The Role of Information Sources on Tourist Behavior Post-Earthquake Disaster in Indonesia: A Stimulus–Organism–Response (SOR) Approach

Abstract

The earthquake disaster has an impact on tourist visit intention. This study aims to investigate tourist behavior in the post-earthquake disaster linkage between information sources (word of mouth and electronic word of mouth) and risk perception toward tourists’ visit intentions to a destination in Indonesia. This study applies the SOR theory to predict tourists’ behavior in the destination aftermath. The Partial Least Squares Structural Equation Model was used to examine the hypothesis of the study. The result found that information sources (electronic word of mouth and word of mouth) significantly influenced visit intention in the time of post-earthquake disaster. The risk perception has not significantly influenced visit intention in post-earthquake disasters. The discussion and conclusion of the study are discussed herein. Overall, the findings of the study may contribute to the theory by adding information sources to predict tourist behavior post-earthquake disaster and also gives a practical contribution to the tourism sector, stakeholders, tourism marketers, and policymakers in Indonesia to enhance the marketing strategy by considering destination promotion through word of mouth (offline) and electronic word of mouth (online) and its mechanism on tourists’ travel decision in the time of aftermath.

1. Introduction

The tourism sector is one of the sectors that has been affected by both natural and human-made disasters, influencing tourist visit intention. Natural disasters such as earthquake disasters, pandemics or endemics, volcanoes, floods, and eruptions can significantly influence tourism and impact tourists’ visit intentions [1]. The tourism industry has turbulence and shock from disasters, such as floods and earthquakes [2], or human-made crises, such as political instability, wars, and terrorist [3]. The current COVID-19 pandemic is one of the disasters in the health sector, which significantly impacts the tourism sector. In addition, a study by [4] stated that natural disasters, such as the risk of the COVID-19 pandemic, could influence tourists’ visit intention during crises. Furthermore, the disaster can lead to a decline in the number of visitors in the case of the Asia-Pacific region by around twelve million [5].

Indonesia is one of the countries with the highest number of disasters globally. This was supported by studies [6,7,8] revealing that Indonesia is one of the countries most prone to disaster. In addition, Indonesia is one of the countries with a high risk of natural disasters or “ring of fire” [9]. The number of disasters has significantly increased from 1970 to 2020. Based on the data from the National Disaster Management Authority (BNPB), it was reported that the number of disasters was 197 floods, 27 landslide events, forest fires, three volcanic eruptions, and 13 of the biggest earthquake disasters [10]. These natural disaster events impact the tourism sector and the number of visitors visiting a destination in Indonesia. This result was supported by [11], which states that tourists have some attention to their destination, as it is impacted by disaster.

However, despite the impact of the disaster on tourism, tourists’ perceptions toward the destination in the case of post-disaster need to be explored by scholars. Existing studies have largely discussed tourists’ visit intention in the aftermath of a natural disaster from several theories such as [12] used theory (Pleasure–Arousal–Dominant), cognitive and affective components of visit intention [13], consumer post-disaster behavior [14], and visit intention toward dark tourism theory [15]. In addition, the study on tourist behavior and visit intention in the case of post-disaster needs to investigate the information sources such as word of mouth and electronic word of mouth.

Word of mouth is a marketing promotion method that involves oral communication between the giver and receiver of information. Several studies in the marketing field used word of mouth to influence purchase intention [16,17], and in tourism and hospitality to influence behavioral intention in service [18,19,20] revealed that communication through word of mouth successfully influences tourists’ visit intention.

Information technology and social media are leading sources of online reviews for deciding whether tourists visit a destination or not [21]. Visitors can easily share information and experience from social media and influence the travelers’ options [22]. Electronic word of mouth is one of the parts of social media with a reliable information source that is effective for travelers in planning to visit a destination [23] and plays a piece of information in the consumer’s decision-making [24,25]. According to some studies, e-WOM on social media can increase travelers’ awareness and travel options when making decisions [26,27,28].

Electronic word of mouth is a type of communication that uses internet technology and has been used in marketing [29]. Despite the concept of “word of mouth”, electronic word of mouth (e-WOM) is communication through internet-mediated communication about products, services, or brands [30]. Electronic word-of-mouth concepts have been used in several areas, such as marketing, tourism, and hospitality [29], and predict tourist behavior [31]. Electronic word of mouth influences tourist behavior in the context of attitudes such as cognitive, normative, and effective [31]. Furthermore, this study used the SOR theory to predict tourist visit intention in the time of post-earthquake disaster. According to SOR theory, cognitive and affective processing as an organism, electronic word of mouth as a stimulus, transfers into behavioral intention response [32,33]. The literature found that e-WOM influences the intention to purchase [34,35]. However, the relationship process must be clarified in the aftermath of a disaster. Some works of literature have explained the relationship between e-WOM and intention through either moderator or mediator variables [31,36,37], but we need to explore the role of e-WOM as destination marketing in the case of a post-earthquake disaster.

Word of mouth and electronic word of mouth (e-WOM) have been discussed by several scholars in the tourism field, such as [31,38,39] revealed that word of mouth and electronic word of mouth influence the tourist’s decision to visit a tourist destination. In addition, WOM is one of the methods for promoting a destination and influencing tourists to visit a destination. Moreover, tourist visit intentions and attitudes were influenced by electronic word of mouth (e-WOM) in making the decision to visit a destination [40].

However, scholars have yet to concern more on tourist behavior or tourists’ visit intention in a destination in post-earthquake disaster using SOR theory. Thus, this study tried to investigate the SOR theory to examine the tourist visit intention by using word of mouth, electronic word of mouth, and risk perception in the case of post-earthquake disaster (natural disaster) in Indonesia. According to SOR theory, word of mouth refers to a “stimulus” that transfers into a behavioral intention as a “response” [32,33]. Moreover, in the context of post-disaster, no study has been conducted to explain the influence of word of mouth and electronic word of mouth on visit intention in the case of a post-disaster using the SOR theory.

This study has several objectives. First, we examine two factors of the stimulus of visit intention in the post-earthquake disaster, namely the role of electronic word of mouth and word of mouth, using the essential conceptual background of SOR theory. The findings expand on existing knowledge as information sources in post-disaster marketing destinations as a “stimulus” for tourists to visit a destination. Second, we investigate the role of risk perception as an “organism” in the SOR theory on visit intention in the post-earthquake disaster. Consequently, by extending the theoretical framework, this study formally analyzes the relationship between word of mouth, electronic word of mouth, risk perception, and visit intention. Moreover, this study used a Partial Least Squares Structural Equation Modeling (PLS-SEM) to propose the empirical model because this work examines the causal interrelationships of its constructs. Thus, our work can extend the theoretical framework by extending the information sources in destination marketing (WOM and e-WOM) in the case of the post-earthquake disaster, in which the effect of risk perception on visit intention is investigated.

This study provides insight into the effectiveness of information sources in destination marketing (WOM and e-WOM) and perceived risk on visit intention in the aftermath of a natural disaster. As the risk of the earthquake disaster affects risk perception and behavior patterns, using information sources in destination marketing post-disaster can play a central role in helping recover the destination and building the tourism industry’s resilience. As a result, this study contributes to post-disaster destination marketing research by shaping tourists’ psychological decisions to visit a destination. In addition, the study’s findings contribute to the literature linkage to SOR theory by applying it in a novel setting, especially for tourist behavior in the post-earthquake disaster. Furthermore, this study explores the tourist’s emotional and behavioral responses to visiting a destination post-disaster. It also demonstrates the effectiveness of online and offline destination marketing in influencing tourist visit intention. In addition, tourism resilience and sustainability are on the agenda in the tourism sector in case of post-disaster, especially in case of post-earthquake disasters due to the negative impact of the disaster on tourism. However, marketing toward business and sustainable tourist destinations has the role of increasing tourist visit intention in the post-disaster. Good marketing is responsible for the destination’s sustainability after a disaster to contribute to the business and economic sector [41]. Thus, the study’s result contributes to destination marketers and stakeholders arranging the promotion of destination marketing to affect tourists to visit a destination in the post-disaster period.

2. Literature Review and Hypothesis Development

2.1. Stimulus–Organism–Response Theory

This study adopted the theory of SOR based on tourist visit intention in case of an earthquake disaster. The concept was founded by [32], who stated that the SOR theory was divided into the environmental aspect (stimulus), generates the individual psychology (organism), and then can shape the behavior of (response). Based on the SOR framework, the consumer’s behavioral intention is formed by the stimulus from the external factor, such as the environment, and the internal aspect, such as psychology and physiology [32]. The concept of SOR refers to the external stimulus of the individual to influence the personal perception, attitude, and shaping their behavior [42]. The SOR theory was expanded in several areas of study, including marketing and tourism [43]. Several studies have been conducted in tourism using the Stimulus–Organism–Response (SOR) theory, particularly in the hospitality industry [44], destination image [45], rural tourism experience [46], and solo outbound travel [47]. “Stimulus” are factors that predict the consumer perception as starting point for making a decision. However, the framework of SOR has yet to be used in the context of tourist visit intention post-earthquake disaster based on the information sources. Thus, this study tried to investigate the framework of SOR as a leading theory for tourists’ behavior in post-earthquake disasters.

“Organism” is a component of the SOR theory. The organism refers to the internal processes of the individual that intervene between the external stimulant and individual action and response [32]. In the context of the original framework, the organism mainly refers to emotional and cognitive aspects [32]. A framework or construct such as emotions or feelings [44,48,49], customer value [50], and corporate image [51] have been used to represent the aspect of the organism in consumers in empirical studies on the tourism and hospitality industries. Furthermore, the component of the SOR theory is “organism”. The term “organism” in SOR theory has been used to examine the perceived safety and emotional solidarity as an organism by [47], tourism experience, environment awareness, and connectedness to nature [46] and social perception of destination image [45], perceived enjoyment and perceived usefulness in travel applications [52].

The last component in SOR theory is “response”. The final component is an outcome or result and decision that refers to the “approach or avoidance of behaviors” [32]. Several studies used the response constructs in tourism and hospitality studies, such as those [42] in the case of green purchase intention, behavioral intention to visit [47], green consumption [46], support for to use of a mascot in destination image [45,50,51] intention to use WOM, and using electronic word of mouth on tourism destination [53].

Following the logic of the studies above, this study employed the framework of SOR theory. It aimed to clarify the decision of tourists to visit a destination in the case of a post-earthquake disaster. In this study, the “stimulus” involved electronic word of mouth and word of mouth. Second, risk perception is an “organism” indicator in the SOR theory. The last visit intention in the post-earthquake disaster is as a “response” in the SOR framework. Moreover, the SOR can predict that tourist visit intention in the post-earthquake disaster based on information sources (WOM and e-WOM).

2.2. Word of Mouth and Visit Intention

The study [54] found the concept of word of mouth (WOM). The concept of word of mouth (WOM) refers to a type of communication between a giver and a receiver of information about a brand, product, or service. From this explanation, “word of mouth” means having positive or negative information. Word of mouth (WOM) refers to communication by individuals or people to share their experiences and assessment [55]. Furthermore, word-of-mouth communication refers to expressing positive and negative service experiences, which impacts purchasing behavior [55]. Positive and negative experiences indicate that positive word-of-mouth communication relates to satisfaction with service and experience.

In contrast, negative communication generally refers to dissatisfaction with a product or service [56]. Consumer satisfaction can be increased through communication or promotion through word of mouth [57] and revenue or income [58]. Consumers also revealed that receiving recommendations from friends, family, and colleagues who are satisfied through word of mouth about the product, service, and company impacts sales and income. Thus, according to several explanations, “word of mouth” in this study refers to communication or providing information about the destination to influence tourist visit intention in the case of a post-earthquake disaster.

Tourists will receive information about the experience from other tourists based on word of mouth in the context of the tourist destination, and strong enough to make a decision refers to tourist satisfaction with both positive and negative information [59]. However, the negative image of the destination can be framed by word of mouth [60,61]. The negative image of a destination in the tourism sector can be influenced by the damage of a disaster that cannot be avoided [61], and the negative image of a destination can be influenced by word of mouth when deciding to visit a destination [60].

One study [62] revealed that word of mouth could influence tourists’ perceptions and behavior regarding their visit intention. It can be concluded that word of mouth is an information source [38]. Word of mouth can predict and influence tourist behavior [63]. Positive word of mouth can provide a competitive advantage for tourists [38].

On the other hand, the negative impact of word of mouth can be attributed to the negative perspective of the customers toward loyalty, brand, reputation, and the management of the company [62]. In general, the positive information through word of mouth on tourist destinations in post-disaster cases will benefit tourists’ visit intention. Therefore, this study uses word of mouth as a “stimulus” and visit intention as a “response” in the - S–O–R framework in the context of post-earthquake disasters. Thus, from several explanations above, this study predicts that:

Hypothesis 1a.

Word of mouth is positively related to visit intention in case of a post-earthquake disaster.

Hypothesis 1b.

Word of mouth is positively related to the perceived risk in the case of a post-earthquake disaster.

2.3. Electronic Word of Mouth (e-WOM) and Visit Intention

One of the information sources in communication sources is electronic word of mouth (e-WOM). Recently, many studies have been conducted using electronic word of mouth in marketing and tourism [38]. Electronic word of mouth is “all information communication through internet-based technology or directly to consumers about the usage of specific goods and services”. Furthermore, social media is also well known for social networking, such as social media platforms including Facebook, Twitter, and social media website networking. Electronic word of mouth or social media can share information on medical and healthcare issues [64] and disaster information [65].

Electronic word of mouth is one of the information sources related to positive and negative product and service reviews that can be used to attract customers via the internet [66]. In addition, e-WOM, as a social media platform, is one platform that can share and post information content [35]. Additionally, electronic word of mouth allows users to share and post their thoughts, content, pictures, and videos in their account or application [67]. In marketing, consumers can receive and consider information about the products and services through electronic word of mouth [35].

Electronic word of mouth (e-WOM) impacts tourists’ decisions in the tourism sector. The first study from [68] revealed that tourists could share good information and experiences. It refers to the search for information by tourists based on tourist information. Second, the tourism industry requires sharing up-to-date information with users, tourists, or visitors [69], as e-WOM is an urgent information source from which tourists can obtain trustworthy information. Furthermore, tourists can access information such as tourist destinations [34].

Some studies have been conducted on the relationship between e-WOM and visit intention [40], revealing that e-WOM affects tourist visit intention to a destination. In addition, the study from [70] in medical tourism found that e-WOM can affect tourist visit intention. Previous studies have found that positive information will affect tourist behavior. Positive sharing on e-WOM can influence tourist behavior toward visiting a destination [71]. Using electronic word of mouth to provide the destination information post-disaster will increase the tourists’ intention to visit a destination. From several explanations in the literature, it can be concluded that the tourists’ visit intention is based on e-WOM. Thus, this study used electronic word of mouth (e-WOM) as a “stimulus” and visit intention as a “response”. With the same logic, it is predicted that in the post-earthquake disaster context, tourists’ will be gained attention when using e-WOM to post or give information related to a tourist destination, tourists have attention to visit destination. Thus, the hypothesis proposed is:

Hypothesis 2a.

Electronic word of mouth is positively related to visit intention in the case of a post-earthquake disaster.

Hypothesis 2b.

Electronic word of mouth is positively related to the perceived risk in the case of a post-earthquake disaster.

2.4. Risk Perception and Visit Intention

The perception of risk is one of the factors considered in the tourism industry when deciding whether or not to visit a destination [72]. Some risks can influence tourists traveling to their destination, such as those in the health sector, natural disasters, political instability, and terrorism [73,74]. The disaster can influence tourist behavior, alter perceptions of the destination’s risk, and reduce tourist arrivals [75]. In marketing, risk perception relates to the products, financial situation, mode of purchase, psychological consequence, and personal perception [73,76]. However, in the tourism and hospitality fields, the perceived risk occurs when the tourist has a congruence difference with the actual image of their travel and visit intention [77].

Risk perception refers to the feeling of someone to fear and worry about risk [78]. Risk perception has an impact on tourist visit intention [79]. Several studies have been conducted on risk perception and visit intention. The result found that the tourist’s intention to visit the destination will consider the risk as the main factor in deciding while traveling [80]. Furthermore, the study from [81] revealed that citizens considered the destination to be safe while visiting for the event of the Olympic Games.

However, the study from several scholars on risk perception, such as in risk of COVID-19 and visit intention [4,82,83], revealed that the risk of COVID-19 significantly influences tourists’ visit intention during the crisis, including risk perception in times of crisis, such as unstable politics [80], health issues [84], and pandemic issues [85].

Based on the SOR theory in this study, risk perception is an “organism” mechanism toward the decision of tourists to visit a destination in cases of post-earthquake disaster. Risk perception refers to the feeling of tourists about the risk that will be perceived to travel while planning to visit a destination in post-disaster. When tourists perceive the risk when traveling, it will reduce their intention to visit the destination. Based on the explanation, the hypothesis is proposed:

Hypothesis 3.

Risk perception is negatively related to visit intention in the case of a post-earthquake disaster.

Figure 1 depicts the SOR research model based on the hypothesis developed in the literature review. In this study, the information sources are word of mouth and electronic word of mouth as a “stimulus”, risk perception is an “organism”, and visit intention is a “response” in the framework of SOR theory.

3. Methodology

3.1. Research Design

This study used the quantitative analysis approach to answer the research hypothesis model in Figure 1. The quantitative method with questionnaire survey methods is one of the methods with adequate time and efficient low-money budgeting [86]. Furthermore, a cross-sectional data collection method was applied because it is easier to reach research targets, does not require an individual-identities, and respondents feel free to answer the questions [87]. Thus, from several explanations, this study used the quantitative method with a questionnaire survey and cross-sectional data collection methods. Several procedures were conducted to reach the purpose of the study, such as validity and reliability testing, testing the structural equation model, and hypothesis testing (t-test).

3.2. Data Collection and Sampling of the Study

This study used an online survey (an online questionnaire) to collect the potential respondents. Some scholars used an online survey to collect data on tourism and hospitality [88,89]. The online questionnaire was sent to the potential respondents or Indonesian tourists who desired to visit the destination of Lombok, West Nusa Tenggara, Indonesia. Based on these criteria, the tourists or visitors from outside of Lombok, West Nusa Tenggara, are the respondents in this study. In this context, the online questionnaire was created in this direction to reach the respondent outside of the study area. The online questionnaire was shared through social media platforms (WhatsApp, Facebook, mail, etc.) to obtain potential respondents. The respondents in this study were selected through screening questions to identify potential respondents. Only those who intended to visit Lombok, Indonesia, were asked to complete the questionnaire, and in the body of the questions applied guidelines and instructions for respondents. The data were collected from 2020 to 2021. Ethical approval was granted, and consent from all the respondents in this study was obtained. The researchers assured the respondents that the anonymity of the responses would be maintained.

This study aimed to explain tourist visit intention based on word of mouth, electronic word of mouth, and risk perception in the case of a post-earthquake disaster in Indonesia. Thus, tourists in Indonesia were the target population of the study. In the probability sampling technique, every element of the population had an equal chance of being selected as a respondent when the sample frame is available [90]. However, when no sampling frame was available, the study used a non-probability purposive sampling technique [91]. Thus, this study used the purposive sampling method to analyze the tourists who desired to visit Lombok’s destination in case of a post-earthquake.

Furthermore, for the sample size in this study, the state from [92] revealed that the small sample size is mainly used in the Structural Equation Model-Partial Least Squares (SEM-PLS). In addition, the study [93] cited in [94] revealed that most structural equation modeling studies used samples from 30 to less than 500 samples. Initially, we received 290 responses in this study, of which only 208 were considered valid for those who desired to visit a destination in a post-earthquake disaster. Thus, the sample size in this study was only 208 questionnaires with a 71.72% of response rate. Based on this result of the sample size, Structural Equation Modelling based on [95] stated that the median or minimum sample size is 200 samples. Thus, this sample size was more significant than the minimum sample size in this study.

3.3. Questionnaire Measurement and Development

The items or dimensions of the variables were developed from several studies. The constructs are divided into several variables, such as electronic word of mouth, word of mouth, risk perception, and visit intention. The electronic word-of-mouth variable with seven questions was developed from [65,70,96] and divided into several dimensions, such as information adoption, the credibility of the information, objectivity, and trust in e-WOM. The variable of word of mouth with four questions was adopted from [39,97,98,99] and divided into dimensions of source credibility and trust in information. The variable of risk perception with three questions was developed from [100,101] and divided into the worrying, high-risk, and dangerous dimensions. The last variable of visit intention with three questions was adopted from several sources [4,102,103,104] and divided into indicators of visit intention and willingness to visit. The study mainly contained 17 items to measure the variable of the study.

Furthermore, this questionnaire used the Indonesian language to assist Indonesian tourists in understanding the questions. In addition, this study applied seven-step Likert-type scales for the questions, with scales ranging from 1: “Strongly Disagree” to 7: “Strongly Agree”.

3.4. Data Analysis

The Structural Equation Model-Partial Least Squares examined the study’s hypotheses. First, PLS-SEM is one of the methods with higher statistical power and is useful for exploratory research, which aims to develop a theory [105]. Therefore, the current study is exploratory, and PLS-SEM, a variance analysis technique, is more appropriate than CBS-SEM. Second, SEM-PLS does not require strict data with normal distribution and is suitable for single-item measurement [106]. In addition, SEM-PLS is considered for the data analysis because it can cover complex and many constructs [105]. Thus, based on these explanations, this study used exploratory factor analysis with SEM-PLS.

There are three indicators to measure the construct in SEM-PLS: (1) composite reliability (C.R.), the value of C.R. should be more than 0.7 [107]; (2) AVE (Average Variance Extracted), the value of AVE should be more than 0.5 [108]; and the last indicator, (3) the value of factor loading is more than 0.6 [107]. In addition, descriptive statistics were used to describe the characteristics of the respondents, such as age, education, etc. Furthermore, the Smart PLS v3.2.6 package [109] was used to examine the structural model for the outer (validity and reliability testing) and inner model (path coefficient), and SPSS was used to analyze descriptive data in this study.

4. Result and Discussion

4.1. The Demographic of Respondents

The respondents in this study were Indonesian tourists who intended to visit the destinations in the post-earthquake disaster in Lombok, West Nusa Tenggara, Indonesia, that have been affected by the disaster since 2018. General information must be filled out in this study by the respondents. The general information of respondents refers to gender, age, education, income, and occupation. The proportion of respondents was divided into two types of gender: male and female, based on

Table 1. Demographics of Respondents.

Variable	Responses	Frequency	Percentage
Gender	Male	142	68.3
	Female	66	31.7
Age (year)	18–24 years	58	27.9
	25–34 years	117	56.3
	35–44 years	26	12.5
	45–54 years	6	2.8
	More than 55 years	1	0.5
Education	Senior High School	24	11.5
	Bachelor	88	42.31
	Master	86	41.3
	Doctoral	6	2.9
	Vocational	4	1.9
Income monthly	Less than IDR 1,000,000.	44	21.1
	IDR 1,000,000–IDR 5,000,000.	115	55.3
	IDR 5,000,000–IDR 10,000,000.	37	17.8
	IDR 10,000,000–IDR 15,000,000.	5	2.4
	IDR 15,000,000–IDR 20,000,000	7	3.4
Occupation	Students	58	27.9
	Full-time workers	52	25
	Part-time workers	19	9.1
	Household	8	3.8
	Businessman	5	2.4
	Professional	23	11.1
	Others	43	20.7

Source: Data analyzed.

of respondent demographic information. Most of the respondents in this study are male (68.3%). The majority of respondents in this study (56.3%) are between the ages of 26 and 34. Table 1 shows the respondents’ profiles.

4.2. Statistic Descriptive

Details on how statistics were applied to determine the normality and skewness of the data connected to the study’s variables are provided in

Table 2. Descriptive statistics.

Variables	Mean	Median	Std. Deviation	Skewness	S.E of Skewness	Kurtosis	S.E of Kurtosis	Min	Max
WOM1	5.63	6	1.205	−0.882	0.169	0.756	0.336	1	7
WOM2	5.8	6	1.141	−1.015	0.169	1.14	0.336	1	7
WOM3	5.78	6	1.067	−0.703	0.169	0.072	0.336	2	7
WOM4	4.71	5	1.388	−0.38	0.169	−0.069	0.336	1	7
e-WOM1	5.15	5	1.59	−0.882	0.169	0.368	0.336	1	7
e-WOM2	4.55	5	1.744	−0.393	0.169	−0.633	0.336	1	7
e-WOM3	4.81	5	1.607	−0.607	0.169	−0.185	0.336	1	7
e-WOM4	5.05	5	1.593	−0.615	0.169	−0.364	0.336	1	7
e-WOM5	5.28	6	1.51	−0.866	0.169	0.425	0.336	1	7
e-WOM6	5.19	5	1.618	−0.695	0.169	−0.204	0.336	1	7
e-WOM7	5.35	6	1.375	−0.833	0.169	0.601	0.336	1	7
R.P. 1	3.39	3	1.671	0.28	0.169	−0.766	0.336	1	7
R.P. 2	3	3	1.596	0.419	0.169	−0.637	0.336	1	7
R.P. 3	3.23	3	1.449	0.398	0.169	−0.195	0.336	1	7
R.P. 4	5.37	6	1.429	−0.774	0.169	0.128	0.336	1	7
VI 1	5.32	5	1.365	−0.715	0.169	0.394	0.336	1	7
VI 2	5.58	6	1.346	−0.946	0.169	0.444	0.336	1	7
VI 3	5.67	6	1.147	−0.798	0.169	0.69	0.336	1	7

Source: Author’s Analysis.

. The table below displays the findings of the study’s statistical analysis. According to Table 2, all factors had scores ranging from 1 to 7, with the mean falling between 3 and 5.58. The skewness value was employed to assess if the data in this study were normally distributed, and it should be less than 3 for each variable [110]. We can infer that the study’s data was regularly distributed.

4.3. Measurement and Common Method Bias

Exploratory factor analysis was used to test the questionnaire with Smart-PLS software. The threshold of factor loading greater than 0.70 for each item of the question [111]. According to [112], exploratory factor analysis was examined for common method bias. To certain whether the sample size was adequate for factor analysis, the KMO (Kaiser–Meyer–Olkin) test was applied. The findings demonstrate that every magnitude on the diagonal of the matrix was more than 0.5, and the KMO coefficient value was 0.832, which satisfied the requirement [113]. Second, a sample screen test was run while the Kaiser–Guttman criteria were employed to ascertain the number of variables. By using Harman’s one-factor analysis, the model explained roughly 20.1% of the variation with a single factor that was less than 50% [114]. Based on [115], a marker variable (unmeasured) was added to test for common technique bias. Thus, it was determined that common technique bias was not a problem.

4.4. Multicollinearity

According to [116], one helpful method for identifying the presence of multicollinearity among independent variables is to evaluate the Variance Inflation Factor (VIF). According to the regression analysis’s findings (see

Table 3. Variance Inflation Factor.

Variable	VIF
RP1	1.827
RP2	1.756
RP3	2.226
WOM1	2.033
WOM2	1.855
WOM3	1.171
eWOM1	2.074
eWOM2	2.120
eWOM3	2.145
eWOM4	1.768
eWOM5	3.438
eWOM6	2.723
eWOM7	2.602
V1	1.267
V2	2.095
V3	2.181

Source: Author’s Analysis.

), the VIF ranged from 1.0 to 3.615, indicating values between one and five [117]. The multicollinearity was not an issue in this investigation.

4.5. Validity and Reliability

Validity and reliability are two processes to examine the outer loading in the structural equation model. Three indicators are used to assess the construct’s reliability. The first is that the factor loading value should be greater than 0.70 [111].

Table 4. Factor loading, Cronbach’s alpha, C.R. and AVE.

Indicators	Factor Loading	AVE	CR	Cronbach Alpha	Note
Word of Mouth
When I want to travel to Lombok’s tourist destination, I will feel safe because I follow the advice given by friends or tourists.	0.856
2. When I want to visit the Lombok tourist destination in the post-earthquake disaster, I will hear recommendations from friends/tourists.	0.738
3. Lombok island is safe for travelers in the post-earthquake disaster.	0.760	0.618	0.829	0.700
4. I decided to advise my friends or relatives to visit Lombok tourist destination post-earthquake disaster, because safe from the disaster.					Deleted
e-WOM
I might visit a destination in Lombok in the post-earthquake disaster because my friends uploaded Lombok’s destination on Instagram.	0.717
2. I might visit a destination in Lombok in the post-earthquake disaster because my friends uploaded Lombok’s destination on Facebook.	0.718
3. I plan to visit a destination in Lombok post-earthquake disaster after watching Lombok’s destination on YouTube.	0.770
4. I often read online travel reviews of tourist destinations to make a good impression on myself and others.	0.702	0.577	0.905	0.877
5. I often read online travel reviews to convince myself of the right choice for Lombok’s destination post-earthquake disaster.	0.846
6. I often collect information from tourist’s online reviews before traveling to Lombok tourist destination post-earthquake disaster.	0.749
7. When I travel to Lombok’s tourist destination, online travel reviews make me confident to travel to Lombok destination in post-earthquake disasters.	0.802
Risk perception
Your friends will worry about your safety when traveling in Lombok post-earthquake disaster.	0.857
2. Based on your experience, Lombok’s tourist destination is the high risk for tourists.	0.843	0.739	0.895	0.824
3. Your friends or relatives see or assume that Lombok’s tourist destination is a dangerous place to visit.	0.879
Visit Intention
I want to visit Lombok’s destination rather than others.	0.764
2. I predict that I plan to visit Lombok’s destination in the future post-earthquake disaster.	0.837	0.682	0.865	0.765
3. I will visit Lombok’s destination in the future post-earthquake disaster.	0.872

shows that the factor loading value is greater than 0.70 (ranging from 0.702 to 0.879). The composite reliability and Cronbach’s alpha values should be greater than 0.70 [107,108]. Table 4 shows that the composite reliability and Cronbach’s alpha values were above 0.7 (ranging from 0.829 to 0.905 and from 0.700 to 0.877, respectively). Furthermore, the average variance extracted (AVE) values should be greater than 0.5 [108]. The results of AVE in Table 4 show that the values are greater than 0.50 (ranging from 0.577 to 0.739).

In this study, the discriminant validity was measured using three methods. First, we examined the Fornell–Larcker Criterion by comparing the square root of the AVE for each construct to the inter-construct correlation value. According to the researchers, the expected value of the AVE root should be greater than or equal to 0.70 [118]. As shown in

Table 5. Inter construct correlations and square Roots (Fornell–Larcker Criterion).

	Risk Perception	Visit Intention	WOM	e-WOM
Risk perception	0.860
Visit intention	0.065	0.826
WOM	−0.229	0.500	0.786
e-WOM	0.141	0.567	0.501	0.759

Note: square root of AVE is shown in the diagonal and in bold. AVE: Average variance extracted.

, the discriminant validity was accepted, as were all the diagonal elements, for which the square root of AVE exceeds the inter-construct correlation. Thus, the discriminant validity can be reached.

Second, we used the cross-loading matrix approach to measure discriminant validity in

Table 6. Cross Loading Matrix.

	Risk Perception	Visit Intention	WOM	e-WOM
RP1	0.857	0.093	−0.148	0.195
RP2	0.843	0.048	−0.221	0.102
RP3	0.879	0.022	−0.227	0.057
V1	0.000	0.764	0.430	0.511
V2	0.084	0.837	0.370	0.396
V3	0.083	0.872	0.428	0.481
WOM1	−0.101	0.451	0.856	0.497
WOM2	0.071	0.355	0.738	0.394
WOM3	−0.401	0.367	0.760	0.305
eWOM1	0.201	0.454	0.216	0.717
eWOM2	0.051	0.424	0.290	0.718
eWOM3	0.118	0.420	0.262	0.770
eWOM4	0.115	0.338	0.376	0.702
eWOM5	0.128	0.516	0.442	0.846
eWOM6	0.071	0.371	0.489	0.749
eWOM7	0.049	0.452	0.609	0.802

. The result found that all of the constructs of factor loadings were more significant than the correlation coefficient with other constructs. Thus, the discriminant validity was acceptable in this study.

The last approach is based on the Heterotrait–Monotrait Ratio (HTMT), representing the new technique for testing the discriminant validity. In order to assess the HTMT value [119], the used A threshold value of lower than 0.90 was suggested. Table This study found that the value of HTMT is lower than 0.90, which indicates that the study has strong discriminant validity.

Table 7. Heterotrait–Monotrait Ratio.

	Risk Perception	Visit Intention	WOM	e-WOM
Risk perception	-
Visit intention	0.109	-
WOM	0.317	0.671	-
e-WOM	0.161	0.677	0.649	-

shows the discriminant validity HTMT ratio.

4.6. Hypothesis Testing

This study used Structural Equation Modeling-Partial Least Squares (SEM-PLS) to examine the research framework and hypotheses. SEM-PLS has two perspectives: the standardized path coefficient and the explanatory model with R [108,111]. The standardized path coefficient refers to statistical significance, and the R coefficient is determined to explain the capability of the independent variable toward the dependent variable.

This study used the Smart PLS v3.2.6 package [109] to examine the structural model, and bootstrapping with 5000 samples was performed to assess the path coefficient of the study. The research model was measured with an R-value to explain how the variance of an independent variable can explain the dependent variable. The explanatory capability of R square is to determine the framework of the study R [108].

Table 8. Result of the Structural Equation Model.

	Original Sample (Estimated β)	T Value	p Values	R2	Decision
WOM → Risk perception	−0.401	4.552	0.000 ***	0.140	Supported
e-WOM → Risk perception	0.342	5.348	0.000 ***		Supported
WOM → Visit intention	0.321	3.714	0.000 ***	0.389	Supported
Risk perception→ Visit intention	0.083	1.579	0.114		Not supported
e-WOM → Visit intention	0.394	4.507	0.000 ***		supported

Note: *** Significant at the p < 0.05 level (two-tailed).

shows the result of the structural equation model of the study. Table 8 presents that Hypothesis 1 was supported in this study (β_WOM → _{Risk perception} = −0.401, p = 0.000). Hypothesis 2 was accepted or significant (β_e-WOM → _{Risk perception} = 0.342, p = 0.000). Hypothesis 3 was accepted in this study (β_WOM → _{visit intention} = 0.321, p = 0.000). Furthermore, this study did not support the hypothesis of (β_{risk perception} → _{visit intention} = 0.083, p = 0.114). The last hypothesis in this study (β_e-WOM → _{visit intention} = 0.394, p = 0.000) or the hypothesis was supported in this study. Therefore, the variance explained or R square by presence word of mouth, electronic word of mouth, and risk perception were 0.140, and WOM, e-WOM, and visit intention were 0.389, respectively. Figure 2 shows the result of PLS-SEM.

4.7. Discussion

This study applied SOR theory from the perspective of tourism destination marketing through information sources to provide an understanding of tourist behavior in post-earthquake disasters. The followings are the key findings:

First, this study examined the effect of word of mouth (WOM) on visit intention in the post-earthquake disaster as a “Stimulus” in SOR theory. The tourist’s stimulus through word of mouth significantly influenced visit intention. This result indicated that word of mouth is one of the marketing strategies that can influence tourists to visit a destination post-disaster. In addition, the result implied that word of mouth played an essential role in the promotion of a destination to influence tourists’ visit intention. Some scholars have examined the influence of word of mouth on visit intention, such as [39,97], who stated that word of mouth (WOM) influences tourists’ travel intention. Word of mouth is a mechanism to shape behavioral intention in crisis [99]. In times of crises or post-disaster, promoting the destination through word of mouth is one of the ways to make it convenient for tourists to visit a destination. This study’s result, supported by scholars [39,97], revealed that word of mouth influences tourists’ travel decisions. The result emphasizes that word of mouth is thriving as destination marketing post-earthquake disaster to gain the attention of tourists’ visit intention. Promoting the destination through word of mouth is a way to promote the destination in the post-disaster period, such as the destination being safe from disaster, giving the recommendation to visit a destination, giving advice to friends or relations to visit Lombok’s safe destination. These are the agenda of the tourism destination promotion that should be emphasized by all stakeholders to all visitors through word of mouth to visit Lombok’s destination.

Furthermore, this study examined word of mouth influences risk perception. The finding indicated that word of mouth significantly and negatively impacted risk perception. The result of this study indicates that the credibility, trust, and biases of information given through “word of mouth” impact negative risk perceptions for destinations. The negative perception of tourist destinations in the post-earthquake disaster influenced tourists to visit a destination. This study is in line with the study from [120] that stated that information biases influenced negative risk perception.

Second, the study found that electronic word of mouth had a positive and significant influence on visit intention in the aftermath of the earthquake disaster in West Nusa Tenggara, Indonesia, which confirms that this research conclusion was supported by [40,121] revealed that electronic word of mouth has an impact on visit intention. The result of this study was in line with [122], stating that e-WOM significantly impacts travel intention. In promotion, electronic word of mouth is easier and more reliable to access than other conventional methods for promoting and marketing the destination post-disaster among social media users such as Facebook, Instagram, YouTube, online travel reviews, etc. Several scholars, including [122,123], found that e-WOM influences tourists’ preferences for destinations in the tourism industry. Also, electronic word of mouth is one of the sources in promoting and restoring the destination image in crises or disasters such as natural disasters, political instability, and social crises (terrorism risk). It can increase the tourist visit intention [124,125].

The result implied that electronic word of mouth is a medium to promote the destination in the case of post-earthquake disaster through social media such as Facebook, Instagram, YouTube, online travel review, website travel, and other platforms to gain attention from the visitors. The promotion through electronic word of mouth can be conducted by the stakeholders (governments, destination marketing organizations, and tourism marketers) to promote Lombok’s destination with social media content to inform the tourist that the destination is safe. With numerous promotions of the destination through electronic word of mouth (social media platforms) contain tourist information, promotion, and regular information in the news, official media, and tourist destination information on behalf of Lombok’s tourism. Furthermore, the significance of destination promotion by the governments, tourism marketers, and destination management organizations toward the Lombok West Nusa Tenggara destination is considered more credible information and trusted by tourists who intend to visit a destination. It was supported by the study from [124], which stated that e-WOM (social media, online review, online travel website) was highly effective in reducing the negative perception of crises such as political instability and terrorism.

In addition, the result found that the e-WOM had a positive and significant effect on risk perception. The dimensions of electronic word of mouth, such as information credibility, trust, and objectivity, influenced tourists’ perceptions of risk. The study from [126] supported this result of the study revealed that the e-WOM influences the perceived risk. In addition, electronic word of mouth (e-WOM) can mitigate risk perception and positively influence the tourists’ visit intention during crises [125]. In addition, the online campaign using e-WOM through social media and online travel bloggers can reshape the destination image in the following crises and disasters, influencing tourists’ visit intention [125]. It was also supported by [124], revealed that e-WOM, such as social media (Facebook, Instagram, YouTube, online travel reviews), online travel bloggers, and online travel communication effectively decrease or elevate the tourists’ risk or negative perception of political and terrorism crises toward destination image. It is similar to a study that suggests that tourists’ reviews on e-WOM may also increase the positive perception of tourists toward tourist destinations [96].

Furthermore, the last hypothesis is the effect of risk perception on visit intention in post-earthquake disasters. The result found that risk perception did not influence a tourist’s decision to visit a destination during a crisis or post-earthquake disaster. According to this study, tourists’ desire to visit Lombok’s destination after the disaster is unaffected by their perception of disaster risk (earthquake disaster). While traveling during post-earthquake disasters in Indonesia, it is indicated that tourists know about disasters. Based on the study’s findings, the Indonesian government attempted to prevent disasters in the destination area through disaster management for tourists. In addition, the tourists believed that Lombok’s destination was safe for travel in the post-earthquake disaster. The result of this study was in contrast with those [73], who stated that tourists tried to avoid high-risk destinations while traveling and chose less dangerous destinations. The risk perception in disasters such as earthquakes, volcanic eruptions, disease (pandemic or endemic), terrorism, political instability, and others with high perceived risk while traveling is one indicator of a tourist’s decision to visit a destination during the disaster period. Several studies also have been conducted related to risk perception during a disaster, such as from [127,128,129,130] found that risk perception influences tourists’ decision to visit a destination. Moreover, the result of this study implied that the marketing campaign of the destination as a response from the stakeholders (governments, destination marketers, destination management organizations) to the tourist risk perception on tourist destinations in post-earthquake disaster to initiate a destination crisis campaign program to reshape or reestablish tourist’s trust toward a tourism destination. Several agendas to campaign for tourism destinations post-earthquake disasters include disaster management systems, disaster mitigation in tourist destinations, disaster information, and disaster awareness and disaster preparedness.

This study views risk perception as an organism within the SOR theory. The study discovered that risk perception did not significantly impact tourists’ visit intention in the aftermath of an earthquake in Indonesia. Risk perception in this study represents how tourists feel while traveling to their destination following a disaster. A framework or construct such as emotions or feelings [44,48,49], customer value [50], and corporate image [51] have been used to represent the aspect of the organism in consumers in empirical studies on the tourism and hospitality industries.

Furthermore, the study’s findings emphasize that marketing tourist destinations post-earthquake disasters through electronic word of mouth and word of mouth as a stimulus in the SOR theory has successfully impacted tourist visit intention in the case of a post-earthquake disaster in Indonesia. As a stimulus in the aftermath of the earthquake, information sources such as e-WOM and word of mouth are the primary means of promoting the destination and attracting tourists. For example, promoting a destination using e-WOM on social media platforms (Facebook, Instagram, WhatsApp, online travel reviews, etc.) can increase tourists’ interest in visiting a destination in the aftermath because the tourists believe that the destination is safe. Moreover, in this study, word of mouth as a “stimulus” can impact local tourists’ intentions. These information sources effectively stimulate visit intention after the earthquake disaster.

4.8. Theoretical and Practical Implications

The study’s findings highlight the importance of understanding tourists’ behavior toward visit intention in the aftermath of Indonesia’s post-earthquake disaster. The main theoretical implication in this study has several theoretical implications. First, this study provides insight into applying the SOR theory in the case of a post-disaster, especially a post-earthquake disaster, by introducing word of mouth and electronic word of mouth (as a stimulus), risk perception (as an organism), and visit intention (as a response). Second, the findings of this study can enrich the theories of word of mouth and electronic word of mouth in the time of disaster or aftermath. As the central variables, word of mouth and electronic word of mouth have a central role in risk perception and visit intention. In addition, the study’s findings can enrich and extend the marketing theories for tourism literature in times of disaster or crisis. Furthermore, the findings articulated that the risk perception of disaster does not influence tourist behavior. This result can enrich the contribution of the literature in psychology and tourism that the tourists do not consider risk perception because the destination was safe in the post-disaster case.

Based on these study findings, the empirical findings show that word of mouth and electronic word of mouth plays a central role in tourist visit intention post-disaster. It can be stated that WOM and e-WOM marketing as tools for promotion in the disaster to restore the destinations. The tourism marketing authorities, such as managers or marketers, and governments can consider the promotion through word of mouth and electronic word of mouth (social media, online travel reviews, online travel bloggers) for marketing ideas in the time post-disaster.

Based on the study’s findings, several policy recommendations for the stakeholders related to tourism marketing post-earthquake disaster can be made. First, the tourism marketers or managers can provide gifts, bonuses, and other incentives to encourage the tourists to spread the information about Lombok, Indonesia destination through word of mouth, like the promotion “buy one ticket, gain the free stay one night at the destination or free tours for one day.” Second, the tourism manager can effectively apply the promotion using electronic word of mouth through social media such as Facebook, Twitter, Instagram, YouTube, and online travel reviews. These platforms can draw attention to a location by informing tourists that it is safer after an earthquake disaster. The tourism authorities also can collaborate with the “traveling influencers or famous influencers” who have more than one million followers on Instagram and Facebook fan pages to promote Lombok, Indonesia destination that the destination is safe and can enjoy the destination. Third, stakeholders in tourism, such as managers, marketers, and the government, can use social media (Tik-Tok, Micro-Blogs, What-Apps, and Instagram) to increase the information about Lombok as an Indonesian destination.

5. Conclusions, Limitations and Future Research Agenda

Using the SOR theory, this study examined the role of destination marketing through information sources to understand tourist behavior in a post-earthquake disaster. This research then focuses on tourist behavior in the aftermath of an earthquake in Indonesia. Beyond the study, this paper provides information for stakeholders, including the government, marketing strategies in tourism, and the tourism and leisure industries, which might read this paper to recognize tourist behavior in the case of a post-earthquake disaster in Indonesia. Several conclusions from this study, including the following:

This study gives knowledge of the importance of the SOR theory approach to shaping tourist behavior to visit a destination in the case of a post-earthquake disaster in Indonesia. From the perspective of information sources such as e-WOM and WOM, they can influence tourist behavior to visit a destination in a post-earthquake disaster. The result implies the importance of electronic word of mouth in providing information to tourists through several social media platforms. Furthermore, in this study, the perception of risk was not considered a factor in a decision to visit a destination in the event of a disaster. It implies that tourists know about disasters while traveling in post-disaster situations.

It should be noted that this study has some limitations that are to be highlighted. First, this study or the previous studies have focused on the effect of word of mouth and electronic word of mouth, such as how the WOM and e-WOM work as destination marketing in the case of the disaster and the process of influencing the visit intention. However, only a few studies have been done in the context of the effectiveness of WOM and e-WOM as marketing destinations toward risk perception and visit intention. In addition, this study focuses on the kinds of information sources (WOM and e-WOM) that are the main variables; thus, it is hoped that the dimension of each variable can be explored for future agenda. Some dimensions can be explored, such as information credibility and accuracy, understandability of the information, and trust [131]. Second, from the methodology perspective, this study used an online survey via an online questionnaire (Google form) to collect the data. Respondents who completed the questionnaire in this study were small and may not be representative of all tourists in post-disaster in Indonesia. Moreover, this study focuses on the perception of Indonesian tourists toward one specific disaster (an earthquake disaster). It may not be generalized to the other disaster in Indonesia. In addition, this study focuses on tourists’ behavior in the case post-earthquake disaster several years ago from the framework of SOR theory and may give bias.

To that end, for future study, this study focuses on the role of information sources (word of mouth and electronic word of mouth) as a stimulus for tourists’ visit intention in the context of post-earthquake disasters. In the future, a similar could be conducted on tourist visit intention in the context of man-made disasters to see whether a different result would be reached. Second, future studies should investigate the relative efficacy of the information sources (word of mouth and electronic word of mouth) for tourist trust and willingness to visit a destination post-disaster. In addition, this study focuses on the factors that influence tourist visit intention by using WOM and e-WOM information. Further research can identify what formed tourist behavior and intention through various social media information platforms [132]. Third, this study also considered the effect of WOM and e-WOM on risk perception and visit intention. Future research into the quality of word of mouth and electronic word of mouth as an intervening model could be considered for future research with the different kinds of disaster or disaster groups.