The Impact of Psychological and Risk Factors on Tourists’ Loyalty Toward Nature-Based Destinations

Abstract

Tourist loyalty is vital for destination success, fostering repeat visits and positive word-of-mouth. This study explores the psychological and safety-related factors driving tourist loyalty to natural attractions in Oman, a rising destination known for its stability and safety. Using Social Cognitive Theory as a foundation, the research incorporates perceived risk and novelty seeking as key moderating variables. Data were collected via an online survey of 165 international tourists and analyzed through Partial Least Squares Structural Equation Modeling (PLS-SEM). Findings show that attachment, satisfaction, and novelty seeking significantly affect both attitudinal and behavioral loyalty. While perceived value strongly influences behavioral loyalty, its impact on attitudinal loyalty appears more complex, suggesting possible unobserved mediators. Additionally, risk perception and novelty seeking moderate the link between destination familiarity and loyalty, underscoring the role of tourists’ internal evaluations of safety and desire for new experiences. This study advances the limited literature on tourist loyalty in developing countries by integrating psychological and risk-related dimensions. It offers actionable insights for tourism planners and marketers in Oman: emphasizing the country’s safety reputation, improving satisfaction levels, and crafting experiences that blend familiarity with novelty can enhance tourist loyalty and ensure sustained competitiveness in the global tourism market.

1. Introduction

Enhancing the tourism sector is crucial for any country with natural attractions and heritage sites, as a thriving tourism industry offers significant socioeconomic benefits (Hall et al., 2025; Cossío-Silva et al., 2019). For example, Italy offsets its energy trade deficit through a surplus in its tourism trade balance, which highlights the strategic value of tourism as a compensatory economic asset (Alegre & Juaneda, 2006). Thus, it is imperative for countries to strategically develop tourism by leveraging their natural and historical assets. Successful tourism planning requires an in-depth understanding of various aspects of the sector, including tourism demand, which is essential for crafting effective strategies and policies (Tomić et al., 2019).

While traditional demand analysis often focuses on economic indicators such as income, prices, and exchange rates, less attention has been given to critical psychological factors like travel experience, frequency, and intensity (Aarabe et al., 2025; Crouch, 1992). These factors, however, play a pivotal role in shaping future travel decisions, yet they remain underexplored in tourism demand forecasting. Exploring psychological elements that influence tourists’ behavioral (intention to revisit) and attitudinal loyalty (recommending and giving positive feedback) can significantly enhance the ability to predict future tourism demand (Oppermann, 1999). One effective measure of tourism demand is the duration of stay, which is closely linked to destination loyalty (Alazaizeh et al., 2024; Cossío-Silva et al., 2019).

In developing countries like Oman, there is a notable gap in foundational research on loyalty factors, which are critical for precise tourism demand estimation (Al Mahruqi, 2023). This study seeks to fill that gap by investigating the psychological factors that impact tourists’ loyalty to Oman’s natural attractions, thereby providing insights into key drivers of loyalty. Additionally, the study examines the important role of safety and security in fostering tourist loyalty, particularly in light of the political unrest and conflicts in the surrounding region. Oman’s reputation as a safe and stable destination enhances tourists’ confidence and willingness to revisit, making the examination of risk factors a secondary objective of the study. Moreover, the study delves into how certain moderator variables, such as novelty-seeking behavior and risk perception, influence the relationship between familiarity with a destination and tourist loyalty. Unlike previous studies that primarily focused on the negative impact of social unrest on tourism, this research extends the scope by evaluating how political neutrality in a country located in an unstable region can positively affect tourism growth. It is one of the first few studies to explore how Oman’s unique characteristics mitigate regional risks, fostering long-term tourism development.

This study employs Partial Least Squares-Structural Equation Modeling (PLS-SEM) as an appropriate tool for examining complex relationships between psychological and risk factors influencing tourist loyalty. PLS-SEM is well-suited for examining intricate variable relationships, making it ideal for research in a region like Oman where foundational tourism data is limited. The findings offer practical insights for decision-makers in the tourism sector, particularly for tourism agencies and private enterprises, by identifying strategies to attract tourists and increase return visits despite regional instability.

In summary, this study investigates the influence of psychological and risk factors on tourists’ attitudinal and behavioral loyalty to nature-based destinations in Oman, with the aim of informing more effective tourism development strategies. This study addresses a key theoretical gap by examining how novelty seeking and risk perception moderate the relationship between familiarity and tourist loyalty, a mechanism largely overlooked in prior research. By doing so, it extends Social Cognitive Theory (SCT) to tourism contexts in which safety perceptions and the search for novelty critically shape loyalty outcomes. Moreover, the research uncovers valuable insights that can guide decision-making and policy formulation in Oman’s tourism sector, contributing to the country’s long-term tourism success.

The remainder of this paper is structured as follows. Section 2 reviews the theoretical foundations and previous research on psychological and risk-related factors influencing tourist loyalty. Section 3 introduces the research model and hypotheses. Section 4 describes the methods and materials. Section 5 presents the results of the analysis. Finally, Section 6 discusses the findings and concludes with the main contributions and policy implications.

2. Literature Review

2.1. Theoretical Framework

In exploring tourist loyalty to nature-based destinations, several foundational economic and behavioral theories offer valuable explanatory power. The conceptual foundation of this study is grounded in established frameworks of consumer behavior and tourism loyalty, with particular emphasis on SCT (Bandura, 1991). To strengthen this foundation, we explicitly map SCT constructs onto the study variables and clarify why SCT provides the most appropriate lens. SCT highlights how behavior is shaped by the reciprocal interaction of cognition, environment, and behavior. In this study, satisfaction and attachment reflect reinforcement and observational learning, while perceived value corresponds to outcome expectations that guide decision-making. Familiarity is linked to accumulated knowledge that enhances self-efficacy in travel choices. The moderators, novelty seeking and risk perception, represent cognitive evaluations that alter the strength of familiarity’s effect on loyalty, with novelty seekers discounting repetitive experiences and risk-averse tourists interpreting familiarity as a safety signal. Compared to alternative frameworks such as the Theory of Planned Behavior, which focus primarily on intention, SCT better captures the dynamic interplay between psychological states, environmental context, and behavioral outcomes, making it especially well-suited to explaining tourist loyalty in the context of Oman’s nature-based tourism.

Loyalty itself is commonly conceptualized along two dimensions: attitudinal and behavioral. Attitudinal loyalty encompasses tourists’ emotional attachment to a destination and their inclination to advocate for it, whereas behavioral loyalty pertains to the actual intention or decision to return (Whitehead & Wicker, 2025; Zeithaml, 2000; Oppermann, 1999). These dimensions are critical to understanding repeat visitation patterns and word-of-mouth promotion. Furthermore, aligned with the Theory of Planned Behavior (Ajzen, 1991), psychological constructs such as satisfaction, emotional attachment, perceived value, and familiarity are identified as principal antecedents of loyalty (Mechinda et al., 2009). Satisfaction denotes the extent to which a destination meets or exceeds visitor expectations, while attachment reflects the affective bond formed between the tourist and the place (Williams & Vaske, 2003). Perceived value refers to the perceived balance between the benefits received and the costs incurred (Cronin et al., 2000). Additionally, this study incorporates elements from Risk Perception Theory, which highlights the role of perceived safety and political conditions in shaping travel behavior (Fletcher & Morakabati, 2008). In this regard, Oman’s image as a politically stable and secure destination is considered a vital influence in cultivating both attitudinal and behavioral loyalty, particularly within a region marked by geopolitical uncertainty.

2.2. Loyalty in the Context of Tourism

In contrast to brand loyalty, destination loyalty research is relatively new phenomenon. Since the late 1980s, loyalty has been used by several researchers to indicate a positive experience toward a travel destination. It can be helpful in predicting revisits and future revenue (Oppermann, 1999). However, even destinations with a high reputation and strong branding may have trouble figuring out what factors, other than novelty-seeking behaviors, influence loyalty (Lv et al., 2020). Tourism loyalty is defined as a tourist’s inclination to return to a destination, give it positive feedback, and suggest it to other travelers. In other words, the concept of “tourist loyalty” refers to the future behavioral intentions of visitors as influenced by their travel experiences (Al Mahruqi, 2023; T. H. Lee & Hsu, 2013). Loyalty can be divided into two types: attitudinal and behavioral (Zeithaml, 2000). While attitudinal loyalty is associated with positive feelings toward the place and willingness to recommend it, behavioral loyalty goes beyond that (Fajriyati et al., 2022). Behavioral Loyalty is reflected by intentions to revisit the place (Fajriyati et al., 2022). Previous studies have investigated the impact of six psychographic variables in addition to demographic variables on the loyalty of tourists (Mechinda et al., 2009; Al Abri et al., 2023b). The six psychological variables included satisfaction, perceived value, attachment, familiarity, novelty seeking and perceived risk.

2.2.1. Satisfaction

Satisfaction in the broader context is a vital element for maintaining long-term relationships with customers (Ardani et al., 2019). In the context of tourism, satisfaction is found to be a strong determinant of tourism destination loyalty. The attractiveness level of the destination is one of the factors that influence tourist satisfaction (Mlozi & Pesämaa, 2013; Morais & Lin, 2010). The attractiveness of a destination is a combination of natural attractions and man-made attractions. For some tourists, the availability of culturally appropriate food options, prayer facilities, and other religious accommodations positively influences their overall satisfaction (Sánchez-Sánchez et al., 2025; Miguel et al., 2025; Rahman, 2014; Cheng et al., 2018). Kim and Park (2017) suggest that highly satisfied tourists tend to have loyalty toward the destination. Based on this literature, we hypothesize that satisfaction positively influences both attitudinal and behavioral loyalty (H5 and H6).

2.2.2. Perceived Value

Perceived value is defined as the gap between the benefits gained and the sacrifices made by the consumer (Cronin et al., 2000). The perceived value is subjective, meaning that each tourist has a different attitude toward the good consumed or service provided. For instance, some tourists appreciate lower prices more than quality, while others appreciate quality more. Previous research suggests that perceived value influences future intentions and loyalty (Ozturk et al., 2016; Bojanic, 1996). It has been confirmed that there is a strong link between loyalty, profits, and the value provided to customers (Salem Khalifa, 2004). In addition, researchers also found that perceived value affects satisfaction (El-Adly, 2019; Slack et al., 2020; Elshaer et al., 2025). Accordingly, we propose that perceived value positively influences both attitudinal and behavioral loyalty (H7 and H8).

2.2.3. Attachment

Attachment to a destination is another factor influencing loyalty. It is connected to tourists’ emotions and memorable experiences. Tsai (2012) mentioned that tourists form emotional bonds between themselves and the destinations they visit. In addition, attachment is a strong predictor of behavioral loyalty (Mechinda et al., 2009; Huang & Lin, 2023; Vera et al., 2025). Some researchers used a two-dimension model of place attachment, which includes place identity and place dependence (Williams & Vaske, 2003). Place identity describes a strong bond between a person’s identity and a place (Proshansky, 1978). Place dependence recognizes a place’s capacity to satisfy a person’s functional demands, which may not always be fully satisfied by another destination (Stokols & Shumaker, 1981). Attachment is also considered as a consequence of past visit satisfaction (Alegre & Juaneda, 2006). Thus, we hypothesize that attachment exerts a positive effect on both attitudinal and behavioral loyalty (H1 and H2).

2.2.4. Familiarity

Information and knowledge about the destination are referred to as familiarity. G. Lee and Tussyadiah (2012) showed that familiarity is a multidimensional construct that has many determinants including general knowledge and previous visit experiences. Even though familiarity can affect destination choice positively and negatively, it remains a major factor in predicting tourist plans (G. Lee & Tussyadiah, 2012). While Casali et al. (2021) remarked that the significance of familiarity in shaping tourists’ loyalty remains uncertain, Mechinda et al. (2009) found that familiarity is the strongest predictor of behavioral loyalty of tourists in Thailand, followed by attachment. Familiarity also affects tourists’ satisfaction and loyalty through its impact on the overall image of the destination (Wang et al., 2022). Building on these insights, we posit that familiarity positively influences attitudinal and behavioral loyalty (H3 and H4).

2.2.5. Novelty Seeking as Moderator

Tourists have different motivations for traveling. Literature on tourism identified various push and pull motivating factors that influence the choices of tourists including the search for novelty (Manhas et al., 2025; Blomstervik et al., 2021; Carvache-Franco et al., 2022, 2023; Pestana et al., 2020). The concept of novelty pertains to a deviation from one’s ordinary routine or lifestyle through the exploration of new and diverse experiences, suggesting a tendency toward seeking out variety (T. H. Lee & Crompton, 1992; Mitas & Bastiaansen, 2018). Tourists who seek novelty and change are less likely to be loyal to a particular destination. As a result, they are less interested in revisiting the same destination (Alegre & Juaneda, 2006). Given that novelty seekers are less likely to revisit familiar destinations, we hypothesize that novelty seeking negatively moderates the relationship between familiarity and both attitudinal and behavioral loyalty (H11 and H12).

2.2.6. Perceived Risk as Moderator

The political environment, which encompasses both the regulatory environment, political instability, and terrorism, has a negative impact on the development of the tourism business (Fletcher & Morakabati, 2008; Mohammed et al., 2022; Richter & Waugh, 1986; Saha & Yap, 2014). Some researchers argue that political stability is required for tourism to thrive (Drakos & Kutan, 2003). Political disturbance and riots can sometimes result from political instability, which has a negative influence on tourism (Drakos & Kutan, 2003; Mohammed et al., 2022). Richter and Waugh (1986) also demonstrated that the consequences of political instability on tourism are significantly greater than the effects of isolated terrorist incidents. Furthermore, the geographical location of terrorist acts can have an impact on surrounding nations, resulting in a decrease in tourism demand (Dann, 1977; Y. Yang & Wong, 2012). Recent post-pandemic studies highlight that global crises such as COVID-19 have intensified tourists’ sensitivity to safety and health risks, further reinforcing the importance of risk perception in shaping travel intentions and loyalty (Altınay & Kozak, 2021; Cambra-Fierro et al., 2021; Santos et al., 2024). Since perceived risk can diminish tourists’ confidence and loyalty, we hypothesize that risk perception negatively moderates the relationship between familiarity and both attitudinal and behavioral loyalty (H9 and H10).

3. Research Model and Hypotheses

Figure 1 depicts the hypothesized relationship between the constructs. The hypothesized relationships were built based on the literature on tourism loyalty, which was reviewed in the previous section. Previous studies have addressed different factors influencing tourists’ loyalty. This study examines the impact of satisfaction, familiarity, attachment, and value perception, as studied by previous authors in different destinations. Having that familiarity can affect destination choice positively or negatively (G. Lee & Tussyadiah, 2012), so this study investigates the moderating impact of novelty seeking and risk perception on familiarity to explore the circumstances in which familiarity is not a positive predictor of loyalty.

The arrows pointing to the squares reflect the impact on the dependent variables by the independent variables (path coefficient). Since familiarity can affect destination choice positively and negatively (G. Lee & Tussyadiah, 2012), this research assumes that in the presence of risk and the search for novelty, familiarity tends to have a weaker impact on loyalty. Therefore, novelty seeking and risk perception were introduced as moderators for familiarity to reflect this assumption. A moderator variable is a variable that affects the strength or direction of the relationship between two other variables (Hair et al., 2021). Based on the literature and the above discussion, this research has the following hypotheses:

H1: 

Attachment influences attitudinal loyalty positively.

H2: 

Attachment influences behavioral loyalty positively.

H3: 

Familiarity influences attitudinal loyalty positively.

H4: 

Familiarity influences behavioral loyalty positively.

H5: 

Satisfaction influences attitudinal loyalty positively.

H6: 

Satisfaction influences behavioral loyalty positively.

H7: 

Perceived value influences attitudinal loyalty positively.

H8: 

Perceived value influences behavioral loyalty positively.

H9: 

Risk perception negatively moderates the relationship between familiarity and attitudinal loyalty.

H10: 

Risk perception negatively moderates the relationship between familiarity and behavioral loyalty.

H11: 

Novelty-seeking negatively moderates the relationship between familiarity and attitudinal loyalty.

H12: 

Novelty-seeking negatively moderates the relationship between familiarity and behavioral loyalty.

4. Materials and Methods

The study focuses on Oman given its historical and natural resources that are major aspects of the country (Henderson, 2015; Al Abri et al., 2023c; Al Ismaili et al., 2024) as well as the country’s neutral political stance in the Middle East. Oman, like other oil-dependent countries in the MENA region, is working to diversify its economy as outlined in its Vision 2040 strategy (Al-Abri et al., 2019). A key pillar of Oman’s 2040 plan is substantial investment in the tourism sector, focusing on sustainable tourism to attract global visitors. Nature-based tourism is a potential in Oman which includes, for example, activities such as exploring the diverse landscapes of deserts, mountains, and wadis, visiting nature reserves, observing wildlife like the Arabian Oryx and sea turtles, and participating in eco-friendly adventures in its protected natural and marine areas.

A questionnaire was employed to collect the information. The first two sections assessed loyalty factors such as repeat visits, duration spent, and intentions to recommend Oman to others. The questionnaire also included questions assessing factors that are considered exogenous (independent) variables influencing tourists’ loyalty: satisfaction, attachment, perceived value, familiarity as proposed by Mechinda et al. (2009) and Al Abri et al. (2023b) where the latter focused on Oman. Along with that, the questionnaire included a section on novelty seeking to address the factors that urge tourists to travel. The question about risk-perception was derived from previous research (Rejan & Ahn, 2024; Fletcher & Morakabati, 2008; Mohammed et al., 2022; Powell, 2011; Tung & Ritchie, 2011; Y. Yang & Wong, 2012). In addition, a small number of items were self-developed to capture context-specific aspects of Omani tourism. These items were reviewed and refined during the pilot test to ensure clarity, content validity, and comprehensibility for both English- and Arabic-speaking respondents.

The respondents’ profiles were also covered in a section because demographic factors affect and influence travel choices. The questionnaire utilized a combination of yes-no, open-ended questions, and a five-point Likert scale. A statistical consulting firm in Oman administered the questionnaire. Data collection took place from November 2023 to March 2024, which corresponds to the winter season in Oman and coincides with the country’s peak tourist period. The survey was administered online through the consulting firm in collaboration with domestic travel and tourism agencies. Tourists who had recently visited Oman were contacted via email and online platforms using a random sample provided by these agencies. To ensure broad coverage, invitations were distributed at different times of the day and week, and follow-up reminders were issued to enhance the response rate. Both English and Arabic versions of the questionnaire were made available, and respondents could complete the survey on any device (PC, tablet, or smartphone). This approach helped capture a diverse sample of international and Arab tourists, enhanced accessibility, and reduced potential bias from time- or location-specific surveys. Since the purpose of this study is to examine overall tourism in Oman rather than a specific tourism location, interviewing tourists at specific locations during specific times would not have been fully representative.

The questionnaire was prepared and distributed in both English and Arabic to reflect the mixed profile of respondents, which included international as well as Arab tourists visiting Oman. This bilingual approach ensured clarity and accessibility, while also making the survey clear and comprehensible for non-native English speakers. In preparation for the actual data collection, a pilot testing phase was conducted to validate the relevance of the questions to the target population and to establish the dependability of the final results. The pilot study also served to test the questionnaire’s face validity. A group of 20 tourists (both Arab and international) completed the survey and provided feedback on clarity, scale appropriateness, and flow. Based on their input, minor revisions were made to improve wording consistency and to confirm that both the English and Arabic versions were accurate and comprehensible. This process played a crucial role in refining and consolidating the questionnaire, ensuring its comprehensibility, content validity, and addressing concerns regarding questionnaire length. The final survey yielded a total sample size of 165 respondents, with a response rate of approximately 50%, which is higher than the average for online surveys as identified by Wu et al. (2022). The sample reflects the demographic structure of tourists visiting Oman during the peak season but may overrepresent Asian respondents and professionals. While this composition reflects agency records of actual visitors at the time, we acknowledge that such concentration may limit the generalizability of the results to all international tourist groups.

5. Data Analysis and Results

First, the dataset was filtered to solve discriminant validity issues, which led to the exclusion of 44 observations using the standard deviation function in Microsoft Excel. After screening, 121 observations remained. A G*Power 3.1 analysis indicated that for a model with six predictors, at least 98 participants were needed to detect a medium effect size with a statistical power of 0.80 and a significance level of 0.05. The final sample therefore exceeded the minimum requirement, confirming adequacy for hypothesis testing. PLS-SEM is also well-suited for smaller to medium samples, as it does not rely on multivariate normality and can handle complex models with multiple constructs. While the relatively modest sample size may limit the generalizability of the results, it also reflects the realities of Oman’s developing tourism sector and its relatively limited tourist population due to the country’s arid climate and ongoing industry growth. Moreover, the descriptive statistics of the dataset were compared before and after filtration to ensure that the filtered dataset precisely represents the original data and does not significantly deviate from it.

Table 1. Respondents’ profile.

Variables	Descriptions	Frequency (%)
Age	Less than 30	5 (4%)
	30–49	65 (54%)
	50 and higher	51 (42%)
Gender	Female	68 (56%)
	Male	53 (43%)
Income	Less than $50,000	81 (66%)
	$50,000 and more	40 (33%)
	Not specified	0 (0%)
Marital status	Married	101 (83%)
	Single	15 (12%)
	Divorced	2 (2%)
	Separated	2 (2%)
	Widowed	1 (1%)
Children	Yes	74 (61%)
	No	47 (39%)
Education	Less than a bachelor’s degree	5 (4%)
	Bachelor’s degree or higher	116 (96%)
Occupation	Professional	103 (85%)
	Administrative/Managerial/Entrepreneur	5 (4%)
	Production/Agriculture worker	0 (0%)
	Govt. officer state enterprise	7 (6%)
	Housewife/Student/Retired/Unemployed/Other	6 (5%)
Residence	GCC countries	25 (21%)
	Other Arab countries	15 (12%)
	Asia	54 (45%)
	Europe	12 (10%)
	The Americas	2 (2%)
	Oceania	0 (0%)
	Africa	3 (2%)
	Others	10 (8%)

depicts the distribution of individuals by age, gender, income, marital status, presence of children, level of education, occupation, and place of residence. The table provides the number and percentage of participants after screening, enabling a thorough comprehension of the sample’s demographic composition.

To address the potential issue of common method bias in the collected data, the methodology proposed by Kock and Lynn (2012) and Kock (2015) was employed. The analysis involved testing full collinearity by regressing all variables against a random variable. According to Kock (2017), a VIF value of ≤5.0 indicates the absence of bias from the single source data. Applying this method, it was found that the VIF values were below the threshold, indicating that single-source bias is not a significant concern in the dataset.

Table 2. Full collinearity test.

Variable	A	AL	F	PV	RPME	S	NS	BL
VIF	1.12	3.567	2.125	2.158	1.056	3.51	2.978	2.785

Note: A = Attachment, AL = Attitudinal Loyalty, F = Familiarity, PV = Perceived Value, RPME = Perception of risk in the Middle East, S = Satisfaction, NS = Novelty Seeking, and BL = Behavioral Loyalty.

shows the VIF values for each construct in the model.

PLS-SEM was utilized in this study because it does not require a normality assumption as survey research is not normally distributed (Chin et al., 2003). The research employed the software SmartPLS 4 (Ringle et al., 2024) as the statistical tool to investigate the measurement and structural model. According to Henseler (2017), the best tool available for conducting PLS-SEM studies is SmartPLS.

Table 3. Outer loadings, internal consistency, and convergent validity.

Variables	Items	Loadings	CR	AVE
Attitudinal Loyalty	AL1	0.963	0.961	0.924
	AL2	0.960
Attachment	A1	0.871	0.922	0.798
	A2	0.882
	A3	0.925
Familiarity	F1	0.950	0.948	0.900
	F2	0.948
Novelty Seeking	NS1	0.780	0.912	0.722
	NS2	0.897
	NS3	0.848
	NS4	0.865
Perceived Value	PV2	0.953	0.954	0.912
	PV3	0.957
Behavioral Loyalty	BL1	0.923	0.922	0.855
	BL2	0.908
Satisfaction	S1	0.904	0.937	0.831
	S2	0.909
	S3	0.899

Note: F3 and PV1 were eliminated due to low loadings.

presents the factor loadings, composite reliability (CR), and average variance extracted (AVE) for each construct in the model.

The first step in assessing the measurement model is to evaluate indicator reliability using loadings, which should be ≥0.708 to ensure satisfactory item reliability (Hair et al., 2021; Ramayah et al., 2018). As shown in Table 3, all indicators meet this criterion with loadings ranging from 0.743 to 0.96. Next, internal consistency reliability was assessed using Jöreskog’s (1971) composite reliability (CR). In exploratory research, CR should be 0.6 or more. The PLS algorithm results indicate all constructs have CR values above 0.8 (Table 3). Convergent validity was then evaluated using the average variance extracted (AVE), which should be ≥0.5. The results show that all constructs meet this requirement, with the lowest AVE being 0.722 (Table 3).

After that, discriminant validity was assessed, using the heterotrait–monotrait ratio (HTMT) as recommended by (Henseler et al., 2015). The cut-off value suggests that the HTMT ratio for conceptually similar constructs should be less than 0.9 (Hair et al., 2021).

Table 4. Discriminant validity.

Variables	A	AL	BL	F	NS	PV	RP	S
A
AL	0.771
BL	0.869	0.774
F	0.872	0.652	0.834
NS	0.826	0.822	0.690	0.774
PV	0.814	0.601	0.865	0.743	0.716
RP	0.111	0.050	0.031	0.055	0.041	0.109
S	0.832	0.886	0.827	0.826	0.883	0.821	0.065

Note: A = Attachment, AL = Attitudinal Loyalty, F = Familiarity, PV = Perceived Value, RP = Perception of risk in the Middle East, S = Satisfaction, NS = Novelty Seeking, and BL = Behavioral Loyalty.

shows the HTMT ratios for the constructs of the model. The largest HTMT ratios obtained by the PLS algorithm is 0.886, which is less than the cut-off value, indicating that all the constructs are conceptually distinct, and the respondents understood the difference between them.

After ensuring that the constructs are reliable and valid, the data were tested for multivariate normality. As per Hair et al. (2022) and Cain et al. (2017), multivariate skewness and kurtosis were assessed. The results revealed non-multivariate normality of the collected data, indicated by Mardia’s multivariate skewness (β = 54.174, p < 0.01) and Mardia’s multivariate kurtosis (β = 172.214, p < 0.01). Consequently, following the recommendations of Becker et al. (2023), path coefficients, standard errors, t-values, and p-values for the structural model were reported using a 10,000-sample re-sample bootstrapping technique. As noted by Hair et al. (2022), using a larger number of resamples enhances the stability and reliability of the estimated sampling distribution and increases the consistency of results across runs. Additionally, in line with Hahn and Ang’s (2017) critique that p-values alone are insufficient for testing hypothesis significance, they suggested a combination of criteria such as p-values, confidence intervals, and effect sizes.

Table 5. Hypothesis Testing.

Hypothesis	Relationships	Std. Beta	Std. Errors	t-Value	p-Value	PCI LL	PCI UL	f2
H1	A → AL	0.274	0.106	2.584	0.005 ***	0.110	0.460	0.08
H2	A → BL	0.293	0.117	2.503	0.006 ***	0.118	0.504	0.08
H3	F → AL	−0.117	0.097	1.211	0.113	−0.283	0.032	0.02
H4	F → BL	0.211	0.118	1.797	0.036 **	0.001	0.389	0.05
H5	S → AL	0.547	0.076	2.159	0.015 **	0.046	0.293	0.03
H6	S → BL	0.190	0.113	1.673	0.047 **	0.001	0.373	0.03
H7	PV → AL	−0.155	0.081	1.920	0.027 **	−0.301	−0.033	0.04
H8	PV → BL	0.345	0.127	2.731	0.003 ***	0.141	0.554	0.14
H9	RP × F → AL	0.077	0.052	1.486	0.069 *	0.000	0.169	0.02
H10	RP × F → BL	−0.057	0.060	0.944	0.173	−0.153	0.044	0.01
H11	NS × F → AL	−0.123	0.039	3.174	0.001 ***	−0.183	−0.063	0.11
H12	NS × F → BL	−0.029	0.038	0.766	0.222	−0.087	0.034	0.01

Note: A = Attachment, AL = Attitudinal Loyalty, F = Familiarity, PV = Perceived Value, RP = Perception of risk in the Middle East, S = Satisfaction, NS = Novelty Seeking, and BL = Behavioral Loyalty. *** = 1% level of significance, ** = 5% level of significance, and * = 10% level of significance.

presents a summary of the criteria employed to evaluate the formulated hypotheses.

This study investigated the impact of six predictor variables on attitudinal and behavioral loyalty. The findings revealed R² of 0.747 for attitudinal loyalty and 0.696 for behavioral loyalty, indicating that the six predictors accounted for 74.7% of the variability in attitudinal loyalty and 69.6% of the variability in behavioral loyalty. According to Hair et al. (2021), R² values of 0.75, 0.50, or 0.25 for endogenous latent variables can be described as substantial, moderate, or weak in scholarly research, respectively. Based on this criterion, both endogenous variables have substantial explanatory power.

In the prediction of attitudinal loyalty, Attachment (β = 0.274, p < 0.01) and Satisfaction (β = 0.547, p < 0.05) exhibited positive effects, supporting hypotheses H1 and H5. However, perceived value (β = −0.155, p < 0.05) is found to have a negative impact on attitudinal loyalty while familiarity (β = −0.117, p > 0.05) was not significant; thus, H3 and H7 were not supported. In relation to behavioral loyalty, Attachment (β = 0.293, p < 0.01), Familiarity (β = 0.211, p < 0.05), Satisfaction (β = 0.190, p < 0.05) and Perceived value (β = 0.345, p < 0.01) exhibited positive effects, supporting hypotheses H2, H4, H6 and H8.

Next, we tested the moderating effects of Perceived Risk and Novelty Seeking on the relationship posited. Familiarity × NS → Attitudinal Loyalty (β = −0.123, p < 0.01) was significant, indicating it negatively moderated the relationship, while Familiarity × RP → Attitudinal Loyalty (β = 0.077, p < 0.1) was significant but it positively moderated the relationship. Thus, H11 was supported while H9, H10 and H12 were not supported. Figure 2 depicts the inner model and indicator loading for the outer model as extracted from SmartPLS software:

As suggested by Hair et al. (2017), to improve the understanding of the moderation impact, a simple slope analysis was performed. Figure 3 illustrates the slopes of the moderator construct (novelty seeking) at its high and low levels. A low level of moderator construct novelty seeking is represented by a flattened slope, whereas a high level of moderator construct novelty seeking is represented by a steeper slope. This indicates that novelty seeking negatively moderated the relationship between familiarity and attitudinal loyalty, weakening the relationship when novelty seeking is elevated.

Similarly, simple slope analysis was drawn for the moderating impact risk perception about the Middle East region on the relationship between familiarity and attitudinal loyalty as shown in Figure 4.

A high level of moderator construct risk perception about the Middle East is represented by a steeper slope, whereas a low level of moderator construct risk perception about the Middle East is represented by a flatter slope. This indicates that risk perception about the Middle East positively moderates the relationship between familiarity and attitudinal loyalty. Although our hypothesis expected a negative moderation, this positive effect can be explained post hoc: in a region perceived as risky, Oman’s reputation for safety makes familiarity act as a “safety heuristic,” thereby strengthening attitudinal loyalty. In SCT terms, salient environmental cues (regional risk) interact with personal cognitions (familiarity) to increase tourists’ willingness to recommend. Given the marginal significance (p = 0.069), this result should be viewed as tentative and as a possible boundary condition for loyalty theory.

In addition to that, Shmueli et al. (2019) proposed PLS-predict, a holdout sample-based procedure that generates case-level predictions on an item or construct level using the PLS-Predict with a 5-fold procedure to assess predictive relevance. Based on

Table 6. PLS predict.

MV	PLS-SEM_RMSE	LM_RMSE	PLS-LM	Q2 Predict
AL1	0.582	0.611	−0.029	0.649
AL2	0.557	0.624	−0.067	0.641
BL1	0.802	0.842	−0.040	0.538
BL2	0.777	0.772	0.005	0.522

, all the errors of the PLS model for Attitudinal Loyalty were lower than those of the LM model, indicating strong predictive power, while for Behavioral Loyalty a majority of the errors were lower in the PLS model than the LM model, indicating medium predictive power.

6. Discussion and Conclusions

The literature demonstrates that the concept of loyalty within the context of tourism can be categorized into two distinct forms: attitudinal and behavioral loyalty. Attitudinal loyalty is characterized by positive feelings towards a destination, leading to a willingness to recommend it. Behavioral loyalty, on the other hand, refers to the intention to revisit the destination (Fajriyati et al., 2022).

Building on this understanding of loyalty’s different forms and the factors influencing it, the link between tourist loyalty to a destination and the demand for the destination becomes more apparent. A destination with a high degree of tourist loyalty is likely to experience consistent or increased demand as loyal tourists return (behavioral loyalty) and recommend it (attitudinal loyalty) to others. In fact, word-of-mouth recommendations may be more influential in selecting tourism destinations than commercial advertising (Arina et al., 2025; Witt & Witt, 1995). Hence, understanding factors influencing tourists’ loyalty can help destination marketers and managers enhance the appeal and demand for their destinations (Nusair et al., 2023). Candela and Figini (2012) suggested the duration of stays at a destination is influenced by multiple variables, including loyalty.

Predicting potential tourists’ likelihood to visit a destination is crucial for both tourism managers and marketers. It can guide infrastructure development, demand-based pricing, and positioning strategies (Oppermann, 1999). Additionally, tourism marketers can allocate resources more efficiently by targeting potential visitors with higher probabilities of visiting and avoiding unnecessary marketing expenditures on those unlikely to visit or who have already decided to visit (Oppermann, 1999).

The study reveals a set of five predictor variables that significantly impact attitudinal loyalty and a set of four predictor variables that significantly impact behavioral loyalty. Determinants of attitudinal loyalty include attachment, satisfaction, perceived value, the moderating impact of risk perception on the relationship between familiarity and attitudinal loyalty (at the 10% level of significance), and the moderating impact of novelty seeking on the relationship between familiarity and attitudinal loyalty. On the other hand, determinants of behavioral loyalty include attachment, familiarity, perceived value, and satisfaction.

Beyond the statistical results, these findings carry important theoretical implications. The negative effect of perceived value on attitudinal loyalty suggests that while tourists may view Oman as offering value-for-money, this perception alone does not build emotional attachment or advocacy, highlighting the conceptual distinction between behavioral and attitudinal loyalty. This nuance refines loyalty theory by showing that repeat visits can be driven by economic rationality, whereas attitudinal loyalty requires deeper affective bonds such as attachment and satisfaction. In addition, the significant moderating role of novelty seeking reinforces Social Cognitive Theory by demonstrating how cognitive evaluations can weaken the familiarity–loyalty pathway, as novelty-oriented tourists prioritize variety over routine, even when familiarity reduces uncertainty. Conversely, the weak influence of risk perception suggests that Oman’s reputation for safety may buffer the impact of broader regional instability, thereby challenging prior studies that emphasized risk as a major deterrent. Collectively, these results extend existing loyalty frameworks by illustrating the differentiated roles of value, novelty, and risk in shaping tourists’ attitudinal versus behavioral commitments.

The positive effects of attachment and satisfaction align with the literature (Bojanic, 1996; Khoi et al., 2022; Kim & Park, 2017; Mechinda et al., 2009; Vera et al., 2025). However, in contrast to the findings of Mechinda et al. (2009), this study revealed that familiarity has no impact on attitudinal loyalty. Potential explanations for this variation in results might be the differences in sample size and the technique employed in data analysis between the two studies. In studying complex relationships, second-generation techniques like PLS-SEM are favored over first-generation techniques like Multiple Regression Analysis (Hair et al., 2021), which was employed by Mechinda et al. (2009). However, the finding of this study regarding the impact of familiarity on behavioral loyalty is consistent with Mechinda et al. (2009). The level of uncertainty associated with revisiting a country for a holiday is significantly lower compared to traveling to a foreign country that has not been previously visited (Witt & Witt, 1995).

Interestingly, perceived value was found to have a negative impact on attitudinal loyalty. On the other hand, perceived value has a positive impact on behavioral loyalty. This can be explained by both tourism-specific theory and broader consumer research. As McKercher et al. (2012) argue, tourism loyalty is multidimensional and may manifest vertically (toward service providers), horizontally (across competing providers), or experientially (toward holiday styles), rather than being tied exclusively to a single destination. In this sense, tourists who perceive high value may be motivated to revisit (behavioral loyalty) but may not develop stronger attitudinal bonds with a destination, since their sense of value can attach to broader experiences or service providers, encouraging substitution across destinations with similar attributes. This interpretation is reinforced by recent evidence from value-based adoption research outside tourism. Perceived value, shaped by assessments of benefits and risks, tends to encourage repeat behavior but does not necessarily translate into affective loyalty unless mediated by relational factors such as trust or attachment. Together, these perspectives suggest that perceived value in tourism may remain transactional, promoting behavioral intentions but not the deeper emotional attachment required for attitudinal loyalty. Furthermore, these findings suggest that tourists may have greater sensitivity toward prices or elevated expectations regarding value for money.

This also confirms the complex nature of the relationship between attitudinal loyalty and its antecedents, suggesting that the relationship between perceived value and attitudinal loyalty is not linear and may require a moderator or mediator variable to explain the mechanism of impact. Despite the recognized impact of perceived value on loyalty, research suggests that other factors, such as a firm’s actions and strategic focus on market orientation, can influence this relationship (Chen & Tsai, 2008; Z. Yang & Peterson, 2004).

The negative moderating impact of novelty seeking on the relationship between familiarity and attitudinal loyalty can be explained by the fact that novelty seekers are continuously searching for new experiences, which often leads them to avoid revisiting areas they have already explored, viewing a return visit as redundant (Al Mahruqi, 2023). While tourists may have positive opinions about a destination, their inherent drive for novelty may hinder them from returning, negatively moderating the impact of familiarity on behavioral loyalty.

Moreover, the moderating effect of risk perception in the Middle East on the relationship between familiarity and attitudinal loyalty was marginally significant at the 10% level (p = 0.069). This suggests that individuals’ levels of risk perception in the Middle East may positively moderate the effect of familiarity on the intention to recommend Oman, thus potentially strengthening the relationship between familiarity and attitudinal loyalty. The Middle East is often characterized as a high-risk region, attributed to its cultural differences, relatively low levels of economic development, and persistent political instability and conflict (Lepp & Gibson, 2008). However, Oman’s positive reputation as a safe country offers an alternative choice for tourists seeking to visit Arab countries, thereby supporting its plan to diversify the economy beyond the hydrocarbon sector (Al Abri et al., 2023a). This resonates with recent post-COVID evidence showing that tourists increasingly anchor destination loyalty not only on satisfaction but also on strengthened evaluations of safety and resilience (Eid et al., 2019). Given the marginal nature of this result, it should be interpreted as an exploratory indication rather than a robust empirical confirmation, highlighting the need for further research.

Considering the study’s findings, several policy implications emerge. Firstly, differentiating between attitudinal and behavioral loyalty can help Omani tourism policymakers design more effective marketing strategies. Given the significant impact of satisfaction on attitudinal loyalty and behavioral loyalty, policymakers might train local youth to use online platforms offering unconventional experiences, thereby enhancing visitor satisfaction (Gundersen et al., 2024). For example, platforms like Airbnb experiences provide a “Live like a local” experience (Purohit et al., 2023). Engaging youth and local communities in tourism value chains can drive inclusive and sustainable growth (Li et al., 2023) by creating jobs, preserving culture, and ensuring that tourism benefits are widely shared, in line with Vision 2040. Moreover, policymakers in Oman might consider increasing the variety of activities offered in natural attractions, as findings and literature show that satisfaction is influenced by variety and drives loyalty. The negative impact of perceived value on attitudinal loyalty suggests that value-for-money experiences might encourage repeat visits, even without strong emotional ties. Therefore, value-driven promotions or diverse activities in natural attractions could improve perceived value (Candela & Figini, 2012). Furthermore, given the importance of perceived risk in the Middle East, Oman’s image as a safe destination should be leveraged to attract more tourists and investments. To enhance loyalty, policymakers must maintain security and emphasize Oman’s secure and attractive position in the region. Moreover, promoting Oman as a safe and eco-friendly destination for Arabic language learners could make educational tourism a valuable service export for Oman (Hussein et al., 2021). In addition, given the significant impact of satisfaction and attachment on attitudinal loyalty, Omani tourism service providers should utilize word-of-mouth (WOM) marketing techniques such as destination ambassadors and influencer campaigns.

While the study offers valuable insights into factors affecting tourist loyalty in Oman, there are limitations. One limitation of this study is the modest sample size, which may restrict generalizability. Nonetheless, G*Power analysis confirmed that it exceeded the minimum requirement for detecting medium effect sizes with adequate power, and PLS-SEM is appropriate for smaller samples. The sample size also reflects Oman’s tourism context, where visitor numbers remain limited due to the arid climate and the sector’s ongoing development. In addition, potential sampling bias due to the relatively high proportion of Asian and professional respondents should be noted. While this does not undermine the internal validity of the estimated relationships, broader samples in future studies would strengthen external validity and generalizability. Moreover, the reliance on self-reported data may lead to response bias, and the cross-sectional design does not account for changes in preferences over time. Future research could employ a longitudinal design to track changes in loyalty factors, particularly in the evolving tourism landscape. Furthermore, seasonality and timing could influence tourists’ experiences and responses. Therefore, future studies might include the timing of visits to capture potential seasonal variations more explicitly. Additionally, exploring other potential moderating variables beyond novelty-seeking and risk perception could deepen the understanding of loyalty determinants. Further research into the relationship between perceived value and loyalty is also needed. Lastly, as sustainable tourism grows in importance, examining how environmental consciousness influences loyalty could be a promising avenue for future research.