is the most important factor when deciding where to travel. (I1)	0.618
Safety is the most important attribute that destinations in Florida can offer (I3)	0.603
I will only travel to Greece if I believe it is safe from natural hazards (I4)	0.574
When I am trying to decide between destinations in Greece and other States, I would choose the one which does not have active natural hazards (I6)	0.768
If a particular destination in Greece has experienced natural hazards in the past, I will not travel there (I7)	0.677
I’d like to travel to Greece but negative news about natural hazards discourages me (I8)	0.819
Other people’s negative experiences with natural hazards in Greece do not influence my decision to travel (I10)	0.615
When I’m evaluating destinations to travel, the risk of natural disasters is not a factor (I2)		0.687
Natural hazards in Greece have never influenced my decision to travel there (I9)		0.477
Safety is not an important consideration when I’m evaluating different destinations in Greece to travel to (I11)		0.715
I would not let natural hazards keep me from traveling to my final destination in Greece (I12)		0.46

).

4.3. The Role of Sociodemographic Characteristics in Tourism Worries and Behavior Based on the Perceived Travel Risk

To test the role of gender in tourism worries and behavior based on the perceived travel risk, an independent sample t-test was applied. The test shows significant differences between man and woman only in the case of Cautious behavior (t = −2.721, p = 0.007). Results indicate that women more than men (MD = 0.199) show Cautious behavior as a reaction on a perceived travel risk, which leads to the partial acceptance of Hypothesis 11.

Further on, to test the role of respondents’ education, ANOVA test was applied. Findings indicate significant differences in the case of Courageous behavior (F = 3.615, p = 0.028). LSD post hoc test indicates that respondents who have finished secondary school show more Courageous behavior than those with higher education (faculty, master, and PhD) with mean difference (MD = 0.244). This indicates that Hypothesis 13 can be rejected.

When it comes to age, the Pearson test of correlations was performed to test how the age of respondents is related to their tourism worries and behavior based on the perceived travel risk. The results are presented in

Table 4. Correlations between age, tourism worries, and behavior based on the perceived travel risk.

	Tourism Worries	Courageous Behavior	Cautious Behavior	Age
Tourism worries	1	−0.047	0.421 **	−0.131 **
Courageous behavior	−0.047	1	−0.121 *	−0.03
Cautious behavior	0.421 **	−0.121 *	1	0.038

** Correlation is significant at the 0.01 level (2-tailed); * Correlation is significant at the 0.05 level (2-tailed).

.

The results show that respondents’ age is not related to their behavior based on the perceived travel risk, but it is negatively correlated with Tourism worries. This means that the older respondents are less worried about their travel. This indicates that Hypothesis 12 can be partially accepted.

4.4. The Role of Personality in Tourism Worries and Behavior Based on the Perceived Travel Risk

To explore the role of personality in Tourism worries and behavior based on the perceived travel risk, a linear regression analysis was conducted. The findings are presented in

Table 5. Influence of tourist personality on tourism worries and behavior based on the perceived travel risk.

Tourist Personality	Tourism Worries			Cautious Behavior			Courageous Behavior
	F	R2	β	F	R2	β	F	R2	β
Sensation seeking	0.294	0.001	0.026	1.293	0.003	−0.055	0.782	0.002	0.043
Novelty seeking	0.002	0.000	0.002	7.352	0.017	0.130 **	2.808	0.007	−0.081
Conscientiousness	17.715	0.200	−0.010	6.809	0.088	0.166 **	0.334	0.005	0.003
Extraversion			−0.096 *			0.028			0.018
Openness			−0.166 **			−0.162 **			0.003
Agreeableness			−0.040			0.054			0.049
Neuroticism			0.182 **			0.062			0.045
Honesty-Humility			0.293 **			0.167 **			−0.011

** Correlation is significant at the 0.01 level (2-tailed); * Correlation is significant at the 0.05 level (2-tailed).

.

The findings indicate that Sensation-seeking and Novelty-seeking, as well as Conscientiousness and Agreeableness, are not predictors of Tourism worries. On the other hand, Extraversion and Openness are negative predictors of Tourism worries, whereas Neuroticism and Honesty-Humility show significant positive influence. When it comes to Cautious behavior, Novelty-seeking, Conscientiousness, and Honesty-Humility show significant positive influence, whereas Openness negatively predicts Cautious behavior. Interestingly, respondents’ personality did not show significant influence in case of Courageous behavior. This leads to the conclusion that Hypotheses 1, 2, 3, and 6 can be rejected, whereasHypotheses 4, 5, and 7 can be partially accepted and Hypothesis 8 can be fully accepted.

4.5. Comparative Analysis Between Serbian and Romanian Tourists

First, some basic information about spending a holiday in Greece, familiarity with natural hazards, and the possible effect of natural hazards on the future choice of Greece as a holiday destination were compared between two samples. The results are shown in

Table 6. Spending holiday in Greece, familiarity with natural hazards, and the possible effect of natural hazards on the future choice of Greece as a holiday destination.

	Serbian tourists (%)		Romanian tourists (%)
	Yes	No	Yes	No
Have you spent your holiday in Greece in the last few years?	58	42	47.3	52.7
Are you familiar with the wildfires and floods that hit Greece this summer?	94.2	5.8	95.7	4.3
If the answer to the previous question is yes, then will it affect your choice of Greece as a holiday destination?	Serbian tourists (%)		Romanian tourists (%)
Yes, that’s why I will not go to Greece.	1.8		0.5
Yes, I will choose places in Greece that are not affected by natural disasters.	31.3		42.5
Yes, this year I was present during these events in Greece and I do not want to experience it again.	6.2		0
No, it will not affect my destination selection.	60.7		57

.

The results indicate that a slightly higher number of tourists from Serbia have spent their holiday in Greece in the last few years. Moreover, both Romanian and Serbian tourists are highly aware of natural hazards that hit Greece in the last 2years. When it comes to the effect of natural hazards in Greece to the choice of Greece as a holiday destination, more than half of respondents in both samples claim that this would not affect their decision to visit Greece in the future. When comparing Serbian and Romanian tourists, some of the Serbian tourists already have experienced natural hazards in Greece and they did not want to experience this again. On the other hand, a higher percentage of Romanian compared to Serbian tourists would choose places in Greece that are not affected by natural disasters. In order to test the differences between Serbian and Romanian tourists in tourism worries and behavior based on the perceived travel risk, an independent sample t-test was applied. The results are shown in

Table 7. The results of the independent sample t-test.

	t	df	Sig. (2-Tailed)	Mean Difference
Courageous behavior	2.661	429	0.008	0.1725
Cautious behavior	−5.081	429	0.000	−0.3356
Tourism worries	−3.845	429	0.000	−0.2866

.

The findings suggest that Serbian and Romanian tourists differ both in tourism worries and behavior based on the perceived travel risk. Serbian tourists show more Courageous behavior compared to Romanian tourists. On the other hand, Romanian tourists show more Cautious behavior and score higher on tourism worries compared to Serbian tourists. Based on this, Hypothesis 9 can be accepted.

These results were further explored by testing the differences between two samples regarding their determination to change the decision to visit a destination in case some natural disasters (floods and fires) took place there (in the recent past). The analysis shows significant results (t = −2.759, p = 0.006) indicating that Romanian tourists (M = 3.43) are more ready than Serbian tourists (M = 3.15) to change the decision to visit a destination in case some natural disasters (floods and fires) took place there (in the recent past).

Romanian and Serbian tourists, as analysis shows, also differ based on their personality traits—Openness (t = 4.731, p = 0.000), Agreeableness (t = 2.561, p = 0.001), Honesty-Humility (t = −3.381, p = 0.001), and Novelty-seeking (t = −2.730, p = 0.007). Serbian tourists show higher scores on Openness (MD = 0.332) and Agreeableness (MD = 0.190), whereas Romanian tourists show higher scores on Honesty-Humility (MD = −0.271) and Novelty-seeking (MD = −0.188).

4.6. The Moderating Role of Nationality in Regression Between Personality Traits and Tourist Behavior Based on the Perceived Travel Risk

Finally, nationality was explored as a moderator in regression between the influence of tourist personality on tourist behavior based on the perceived travel risk. The findings suggest that Nationality moderates regression between Agreeableness and Cautious behavior. The analysis shows that R2 reported in Model 1 increased significantly in Model 2 (from 0.060 to 0.079).

Table 8. Regression coefficients ^a—moderating effect of Nationality on the effect of Agreeableness on Cautious behavior.

Model	Unstandardized Coefficients		Standardized Coefficients	t	Sig.
	B	Std. Error	Beta
1. (Constant)	2.152	0.206		10.468	0.000
Agreeableness	0.050	0.043	0.055	1.173	0.242
Nationality	0.345	0.066	0.247	5.222	0.000
2. (Constant)	3.148	0.545		5.780	0.000
Agreeableness	−0.203	0.135	−0.225	−1.504	0.133
Nationality	−0.309	0.338	−0.220	−0.913	0.362
Nationality * Agreeableness	0.167	0.085	0.523	1.973	0.049
Romanian Sample Tourists Agreeableness	0.132	0.064	0.142	2.059	0.041
Serbian Sample Tourists Agreeableness	−0.036	0.056	−0.043	−0.637	0.525

^a. Dependent variable: cautious behavior. * multiplied.

shows that the regression coefficient of the moderator (Nationality*Agreeableness) is significant meaning that Nationality moderates the relationship between Agreeableness and Cautious behavior. Further analysis showed that Agreeableness has a significant effect on Cautious behavior in a sample of Romanian tourists, whereas in the sample of Serbian tourists, the regression coefficient is insignificant.

The moderation analysis also showed that Nationality moderates the regression between Openness to experience and Cautious behavior. The findings show that R2 reported in Model 1 increased significantly in Model 2 (from 0.074 to 0.091).

Table 9. Regression coefficients ^a—moderating effect of Nationality on the effect of Openness to experience on Cautious behavior.

Model	Unstandardized Coefficients		Standardized Coefficients	t	Sig.
	B	Std. Error	Beta
1. (Constant)	2.885	0.217		13.265	0.000
Nationality	0.295	0.067	0.211	4.417	0.000
Openness	−0.122	0.045	−0.130	−2.722	0.007
2. (Constant)	1.524	0.527		2.891	0.004
Nationality	1.268	0.350	0.096	3.622	0.000
Openness	0.235	0.134	0.251	1.759	0.079
Nationality*Openness	−0.259	0.091	−0.726	−2.830	0.005
Romanian Sample Openness	−0.283	0.077	−0.248	−3.670	0.000
Serbian sample Openness	−0.024	0.052	−0.031	−0.456	0.649

^a. Dependent variable: cautious behavior. * multiplied.

shows that the regression coefficient of the moderator (Nationality*Openness) is significant meaning that Nationality moderates the relationship between Openness to experience and Cautious behavior. Further analysis showed that Openness has a significant effect on Cautious behavior in a sample of Romanian tourists, whereas in the sample of Serbian tourists, the regression coefficient is insignificant.

When it comes to Courageous behavior, Nationality showed to be a moderator in regression between Consciousness, Novelty-seeking and Sensation-seeking, and Courageous behavior. The findings show that in the case of Consciousness, R2 reported in Model 1 increased significantly in Model 2 (from 0.018 to 0.036).

Table 10. Regression coefficients ^a—moderating effect of Nationality on the effect of Sensation-seeking on Courageous behavior.

Model	Unstandardized Coefficients		Standardized Coefficients	t	Sig.
	B	Std. Error	Beta
1. (Constant)	3.163	0.127		24.878	0.000
Nationality	−0.170	0.065	−0.125	−2.616	0.009
Sensation-seeking	0.029	0.039	0.036	0.747	0.456
2. (Constant)	3.517	0.176		20.036	0.000
Nationality	−0.843	0.241	−0.623	−3.494	0.001
Sensation-seeking	−0.312	0.124	−0.383	−2.515	0.012
Nationality * Sensation-seeking	0.225	0.078	0.647	2.896	0.004
Romanian Sample Sensation-seeking	0.138	0.059	0.160	2.324	0.021
Serbian Sample Sensation-seeking	−0.087	0.050	−0.116	−1.734	0.084

^a. Dependent variable: courageous behavior. * multiplied.

indicates that moderator (Nationality*Sensation-seeking) is significant, meaning that Nationality moderates the relationship between Sensation-seeking and Courageous behavior. In the Romanian sample, Sensation-seeking positively affects Courageous behavior, whereas in the Serbian sample, regression coefficients are insignificant.

The findings show that in the case of Consciousness, R2 reported in Model 1 increased significantly in Model 2 (from 0.018 to 0.036).

Table 11. Regression Coefficients ^a—moderating effect of Nationality on the effect of Novelty-seeking on Courageous behaviour.

Model	Unstandardized Coefficients		Standardized Coefficients	t	Sig.
	B	Std. Error	Beta
1. (Constant)	3.486	0.179		19.473	0.000
Nationality	−0.161	0.065	−0.119	−2.465	0.014
Novelty-seeking	−0.061	0.045	−0.065	−1.349	0.178
2. (Constant)	3.917	0.218		17.993	0.000
Nationality	−1.415	0.375	−1.045	−3.774	0.000
Novelty-seeking	−0.492	0.135	−0.523	−3.656	0.000
Nationality * Novelty -seeking	0.318	0.094	1.097	3.395	0.001
Romanian Sample Novelty-seeking	0.144	0.084	0.120	1.724	0.086
Serbian Sample Novelty-seeking	−0.174	0.049	−0.230	−3.521	0.001

^a. Dependent variable: courageous behavior. * multiplied.

indicates that moderator (Nationality*Novelty-seeking) is significant, meaning that Nationality moderates the relationship between Novelty-seeking and Courageous behavior. In Serbian sample, Novelty-seeking negatively affects Courageous behavior, while in the Romanian sample, regression coefficients are insignificant. Based on this, Hypothesis 10 can be accepted, as nationality was confirmed to moderate the relationship between personality traits and behavior based on the perceived risk.

5. Discussions

The study of Kovačić et al. [44] investigated the inter-relationships of tourist personality, tourism worries, and tourists’ behavior based on the perceived risk. This revealed that some personality traits have a direct influence on behavior patterns in regard to a perceived risk, while the influence of others seems to be mediated by tourism worries. This study called for future research to include sociodemographic characteristics and tourists’ nationality to contribute to a better understanding of these complex relationships. Inspired by this, the principal aim of the current study was to analyze whether tourists’ individual characteristics (personality and sociodemographic characteristics) influence tourism worries and tourists’ behavior based on the perceived risk as well as to explore a moderating role of tourists’ nationality (Romanian and Serbian tourists) on a relationship between personality and tourists’ behavior based on the perceived risk. The contribution of this study also mirrors in the fact that research on individual characteristics such as sociodemographics and personality and their influence on behavior based on the perceived risk is very rare in the tourism field while moderating role of nationality on this relationship was not explored before.

5.1. The Role of Tourists’ Personality

When it comes to the direct effect of tourists’ personality on tourism worries and tourists’ behavior based on the perceived risk, the findings suggest significant negative effect of Extraversion and Openness on Tourism worries and significant positive effects of Neuroticism and Honesty-Humility. This means that extraverts and people open to experiences show less anxiety about travel, which is in line with previous studies [48,57]. On the other hand, people who are more neurotic and show more honesty-humility will show more tourism worries. Previous studies indicated that those individuals who score high at honesty-humility show higher risk-perception, which might be connected with more tourism worries [57] that neurotic individuals are generally more anxious and worried [44,56].

Regarding tourists’ behavior based on the perceived risk, Novelty-seeking, Conscientiousness, and Honesty-Humility show a significant positive influence on Cautious behavior, whereas Openness shows negative effects. So, people who are more conscious, novelty-seekers, and show more honesty-humility will show more Cautious behavior as a reaction to the perceived risk. This is similar to the findings of Kovačić et al. [44], who revealed that Conscientiousness has a direct effect on Cautious behavior, whereas the effect of Honesty-Humility was mediated by Tourism worries. The study of Weller and Tikir [57] also revealed that Conscientious have more tendency to avoid risky behavior, while in their study, Honesty-Humility shows no effect. However, we should bear in mind that tourism risks are a specific field, which might also result in a different relationship between personality and behavior based on the perceived risk. Contrary to what was expected, respondents’ personality did not show significant influence in the case of Courageous behavior. This is a similar finding as in the study by Kovačić et al. [44], who found that only Novelty-seeking was negatively related to Courageous behavior. Based on this, Hypotheses 1, 2, 3, and 6 are rejected, whereas Hypotheses 4, 5, and 7 are partially accepted and Hypothesis 8 is fully accepted.

5.2. The Role of Sociodemographic Characteristics

The findings of this study indicate the significant role of sociodemographic characteristics in tourism worries and tourists’ behavior based on the perceived risk. In such a way, the study contributes to the theoretical knowledge, as there is rather obscure tourism literature exploring the effect of sociodemographic characteristics on risk perception and readiness to engage in risky behavior and visit destinations perceived as risky. The study found that women more than men show Cautious behavior as a reaction to a perceived travel risk, which leads to the partial acceptance of Hypothesis 11. This is in line with previous findings indicating that women are more concerned about travel risk and are less ready to engage in risky behavior [46,62,63,76,77,78,79,80]. An interesting study by Gangestad and Snyder [82] discussed that possible differences between men and women in risk perception may also be influenced by the fact that women are more ready to confess when they feel fear than men. If we consider that, especially for family trips, women are dominantly the ones who make travel decisions; this finding is valuable for marketers, to focus their campaigns on encouraging women to take trips and focus more communication on safety and security.

When it comes to the role of respondents’ education, findings indicate that respondents who have finished secondary school show more Courageous behavior than those with higher education (faculty, master, and PhD). This has led to the rejection of Hypothesis 13, as such findings are opposite to previous studies, which indicate [63,64,88] that the people with higher education worry less and show few fears about uncertainty, whereas people with average education (such as secondary school) show more worries and are more ready to avoid risk-taking. On the other hand, finding is in line with the study of Maciejewski [87], which proved that people with basic education more often take risk when purchasing tourist services. Such finding might be explained by the fact that people who finished secondary education might be less informed about the danger that natural hazards may cause, so they show more courageous behavior. In addition, the current study does not only focus on the perception of risk but also on the behavior based on risk perception. Thus, another explanation may lay in the fact that people with secondary school might be the people with lower income, so they mainly choose their travel destination based on prices and might be less ready to change their travel choices and behavior because of the probability of natural hazards.

When it comes to the respondents’ age, the study shows that age is not related to respondents’ behavior based on the perceived travel risk. However, it is significantly negatively correlated with tourism worries. This means that the older respondents are less worried about their travel, which is in accordance with previous research [61,64,83] claiming that younger respondents show much more concern about various risks than older age groups. This indicates that Hypothesis 12 can be partially accepted.

5.3. The Role of Nationality

One of the main assumptions of the study is that nationality has an important role in behavior based on the perceived risk and tourism worries, as many previous studies proved that different nationalities differ in risk perception and their desire to visit risky destinations [48,65,66,67,68,69,70,71]. For testing this assumption, in the current study, we compared the sample of Romanian and Serbian tourists. These tourism markets were chosen based on the proximity of the Greece and number of arrivals (these are among major tourism markets in Greece). The findings of our study suggest that Serbian and Romanian tourists differ both in tourism worries and behavior based on the perceived travel risk (Hypothesis 9 is accepted). Serbian tourists show more Courageous behavior compared to Romanian tourists. On the other hand, Romanian tourists show more Cautious behavior and score higher on tourism worries compared to Serbian tourists. This can be explained by the fact that the Romanian sample is from NE Romania, where recent big flood events happened alongside Siret and Prut rivers; in the summers of 2005, 2008, and 2010; the mentioned rivers caused severe floods [116,117], which were intensely publicized by media and shared on social networks; a mental fingerprint is possible to be seen on Caution behavior, taking into account the same pattern of highly distributed events from Greece in 2018 and 2019.

In accordance with this, the results also revealed that Romanian tourists are more ready than Serbian tourists to change the decision to visit a destination in case some natural disasters (floods and fires) took place there (in the recent past). This can be explained by the fact that in the case of severe natural hazards, Romanians seems to change much easier their summer destinations because they can choose Romanian Black Sea coast, which traditionally is not associated with major climate-related hazards (there are no any events reported here), being perceived a safe destination. Although some of the previous studies [48,60] claim that differences in perceptions of travel risk level of anxiety and travel intentions between different nationalities can be explained by different level of uncertainty avoidance (Hofstede’s cultural dimension); this is not the case with Serbia and Romania, as they both score very high on this dimension (90 and 92 out of 100, respectively, online: https://www.hofstede-insights.com/product/compare-countries/, respectively). Inspired by Kozak et al. [60] who suggest that tourists’ personality should be researched in order to support and explain nationality differences, we further explored this issue, checking the moderation role of nationality in regression between tourist’s personality and behavior based on the perceived risk. The moderating role of nationality on a relationship between personality traits and behavior based on the perceived risk was confirmed and Hypothesis 10 was accepted. Nationality showed to be a moderator in the relationship between Agreeableness and Cautious behavior and Openness to experience and Cautious behavior. When it comes to Courageous behavior, Nationality showed to be a moderator in regression between Consciousness, Novelty-seeking and Sensation-seeking, and Courageous behavior.

Further analysis revealed that Agreeableness has a significant positive effect on Cautious behavior in a sample of Romanian tourists, whereas Openness shows negative effects on Cautious behavior. This means that Romanian tourists who are agreeable will show more Cautious behavior, whereas those who are open to experience will be less Cautious. Previous studies showed that agreeable individuals show more Conscious behavior [44], whereas individuals open to experience are more ready to take social and recreational risks [57].

In the Serbian sample, Novelty-seeking negatively affects Courageous behavior, whereas Sensation-seeking positively affects Courageous behavior. Sensation seeking is often related to readiness to take risks and visit risky destinations, as well as Novelty-seeking. However, Novelty-seeking in a sample of Serbian tourists, the majority of which traveled to Greece several times till now, in combination with the existing possibility that natural hazards could happen again, may result in their desire to change destination and visit something novel instead. Novelty-seeking also showed to be a negative predictor of Courageous behavior in a study by Kovačić et al. [44], which also included Serbian tourists. Those Serbian tourists that are sensation-seekers would, however, show more Courageous behavior and visit Greece regardless of natural hazards probability. This might explain why Serbian tourists show more Courageous behavior than Romanian tourists.

6. Conclusions

The study aimed to test the total of 13 hypothesis, eight of which were regarding the influence of tourist personality on travel behavior based on the perceived risk. The analysis resulted in rejecting four hypotheses (H1, 2, 3, and 6), partial acceptance of three (H4,5, and 7), and full acceptance of one hypothesis (H8). More specifically, Sensation-seeking, Extraversion, and Agreeableness did not show to be predictors of travel behavior based on the perceived risk, as previously assumed, whereas Novelty-seeking showed positive effect on Cautious behavior, which was contrary to what was expected. On the other hand, as expected, the study confirmed the positive influence of Conscientiousness and Honesty-Humility on Cautious behavior, whereas Openness showed negative effects. Regarding the influence of sociodemographic characteristics on tourist behavior based on the perceived risk, three hypotheses were formulated, one of which was rejected — H13 (as the study showed that less educated respondents show more Courageous behavior, which was contrary to what was initially assumed). On the other hand, as initially proposed, women more than men show Cautious behavior (H11), whereas younger respondents show more concern about various risks (H12).

The study also confirmed the moderating role of nationality on a relationship between personality traits and behavior based on the perceived risk (H10) as well as the existence of difference between Serbian and Romanian tourists in tourism worries and behavior based on the perceived travel risk (H9), indicating that Serbian tourists show more Courageous behavior compared to Romanian tourists, whereas Romanian tourists show more Cautious behavior and score higher on tourism worries compared to Serbian tourists.

Based on such findings, the current study provides several important contributions to theory and practice. First, it provides evidence that not only tourist personality but also sociodemographic characteristics influence tourism worries and tourist behavior based on the perceived risk. In this way, the study shows that individual characteristics of tourists are of paramount importance for predicting differences in their behavior, specifically, for predicting the profile of tourists who will show more courageous behavior and those who will stay cautious. Thus, the study provides important implications for marketing managers for creating promotional campaigns in the time of crisis. The study indicates that especially carefully designed promotional campaigns should be directed towards women and highly educated tourists who show tendencies towards more cautious behavior, as well as younger travelers who tend to be more worried about travel. Such promotional campaigns should provide detailed information about current safety measures at destinations, as well as special customer care to diminish their worries and fear for their safety and quality of their vacation. Regarding personality traits, tourists who are more conscious, novelty-seekers, and show more honesty-humility will show more cautious behavior as a reaction to the perceived risk of natural hazards in Greece.

Additionally, the study is the one of the first to explore and confirm the role of nationality in tourist behavior based on the perceived risk and tourism worries, as well as the moderation role of nationality in regression between tourist’s personality and behavior based on the perceived risk. Finding that Serbian tourists show more Courageous behavior compared to Romanian tourists who are more cautious and score higher on tourism worries also indicates that promotional campaigns on these two markets should be different. Serbian tourists would probably go to Greece and not worry about the risk of natural hazards, whereas Romanian tourists would need more persuasive and intensive promotional campaigns to diminish their worries. Specifically, those Romanian tourists who are agreeable will show more Cautious behavior, whereas those who are open to experience will be less Cautious. Thus, marketing of new offers, experiences, and activities may positively influence Romanian tourists who are open to experience to visit Greece even when they perceive it as risky. On the other hand, in the Serbian sample, sensation-seekers will be especially interested in visiting Greece.

As limitation of this study, we can state that even the present COVID-19 pandemic crisis will dramatically change many of the structural components of international tourism, our study is radiography of prepandemic stage. Structural changes and adaptive measures of tourist businesses in a relationship with public health requirements will be much deeper than local/regional level of the floods and wildfires and, in this sense, the results of our study will respond just partially to these unavoidable and deep changes. Nevertheless, our study explores Serbian and Romanian tourists’ behavior and, in this sense, Greece has an increasing interest taking into account that, in the next years, international tourism will be a regional one (long distances will be avoided).

Future research should test the model with other nationalities, especially those who are culturally different than Serbian and Romanian tourists. To test such regression model of personality traits and travel behavior based on the perceived risk on a sample of many different nationalities, the authors plan to use partial least squares structural equation modeling (PLS-SEM) with the software such as SMART-PLS or AMOS, as could greatly assist in further testing a moderation/mediation role of nationality in the proposed regression model. Moreover, future studies should explore the role of destination image and brand perception on respondents’ behavior based on the perceived risk, with the assumption that better image perception will lead to higher intention to visit destination even in case of possible risk.