The Impact of Consumer’s Adaptation to the Creative Culture of Theme Parks on Review Usefulness

Abstract

In the era of information overload and repetitive reviews, there has been limited exploration into the influence of consumers’ cultural adaptation to creative symbols in theme parks on the usefulness of online reviews, which is significant for enhancing tourism experiences, targeted marketing, personalized services, and informed tourism choices. This study aims to bridge this gap by examining how cultural adaptation factors interact and impact the review usefulness, and by considering the role of cultural adaptation in simplifying information during consumer decision-making processes. Additionally, the study investigates how consumers’ decision reference points, represented by advanced ticket levels, moderate their attention to attribute consistency when evaluating the review usefulness. A sample of 5929 valid consumer reviews of Disney theme parks from 2019 to 2022 on Meituan.com is analyzed using latent semantic analysis and Tobit regression to test the proposed hypotheses. We find that high symbolic creativity reviews stimulate cultural adaptation and increase attention to service attributes when evaluating review usefulness. Moreover, advanced ticket levels do not moderate the usefulness of extremely negative reviews. However, they do have a moderating effect on the usefulness of extremely positive reviews, with the direction of moderation differing based on the levels of symbolic creativity.

1. Introduction

In the rapidly evolving world of e-commerce, the influence of online reviews on consumer decisions and business success in the tourism industry cannot be underestimated [1]. Extensive research has emphasized the direct impact of reviews on theme parks, including visitor numbers, tourism spending, and market competitiveness [2,3]. However, as the volume of online reviews continues to grow exponentially, it has resulted in an overload of information [4] and redundant content [5], which presents challenges for consumers. Consequently, it becomes crucial to understand the usefulness of reviews in order to navigate through this overwhelming abundance of information. Useful reviews not only sway purchase decisions, but also assist tourism managers in enhancing their services, ultimately leading to increased transactions. However, existing research has primarily focused on identifying important attributes of tourism services [6,7,8], while neglecting to fully consider the influence of consumers’ understanding and adaptation to the destination culture, which plays a significant role in enriching travel experiences, fostering emotional resonance, promoting cultural exchange, and establishing meaningful interpersonal connections [9,10]. Therefore, it is necessary to explore the impact of these factors on consumers’ decisions regarding the usefulness of reviews. Furthermore, current research fails to acknowledge consumers’ tendency to simplify information when faced with information overload, which hampers the effective utilization of useful reviews for product and service improvements.

Theme parks possess distinctive features that set them apart from other tourism sectors. Serving as pioneers in the experience economy [11], theme parks create immersive environments that allow visitors to engage with specific cultures or storylines. Symbolic creativity plays a vital role in theme parks, encompassing the creation and expression of symbolic means during cultural dissemination. It not only provides diverse theme elements, but also facilitates visitors’ understanding and acceptance of the cultural significance within theme parks [12]. Symbolic creativity is the key to theme park success and is regarded as a prime example of service innovation.

While previous studies have shown the positive effects of tourist cultural adaptation on cross-cultural communication, understanding, and tourist satisfaction [13,14,15], these studies have primarily focused on immigrants rather than tourists. Additionally, there has been limited research on the topic of theme park cultural adaptation, which is crucial in providing a unique visitor experience. Furthermore, the impact of consumers’ cultural adaptation to creative symbols in theme parks on the review usefulness remains unexplored, particularly in the context of information overload and repetitive reviews. Therefore, it is important to investigate the influence of consumer adaptation to creative theme park culture on the evaluation of review usefulness. Such research is significant for enhancing tourism experiences, targeted marketing, and cross-cultural exchange. It also opens up opportunities for personalized services and informed tourism choices, as well as providing fresh insights into the dynamics of cultural adaptation and consumer behavior [9,10]. Hence, the objective of this study is to investigate the impact of factors influencing consumer cultural adaptation in online reviews and the interactive effects of these factors on the usefulness of theme park online reviews. Additionally, this study will explore how these correlations can be integrated into service improvement and marketing strategies, with the goal of meeting consumer needs and expectations more effectively and efficiently.

Previous research has identified several factors that influence consumers’ perceptions and adaptation to service culture based on reviews, including clarity and specificity [16], attribute consistency [17], and the framing and tone of attribute descriptions [18]. Consistency in attribute descriptions across multiple reviews plays a crucial role in shaping consumers’ understanding of service culture. When descriptions are consistent, consumers perceive the service culture as reliable. However, the impact of attribute review consistency on usefulness evaluations concerning the clarity and specificity of theme park culture descriptions remains unexplored. By analyzing the influence of attribute consistency in highly symbolic and low symbolic creativity reviews, this study aims to shed light on the relationship between review consistency and usefulness evaluations.

Extreme reviews, which are rated as the highest or lowest by consumers, provide valuable insights into the way attributes are described and the overall tone of the reviews. These reviews have a significant impact on consumers’ understanding and adaptation to the theme park culture, making them highly regarded and considered more useful [19,20,21]. However, previous research mainly focused on the influence of ratings or the semantic features of extreme reviews on their usefulness, without thoroughly examining the consistency in attribute descriptions within these reviews, especially in scenarios where there is an overload of information [19,20,21,22,23]. Therefore, this study aims to fill this gap by investigating the role of attribute description consistency in determining the perceived usefulness of extreme reviews. By doing so, it will provide a more comprehensive understanding of the factors that contribute to consumers’ evaluation of extreme reviews and their impact on consumer behavior in the theme park context.

According to reference point theory, consumers’ assessment of review usefulness relies on their ability to adapt and accept symbolic creativity in reviews [24]. However, previous research in consumer decision making has predominantly treated decision reference points as unchanging entities [25,26,27], failing to acknowledge the dynamic nature of reference points across various decision makers and decision-making scenarios [28]. Additionally, the evaluation and decision-making process of theme park consumption heavily relies on the different levels of advanced tickets. These varying ticket levels give rise to distinct expectations and decision-making reference points [8]. Hence, this study aims to examine how different levels of advanced tickets moderate consumers’ attention to attribute description consistency in their evaluations of usefulness, particularly in diverse cultural adaptation scenarios.

The study findings reveal that reviews with high symbolic creativity have a significant impact on cultural adaptation among consumers, leading them to prioritize service attributes when evaluating the review usefulness. Unlike negative reviews, which do not trigger any adjustment in reference points based on advanced ticket levels, extremely positive reviews do elicit reference point adjustment, particularly influenced by advanced ticket prices. The direction of this adjustment varies depending on the level of symbolic creativity. In high symbolic creativity extremely positive reviews, higher advanced ticket levels have a positive moderating effect. Conversely, in low symbolic creativity extremely positive reviews, higher advanced ticket prices have a negative moderating effect. These findings suggest that higher advanced ticket levels amplify cognitive dissonance in high symbolic creativity extremely positive reviews, resulting in increased attention to theme park service attributes. On the other hand, in low symbolic creativity extremely positive reviews, higher advanced ticket levels confuse consumers and impede their ability to make useful evaluation decisions. Previous research has primarily focused on the impact of brand symbolic creativity on consumer satisfaction and loyalty, neglecting the combined effects of symbolic creativity and cultural adaptation [29,30,31]. By exploring the influence of consumers’ adaptation to theme park symbol creative culture on the usefulness of online reviews, this study provides valuable insights into the mechanisms underlying the impact of theme park symbol creative culture on consumer loyalty.

From a management perspective, key attributes should be prioritized when enhancing the theme park experience [32]. Increasing symbolic creativity in theme parks can attract consumer attention to creative-related attributes and foster loyalty [33]. The active cultivation of word-of-mouth promotion and promotion of the theme park’s symbolic creativity culture are recommended. Moreover, raising advanced ticket levels can increase consumer attention to theme park attributes and enhance loyalty [34]. For new users unfamiliar with the theme park’s symbolic creativity culture, cultivating word-of-mouth promotion and offering low advanced ticket levels for basic admission can enhance consumer attention to theme park attributes and loyalty.

The rest of this study is organized as follows. The research status regarding review usefulness and tourist cultural adaptation is presented in Section 2. The research models and hypotheses are described in Section 3. The research design process is shown in Section 4. The data analysis and the results of the experiments are provided in Section 5. Finally, Section 6 concludes the study.

2. Research Status

2.1. Review Usefulness

The usefulness of online reviews plays a crucial role in providing consumers with decision-making information [35]. By reflecting consumers’ perceived value of product or service quality, review usefulness helps consumers make informed purchase decisions. However, the exponential growth of online reviews has given rise to challenges such as varying review quality and the prevalence of fake reviews, resulting in information overload. These low-quality reviews undermine consumer trust in platforms and adversely affect transaction volumes. Consequently, it becomes vital to efficiently identify high-quality reviews amidst the vast number available reviews and locate those that are genuinely useful to consumers.

In existing research, the factors influencing online review usefulness can be categorized into two main groups: review characteristics and reviewer characteristics. Review characteristics studied by scholars encompass aspects such as readability, text length, depth, label-content relevance, and rating. For instance, Korfiatis et al. (2012) found that review readability has a greater impact on review usefulness than its length [36]. Kuan et al. (2015) investigated the effects of various review characteristics, including length, readability, valence, extremity, and reviewer credibility, on review voting and usefulness [37]. Mudambi and Schuff (2010) demonstrated that review usefulness is influenced by factors like review depth and product type, with extreme rating reviews being less useful for experience goods [38]. Zhang et al. (2022) confirmed the positive effects of label-content relevance and content length on review usefulness [39]. Regarding reviewers, scholars have examined demographic characteristics such as gender and nationality, as well as the number of followers, profile image, reputation, and grade. For instance, Cheng and Ho (2015) revealed that reviewers with a larger number of followers and higher expertise levels are perceived as providing more practical and useful reviews [40]. Filieri, Raguseo, et al. (2018) found that extreme reviews accompanied by photos and longer texts are more helpful to consumers [41]. Karimi and Wang (2017) explored the impact of reviewer profile images on review helpfulness and observed a significant enhancement in consumers’ evaluation of review usefulness [42]. However, there is a research gap regarding the exploration of how different degrees of symbolic creativity impact review usefulness.

Scholars have also investigated the moderating factors that influence consumers’ evaluation of review usefulness, such as product price and review ratings. For example, Xu et al. (2023) explored the moderating effect of product price on the relationship between emotions (anger and anxiety) and review usefulness [43]. Zhou and Guo (2017) examined the relationship between review order and review usefulness, along with its moderating effects [44]. Zhu et al. (2014) developed an integrative model of online review usefulness considering the moderating effects of service price and rating extremity [45]. However, few studies have considered the moderating effect of advanced ticket levels in theme parks on consumers’ evaluation of review usefulness.

Additionally, research has explored the impact of extreme ratings on review usefulness. Extreme reviews, characterized by the highest or lowest rating scale, have often been deemed more useful. For example, Filieri, Hofacker, et al. (2018) demonstrated that extreme reviews have a greater impact on large hotels than small ones, especially when accompanied by photos and longer texts [21]. Filieri, Raguseo, et al. (2021) highlighted that extremely negative reviews are perceived as more useful when the hotel has received a certificate of excellence and possesses higher average rating scores and classification [19]. Park and Nicolau (2015) showed that extreme ratings are perceived as more useful and enjoyable compared to moderate ratings [20]. However, inconsistent findings exist, such as the influence of factors like product type and rating dispersion. For instance, Cao et al. (2011) indicated that semantic characteristics have a greater influence on the number of usefulness votes received by extreme rating reviews [22]. Lee et al. (2021) found that in situations of low trust in average ratings (high rating dispersion), extreme ratings have a stronger impact on review usefulness due to the absence of ambiguity in extreme opinions [23].

However, previous studies have primarily focused on the impact of ratings or semantic features of extreme reviews on review usefulness, neglecting the examination of attribute description consistency in extreme online reviews [46,47]. Moreover, in an environment characterized by information overload and repetitive reviews, consumers exhibit varying degrees of cognition and adaptation to theme park culture [13,48]. Therefore, the purpose of this study is to examine the influence of attribute description consistency on the usefulness of extreme reviews, both high symbolic creativity and low symbolic creativity in nature, in comparison to neighboring reviews. Additionally, it aims to investigate how different advanced ticket levels moderate consumers’ attention to attribute description consistency in reviews, particularly when evaluating their usefulness in different cultural adaptation scenarios involving creative symbols.

By addressing these research gaps, this study contributes to the understanding of factors influencing review usefulness and extends knowledge on the impact of cultural adaptation in the context of theme parks. The findings will provide valuable insights for practitioners and researchers to enhance the quality and effectiveness of online reviews, improve consumer decision-making processes, and refine marketing strategies within the theme park industry.

2.2. Tourist Cultural Adaptation

Cultural adaptation refers to the ability of individuals or groups to adjust their behavior and thinking patterns in different cultural contexts, demonstrating their adaptability to external environmental changes. This process involves learning and adapting to new social norms, values, and behaviors, enabling individuals to better integrate into new cultural environments. Numerous studies have explored the phenomenon of cultural adaptation, including its measurement and strategies.

In terms of measuring cultural adaptation, Demes and Geeraert (2014) introduced four acculturation scales, including sociocultural adaptation, psychological adaptation, perceived cultural distance, and acculturation orientation [49]. Their findings revealed positive correlations between sociocultural and psychological adaptation, while demonstrating a negative association with perceived cultural distance. Ward and Kennedy (1999) developed the Sociocultural Adaptation Scale (SCAS), based on extensive data compiled from various sojourner samples, to investigate the relationship between sociocultural and psychological adjustment across different groups [50]. Regarding adaptation strategies, Blynova et al. (2020) conducted empirical research on personality factors influencing the choice of adaptation strategies in different cultural environments among labor migrants [51]. They identified three major strategies: integration, assimilation, and marginalization, with integration and assimilation acting as means to resolve the crisis of social identity. Glukhova (2021) conducted an interdisciplinary study on brand semantics and the cultural adaptation strategies applied when transferring brands to international or local markets [52]. This approach provides a cultural perspective on brand translation studies, facilitating the application of context-based or value-based cultural adaptation strategies.

Within the realm of tourism, cultural adaptation plays a vital role in shaping tourists’ experiences. Tourist cultural adaptation involves the behaviors and processes tourists engage in during their travels to gradually adapt to social norms, values, and behaviors in different cultural backgrounds [53]. This adaptation process facilitates cross-cultural communication, reduces cultural differences and misunderstandings, and enhances the overall tourism experience [54]. It also broadens tourists’ horizons, fosters understanding and respect for different cultures, and improves cultural literacy. Several studies have explored the impact of cultural adaptation on tourist satisfaction. For instance, Antón et al. (2019) examined the influence of the perceived authenticity of local food, adaptation degree, and cultural contrast on memorable tourist experiences [13]. Chen and Rahman (2018) investigated the interplay between visitor engagement, cultural contact, memorable tourism experience (MTE), and destination loyalty in cultural tourism [14]. Their findings indicated positive relationships between visitor engagement, cultural contact, MTE, and loyalty. Zhang et al. (2019) highlighted the moderating role of Confucian culture in the relationship between perceived advantage, satisfaction, user stickiness, and word-of-mouth communication [55].

However, the previous research primarily focuses on the impact of cultural adaptation effects on consumer satisfaction or loyalty, overlooking the exploration of its influence on the usefulness of online reviews [12,56]. Additionally, in an environment characterized by information overload, the effect of cultural adaptation in online reviews on simplifying information during consumer decision-making processes remains largely unexplored [57,58]. Therefore, the objective of this study is to assess the level of cultural adaptation among consumers in theme park culture, specifically in relation to the presence of creative symbols in online reviews, and analyze how this adaptation impacts consumer evaluations of the usefulness of such reviews.

By filling these research gaps, this study contributes to the understanding of how cultural adaptation influences review usefulness and sheds light on the mechanisms through which customers simplify information in an information overload context. The findings will provide valuable insights for practitioners and researchers in enhancing the effectiveness of online reviews, improving consumer decision-making, and developing targeted marketing strategies within the theme park industry.

3. Research Models and Hypotheses

Applying cultural adaptation theory combined with reference point theory, this study explores the impact of the consistency in service attribute descriptions of theme parks on the extreme review usefulness under different levels of symbolic creativity. Also, it considers the moderating effect of the advanced ticket level in the relationship between the consistency in attribute descriptions of reviews and review usefulness when advanced ticket levels are regarded as different reference points. The research framework of this study is shown in Figure 1.

3.1. Influence of Service Attributes of Theme Parks

The attribute-based model, based on cognitive decision-making methods [59], suggests that consumers employ compensatory strategies to evaluate product or service attributes and form expectations regarding their quality. Consumers establish criteria and dimensions for evaluating alternatives and weigh the pros and cons of each alternative based on these criteria. This model assumes that consumers consider all available service attributes when making purchase decisions. The cultural adaptation effect refers to the increasing accuracy in an individual’s understanding and interpretation of cultural values, beliefs, social norms, behavioral standards, and subtle differences as their understanding of cultural significance deepens. This understanding allows individuals to better identify and comprehend implicit information and contextual cues within the culture [60].

Research has shown that high symbolic creativity reviews positively influence consumers’ adaptation to theme park culture [12]. Theme parks often emphasize creativity and imagination, and reviews with high symbolic creativity can stimulate consumers’ imagination, facilitating their immersion in the fantasy world created by the theme park and enhancing their cultural adaptation. As a result, consumers gain a deeper understanding of the information conveyed by the theme park, enabling them to evaluate it more efficiently [61]. In an information overload environment, consumers typically employ simplified decision-making strategies to make quick decisions and overcome the challenges of information overload within limited time constraints [57,58]. Therefore, we hypothesize that, in high symbolic creativity reviews, consumers may not consider all service attributes when evaluating the usefulness of reviews. On the other hand, low symbolic creativity reviews can have a negative impact on consumers’ adaptation to theme park culture [12]. In such reviews, consumers may struggle to comprehend implicit meanings due to the simplicity and commonness of the vocabulary and expressions used, which often lack advanced and abstract language. Consequently, consumers may require additional explanations or clarifications to adapt to the theme park culture. They are more likely to focus on specific service attributes, such as price and performance, while overlooking other attributes. Based on these observations, we propose the following hypothesis:

Hypothesis 1 (H1).

Not all consistency in attribute descriptions plays an important role in the review usefulness, regardless of whether they are in low symbolic creativity reviews or high symbolic creativity reviews.

High symbolic creativity reviews contribute to a more profound, vivid, and emotionally engaging experience for consumers, facilitating their adaptation to the theme park culture [12]. As cultural adaptation increases tourists’ interest in learning about destination-related information [62], high symbolic creativity reviews often convey deeper meanings, prompting consumers to consider a greater number of attributes when evaluating the usefulness of reviews. Conversely, cultural maladaptation can reduce interest in comprehensively understanding destination-related information [62]. Low symbolic creativity reviews typically only focus on basic service attributes, failing to effectively convey the distinctive features and essence of theme park culture. Consequently, consumers find it challenging to comprehend and adapt to the theme park culture, resulting in their tendency to only consider the fundamental service attributes when evaluating the usefulness of reviews, overlooking other aspects and considering fewer attributes. This observation aligns with the perspective of choice processing theory [63]. Based on these considerations, we propose the following hypothesis:

Hypothesis 2 (H2).

In terms of consistency in attribute description, there are a greater number of attributes that significantly influence the review usefulness in high symbolic creativity reviews than in low symbolic creativity reviews.

3.2. Moderating Effect of Advanced Ticket Level

The advanced ticket level refers to a VIP ticket type that offers exclusive service categories and differentiated levels of service. These categories include designated facilities and special services such as early park admission, fast-track access to rides, premium seating for performances, and exclusive tour guides. Classification represents a qualitative assessment of the facility or service quality provided by a supplier [64], leading to varying expectations regarding service quality levels at a destination. When consumers make high-cost purchases (e.g., five-star hotels, membership services, VIP tickets), they have higher expectations for the quality of the products or services they acquire. While research on the impact of ticket types on consumer satisfaction in the tourism field is limited, studies in related travel service industries, such as hotels, have shown that price plays a moderating role in consumer evaluations [8]. Specifically, different price levels associated with differently starred hotels moderate the relationship between service attributes in extreme reviews and review usefulness. Building upon this, our study quantifies theme park ticket types and explores the moderating role of advanced ticket levels in the relationship between the consistency in attribute descriptions in reviews and their usefulness, drawing on reference point theory.

Reference point theory, initially proposed by psychologists Daniel Kahneman and Amos Tversky, stems from the concept of “bounded rationality” in prospect theory, and has been found applications in various fields including economics, marketing, and decision science [24]. According to this theory, people use a reference point as a standard to evaluate and make decisions about different options. Evaluations of outcomes are not solely based on their absolute value, but also on deviations from the reference point [24]. Specifically, individuals tend to derive greater satisfaction from gains that surpass the reference point and greater dissatisfaction from losses that fall short of it.

In our study, the reference point represents the perceived service level and content richness that consumers expect to experience based on past travel experiences or relevant information. Unlike the traditional reference point theory that assumes a fixed reference point during decision making, we propose that the reference point for evaluating review usefulness is dynamically changing [28]. Consequently, we argue that the impact of consistency in service attribute descriptions of theme parks on review usefulness is moderated by advanced ticket levels.

Ticket levels have a significant influence on consumers’ evaluations and expectations in theme parks. Generally, higher-priced tickets generate higher expectations and greater emotional investment from consumers [65], as they anticipate better services and experiences at these theme parks. This higher reference point for decision making regarding review usefulness intensifies consumers’ cognitive dissonance when encountering differing reviews, prompting them to evaluate more thoughtfully. The use of high symbolic creativity in reviews enhances consumers’ adaptability to the theme park culture, and cultural adaptation to tourist destinations fosters a positive impression of the destination [56]. Moreover, extremely positive reviews can lead consumers to align their evaluations with the theme park, surpassing their own decision-making reference points and making more useful evaluations. Conversely, low symbolic creativity reviews may fail to vividly and interestingly convey the atmosphere and cultural characteristics of the theme park, hindering consumers from fully understanding the park’s culture. The cultural gap may dissuade consumers from investing more time and effort in comprehending the theme park’s culture [66]. Additionally, extremely negative reviews may result in an overly negative impression of the theme park, leading consumers to underestimate its actual value, falling far below their expected decision-making reference point. Consequently, consumers feel disappointed and provide more useful evaluations on reviews. Based on these considerations, we propose the following hypotheses:

Hypothesis 3 (H3).

Advanced ticket levels positively moderate the impact of the consistency in attribute descriptions in low symbolic creativity extremely negative reviews on review usefulness.

Hypothesis 4 (H4).

Advanced ticket levels positively moderate the impact of the consistency in attribute descriptions in high symbolic creativity extremely positive reviews on review usefulness.

In high symbolic creativity reviews, consumers tend to develop a better understanding of the culture and concepts associated with a theme park. This heightened understanding increases their interest in and investment towards learning about the tourist destination [56], subsequently leading to more positive perceptions of the park’s services and experiences [67]. However, this can also magnify the disparities among reviews and create cognitive dissonance for consumers, causing them to carefully contemplate their decisions. Conversely, extremely negative reviews can create an excessively negative impression of the theme park, which can hinder cultural adaptation. While high-priced tickets have the potential to raise consumer expectations and emotional investment [65], thereby elevating the decision-making reference point, conflicting reference information can complicate decision making when expectations are high [57]. Consequently, consumers may struggle to form a definitive opinion regarding the theme park’s services and experiences, resulting in a lack of useful evaluations.

On the other hand, low symbolic creativity reviews fail to effectively convey the ambiance and cultural characteristics of the theme park, preventing consumers from developing a deep understanding of its culture and creating a larger cultural gap. This can lead to increased time and effort spent on learning and adapting to the local culture [66], ultimately influencing negative perceptions of the park’s services and experiences [68]. Similarly, extremely positive reviews can create an overly positive impression of the theme park, leading to cultural ambiguity and confronting high decision reference points associated with high-priced tickets. As a result, consumers may find it challenging to form a definitive opinion regarding the theme park’s services and experiences [57], again resulting in a lack of useful evaluations.

Based on these considerations, we propose the following hypotheses:

Hypothesis 5 (H5).

Advanced ticket levels negatively moderate the impact of the consistency in attribute descriptions in high symbolic creativity extremely negative reviews on review usefulness.

Hypothesis 6 (H6).

Advanced ticket levels negatively moderate the impact of the consistency in attribute descriptions in low symbolic creativity extremely positive reviews on review usefulness.

4. Research Design

The method used in this study is shown in Figure 2, including five main steps: data collection, symbolic creativity quantification, latent semantic analysis, data operationalization, and regression analysis.

4.1. Research Background and Data Collection

Meituan.com is a leading online platform for life services, offering a broad range of business coverage and catering to nearly 700 million transaction users. Due to the platform’s enormous transaction volume, online consumer reviews have grown exponentially. Meituan.com is widely used in online review research, and hence, we collected online consumer reviews through this platform to ensure the replicability and generalizability of our research findings. We chose Disneyland as our research object due to its worldwide fame as the most-visited tourist destination and its widely acknowledged theming and creativity.

To gather data for our study, we collected review data of the Disneyland review module from Meituan.com between 2019 and 2022. The data dimensions we obtained included the reviewer ID, ticket type, review time, review text, number of pictures, review rating, number of likes, number of replies, and serious evaluation sign. We obtained a total of 30,712 original data. However, since short review texts contain minimal useful information, which could affect the accuracy of our model and the extraction of the review text’s topic, we only analyzed the reviews with a text length of over ten characters.

Figure 3 depicts a typical online consumer review, consisting of three parts: the consumer’s overall rating level, the consumer’s review content, and other consumer’s feedback. Part A of the review, as shown in Figure 3, refers to the overall rating evaluation given by the consumer on the service, which represents the consumer’s overall attitude towards the purchased or experienced service. Part B refers to the text and photos of the consumer’s review, providing a specific and detailed description of service attributes, such as “Tron Lightcycle Power Run,” “Musical Theater,” “Crossing the horizon,” and other specific service attribute words. Part C is the feedback provided by other consumers on this review, where the number of likes represents the overall usefulness votes of other consumers on this review, and the response refers to the interactive messages between consumers.

4.2. Quantification of Symbolic Creativity

According to the division of semiotic morphology [69], the symbolic creativity features in theme park entertainment items can be categorized into three groups: cultural symbols, identification symbols, and group symbols. Cultural symbols refer to a series of movies and performances that display the unique culture of theme parks, such as Disney Princesses, Frozen Balls, and others. Identification symbols refer to specific theme activities in theme parks, such as Halloween-themed activities, firework shows, singing festivals, and more. Group symbols refer to service groups and group interactions offered by amusement facilities and infrastructure, such as performers, role dolls, and interactive photos.

In order to measure the polarity of symbolic creativity in the review text, it is necessary to perform word tokenization and matching recognition processing on the review text based on the feature dictionary. Since the service products and related vocabulary of theme parks are domain-specific, existing dictionaries are not suitable for this study. Therefore, a symbolic creativity feature dictionary is constructed from the three dimensions mentioned earlier (cultural symbols, identification symbols, and group symbols), and added to the default Jieba word segmentation system. Python programs are used to tokenize text words, extract keywords, and calculate the word frequency ratio.

Table 1. Examples of quantification results of symbolic creativity features of review texts.

Review Text	Feature Word List	Word Frequency
It’s the second time I’ve come to Disney … After eating, we swipe Tron Light Wheel, Pirates of the Caribbean for the second time. Watching the float parade around two o’clock in the afternoon … I screamed excitedly when I saw Lulu and Belle … I want to watch the fireworks at night…	Belle, Pirates of the Caribbean, float parade, parade, fireworks, fireworks	6
Suggest you must go to Disney once if you have the chance, it is really great … I recommend a tour guide for you! It’s very cheap and not very expensive. It’s the first time I go to find one. Don’t miss the fireworks at night. It’s really beautiful and shocking. The dance is also super beautiful…	fireworks, tour guides, dance ball	3
It’s my first time to come to Disney, and I have interacted with many dolls to take pictures. I have played almost all the projects. It’s really good…	puppet	1

provides an example of the quantification results of symbolic creativity features of review texts.

To explore the impact of the quality of symbolic creativity impressions and the consistency in descriptions of various attributes of theme parks on the usefulness of extremely positive or negative reviews, this study divides the review texts into four data sets: low symbolic creativity extremely negative review, high symbolic creativity extremely negative review, low symbolic creativity extremely positive review, and high symbolic creativity extremely positive review. This study then conducts latent semantic analysis and regression analysis on each data set.

4.3. Latent Semantic Analysis

To uncover key latent factors in reviews, we employed Latent Semantic Analysis (LSA)—a powerful natural language processing method with technical and theoretical advantages [70]. LSA objectively extracts meaning from text by mapping document vectors, eliminating the need for subjective coding. LSA is linked to cognitive psychology theory and can extract hidden themes and contextual meaning, making it suitable for processing vast online review text data. To interpret the LSA results, we followed a process similar to exploratory factor analysis [71], marking each factor based on high-loaded words and documents and extracting widely discussed and representative service attribute factors. After finding distinct service attribute factors, we performed a regression analysis using a vector space [72].

4.4. Regression Analysis

4.4.1. Data Operationalization

(1)

Dependent variable. The dependent variable in this research model is review usefulness, which is measured by the logarithmic form of the usefulness votes received by online consumer reviews [73]. The logarithmic form can reduce the scale and make the data more stable because the number of usefulness votes has a relatively large degree of dispersion.

(2)

Independent variable. The independent variable is the coordinate data of each online review vector space on each factor, and the factors from the four types of samples are divided into four models according to the extremeness of the rating and the level of symbolic creativity in the reviews. Because the coordinates are orthogonal, there is no correlation between the independent variables. In order to consider the consistency in attribute descriptions among the reviews, the coordinate data of each factor attribute in the result obtained by LSA and the mean value of the corresponding attribute coordinate data of the 10 reviews near the review are processed by subtraction to obtain the consistency in attribute descriptions in reviews as the independent variable.

(3)

Moderating variable. Advanced ticket level is used as the moderating variable. Based on the ticket types obtained in the original data, the basic ticket without any special VIP service is used as the base point, and the VIP score of each ticket type is calculated according to the VIP service items and quantities included in different ticket types. The scores are mapped to advanced ticket levels of 1–4 based on its value range, where 1 represents the most common basic ticket, and 4 represents the highest-level advanced ticket with the largest number of VIP items.

(4)

Control variables. The control variables in this study include the length of the review text, the mark of serious evaluation of the review, the number of photos attached to the review, the number of interactive replies, the number of days the review is published, and the season of the review.

4.4.2. Statistical Analysis

This study utilized the Tobit regression model to analyze preprocessed data by LSA due to the limited range and skewed distribution of the usefulness vote variable and its ability to analyze non-negative dependent variables with potential selection bias. The Tobit model is preferred over OLS (Ordinary Least Square) regression as OLS treats zero values as missing data, but zero votes in this context indicate the usefulness of the review. Stata 15.0 software was used to analyze the data and assess the regression results.

The four groups of review data obtained based on the extremity and symbolic creativity feature classification of reviews are independent of each other. The four groups of review attribute factors (independent variables) obtained by LSA and the number of review usefulness votes (dependent variables) are respectively analyzed by regression, and the four corresponding independent models are as follows:

$$\mathrm{R}\mathrm{e}\mathrm{v}\mathrm{i}\mathrm{e}\mathrm{w} \mathrm{h}\mathrm{e}\mathrm{l}\mathrm{p}\mathrm{f}\mathrm{u}\mathrm{l}\mathrm{n}\mathrm{e}\mathrm{s}\mathrm{s}={\sum }_i\left({\beta }_i\mathrm{F}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}\_\mathrm{l}\mathrm{s}\mathrm{c}\mathrm{n}\_\mathrm{i}\right)+\sum {\alpha }_{i}cv+ϵ$$

(1)

$$\mathrm{R}\mathrm{e}\mathrm{v}\mathrm{i}\mathrm{e}\mathrm{w} \mathrm{h}\mathrm{e}\mathrm{l}\mathrm{p}\mathrm{f}\mathrm{u}\mathrm{l}\mathrm{n}\mathrm{e}\mathrm{s}\mathrm{s}={\sum }_j\left({\beta }_j\mathrm{F}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}\_\mathrm{h}\mathrm{s}\mathrm{c}\mathrm{n}\_\mathrm{j}\right)+\sum {\alpha }_{j}cv+ϵ$$

(2)

$$\mathrm{R}\mathrm{e}\mathrm{v}\mathrm{i}\mathrm{e}\mathrm{w} \mathrm{h}\mathrm{e}\mathrm{l}\mathrm{p}\mathrm{f}\mathrm{u}\mathrm{l}\mathrm{n}\mathrm{e}\mathrm{s}\mathrm{s}={\sum }_m\left({\beta }_m\mathrm{F}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}\_\mathrm{l}\mathrm{s}\mathrm{c}\mathrm{p}\_\mathrm{m}\right)+\sum {\alpha }_{m}cv+ϵ$$

(3)

$$\mathrm{R}\mathrm{e}\mathrm{v}\mathrm{i}\mathrm{e}\mathrm{w} \mathrm{h}\mathrm{e}\mathrm{l}\mathrm{p}\mathrm{f}\mathrm{u}\mathrm{l}\mathrm{n}\mathrm{e}\mathrm{s}\mathrm{s}={\sum }_n\left({\beta }_n\mathrm{F}\mathrm{a}\mathrm{c}\mathrm{t}\mathrm{o}\mathrm{r}\_\mathrm{h}\mathrm{s}\mathrm{c}\mathrm{p}\_\mathrm{n}\right)+\sum {\alpha }_{n}cv+ϵ$$

(4)

where cv is the control variable and ϵ is the random error. We further divide each of the above four groups of sample data based on the moderating variable (i.e., advanced ticket level) into two sets of sample data, that is, advanced ticket level 1 (i.e., low advanced ticket level) and advanced ticket level 2 and above (i.e., high advanced ticket level), and test each hypothesis in the two groups of data separately.

5. Data Analysis and Results

5.1. LSA Results and Descriptive Statistics

We gathered 30,712 online consumer reviews for Disney theme parks from 2019 to 2022 via Meituan.com. After extensive preprocessing, which included removing duplicates, short and meaningless reviews, and neutral ratings, we ended up with 5929 valid data points for subsequent analysis. These data consisted of 1090 low symbolic creativity extremely negative reviews, 709 high symbolic creativity extremely negative reviews, 1563 low symbolic creativity extremely positive reviews, and 2567 high symbolic creativity extremely positive reviews.

Using the LSA analysis procedure, we identified the theme park service attributes in the four data sets, as shown in

Table 2. Factors identified in the four sample groups.

Factor	Label	Singular Value	High-Loading Terms
LSCN1	Waiting time	9.022	Queuing, hours, waiting, time, admission
LSCN2	Ticketing experience	5.288	Tickets, refunds, customer service, Meituan, prices
LSCN3	Experience item	4.322	Horizon Fly, Pirates of the Caribbean, Tron Nimbus, Carousel, Mine Cart
LSCN4	Infrastructure	4.194	Signs, entrances, road signs, directions, toilets
HSCN1	Experience item	7.947	Horizon Fly, Tron Nimbus, Pirates of the Caribbean, Carousel, Mine Cart
HSCN2	Ornamental items	5.318	Show, fireworks, actors, show, light show
HSCN3	Membership	4.292	VIP, Early Access Card, Fast Track, Location, Membership
HSCN4	Paradise service	3.821	Service, staff, tour guide, attitude, customer service
HSCN5	Waiting time	3.637	Queuing, waiting, time, crowd, crowded
LSCP1	Experience item	10.358	Tron Nimbus, Over the Horizon, Pirates of the Caribbean, Thunder Mountain Rafting, Carousel
LSCP2	Waiting time	6.351	Queuing, hours, waiting, time, admission
LSCP3	Infrastructure	5.760	Entrance, signage, signage, restrooms, convenience
LSCP4	Membership	5.750	VIP, Fast Track, Early Access Card, Location, Skip the Line
HSCP1	Experience item	15.304	Horizon Fly, Carousel, Tron Nimbus, Pirates of the Caribbean, Mine Truck
HSCP2	Ornamental items	11.552	Show, fireworks, actors, show, float
HSCP3	Paradise service	9.454	Service, staff, tour guide, enthusiasm, recommendation
HSCP4	Playing experience	7.344	Dreamy, immersive, interactive, magical, atmospheric
HSCP5	Waiting time	7.115	Queuing, waiting, time, fast track, crowding

. The singular value indicates the amount of variance that each factor can explain [74]. A larger singular value indicates that the load items and corresponding attributes of the related factors should be discussed and analyzed in greater detail. The words and phrases corresponding to these factors are mentioned more frequently in online consumer reviews. For each factor, we kept the top five highly loaded words from 900 words, as these words comprised the key components or themes of the factor [75].

The results in Table 2 indicate that the identified factors are both logically consistent and reasonable. This is because their interpretation is based on the analysis of high-loading terms, and they describe each service attribute of the park according to the relevant theme park literature [3]. Each factor’s high-loading terms point to specific theme park service attributes, with the main service attribute factors that receive positive reviews being experience items and ornamental items. Moreover, the singular values reveal that the textual factors’ importance (represented by the frequency of a word in these factors) differs. Additionally, the factor composition of the four groups of samples also differs. Although there are some small differences in terms of high-loading terms, experience items and waiting time are the same for extreme reviews with low and high symbolic creativity. In contrast, game experience is more emphasized in high symbolic creativity reviews compared to low symbolic creativity reviews.

Table 3. Descriptive statistics.

Sample Group	Variable	Minimum	Maximum	Average	Standard Deviation
Low symbolic creativity extremely negative (N = 1090)	Review usefulness	0	3.807	0.293	0.566
	LSCN1	−0.467	0.731	0.000	0.157
	LSCN2	−0.417	0.495	0.000	0.135
	LSCN3	−0.337	0.735	0.000	0.118
	LSCN4	−0.409	0.436	0.000	0.113
	Text length	11	791	89.5	67.05
	Serious evaluation	0	1	0.057	0.232
	Number of photos	0	9	0.638	1.491
	Number of replies	14	1418	930.2	319.2
	Release days	0	10	0.236	0.820
	Season	1	4	2.237	0.943
High symbolic creativity extremely negative (N = 709)	Review usefulness	0	3.989	0.498	0.709
	HSCN1	−0.276	0.344	0.000	0.107
	HSCN2	−0.495	0.991	0.000	0.204
	HSCN3	−0.349	0.602	0.000	0.142
	HSCN4	−0.360	0.429	0.001	0.123
	HSCN5	−0.312	0.468	−0.001	0.118
	Text length	11	1019	156.6	123.3
	Serious evaluation	0	1	0.130	0.336
	Number of photos	0	9	1.158	1.952
	Number of replies	6	1419	870.4	331.3
	Release days	0	17	0.416	1.261
	Season	1	4	2.292	0.945
Low symbolic creativity extremely positive (N = 1563)	Review usefulness	0	4.419	0.167	0.410
	LSCP1	−0.289	0.368	0.000	0.116
	LSCP2	−0.303	0.845	0.001	0.166
	LSCP3	−0.482	0.578	0.000	0.135
	LSCP4	−0.402	0.595	0.000	0.123
	Text length	11	520	87.83	61.18
	Serious evaluation	0	1	0.267	0.442
	Number of photos	0	9	2.779	2.915
	Number of replies	0	1419	850.4	366.6
	Release days	0	27	0.254	1.080
	Season	1	4	2.525	1.106
High symbolic creativity extremely positive (N = 2567)	Review usefulness	0	4.060	0.310	0.572
	HSCP1	−0.269	0.356	0.000	0.092
	HSCP2	−0.473	0.884	0.000	0.147
	HSCP3	−0.538	0.554	0.000	0.124
	HSCP4	−0.372	0.470	0.001	0.119
	HSCP5	−0.445	0.473	0.000	0.117
	Text length	11	643	135.6	105.4
	Serious evaluation	0	1	0.449	0.498
	Number of photos	0	9	3.552	3.073
	Number of replies	6	1419	887.0	359.4
	Release days	0	26	0.549	1.881
	Season	1	4	2.537	1.082

presents the descriptive statistical results of the sample data for each group. The data distribution characteristics of the four groups of samples are similar.

5.2. Analysis of Regression Results

(1)

Regression analysis results of low symbolic creativity extremely negative reviews (LSCN)

Table 4. Tobit regression results for samples of low symbolic creativity extremely negative reviews.

Model	M1	M2	M3	M4	M5	M6	M7	M8	M9	M10	M11	M12	M13	M14	M15
Hypothesis Data packet	H1/H2					H3
	All					Low advanced ticket					High advanced ticket
Independent variable
LSCN1	0.826 **				0.814 **	0.614 *				0.634 *	2.304 **				2.182 **
	(0.31)				(0.31)	(0.32)				(0.32)	(0.97)				(0.97)
LSCN2		0.386			0.353		0.239			0.231		0.632			0.335
		(0.35)			(0.35)		(0.37)			(0.38)		(0.90)			(0.92)
LSCN3			−0.406		−0.465			−0.361		−0.510			−0.833		−0.626
			(0.43)		(0.44)			(0.51)		(0.52)			(0.77)		(0.77)
LSCN4				0.431	0.485				0.656	0.752				0.106	−0.234
				(0.43)	(0.43)				(0.48)	(0.48)				(0.89)	(0.90)
Control variable
Text length	0.005 ***	0.005 ***	0.005 ***	0.005 ***	0.004 ***	0.005 ***	0.005 ***	0.005 ***	0.005 ***	0.005 ***	0.005 **	0.007 ***	0.007 ***	0.007 ***	0.005 **
	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
Serious evaluation	0.430 *	0.430 *	0.421 *	0.426 *	0.448 *	0.418 *	0.421 *	0.413 *	0.420 *	0.432 **	0.274	0.045	0.085	0.052	0.277
	(0.21)	(0.21)	(0.21)	(0.21)	(0.21)	(0.22)	(0.22)	(0.22)	(0.22)	(0.22)	(0.74)	(0.75)	(0.75)	(0.76)	(0.74)
Number of photos	0.081 *	0.077 *	0.073 *	0.075 *	0.082 *	0.095 ***	0.090 **	0.088 **	0.090 **	0.096 ***	−0.099	−0.068	−0.097	−0.085	−0.099
	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)	(0.12)	(0.12)	(0.12)	(0.12)	(0.12)
Number of replies	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001	0.001 *	0.001 *	0.001 *	0.001
	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
Release days	0.378 ***	0.392 ***	0.395 ***	0.392 ***	0.381 ***	0.427 ***	0.439 ***	0.441 ***	0.444 ***	0.433 ***	0.222 ***	0.248 ***	0.248 ***	0.244 ***	0.230 ***
	(0.05)	(0.05)	(0.05)	(0.05)	(0.05)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)
Season	−0.035	−0.032	−0.033	−0.032	−0.034	0.005	0.009	0.007	0.010	0.005	−0.204 *	−0.219 *	−0.201 *	−0.218 *	−0.191
	(0.05)	(0.05)	(0.05)	(0.05)	(0.05)	(0.05)	(0.05)	(0.05)	(0.05)	(0.05)	(0.12)	(0.12)	(0.12)	(0.12)	(0.12)
Fit index
Constant	−2.165 ***	−2.199 ***	−2.200 ***	−2.202 ***	−2.170 ***	−2.371 ***	−2.399 ***	−2.406 ***	−2.407 ***	−2.389 ***	−1.042 **	−1.146 **	−1.159 **	−1.166 **	−1.035 **
	(0.22)	(0.22)	(0.22)	(0.22)	(0.22)	(0.25)	(0.25)	(0.25)	(0.25)	(0.25)	(0.51)	(0.52)	(0.52)	(0.53)	(0.51)
Pseudo R Squared	0.1323	0.1291	0.1289	0.1290	0.1340	0.1447	0.1426	0.1427	0.1436	0.1469	0.1288	0.1122	0.1145	0.1108	0.1316
Number of observations	1090	1090	1090	1090	1090	887	887	887	887	887	203	203	203	203	203

Note: The dependent variable is review usefulness; standard errors are in parentheses; * p < 0.05; ** p < 0.01; *** p < 0.001.

presents the regression analysis results for the samples of low symbolic creativity extremely negative reviews (LSCN). There are 15 models (M) divided into three main groups. M1 to M5 are used to test H1 in the condition of LSCN. The results from M1 to M4 show that only LSCN1 (Waiting time) has significant impact on the review usefulness ($\beta$ = 0.826, $p$ < 0.01). M5, which includes all factors, also confirmed this result. Therefore, H1 is supported in the condition of LSCN.

M6 to M15 are used to test H3, that is, whether the impact of the consistency in attribute descriptions in LSCN on the review usefulness is positively moderated by the advanced ticket level. The results show that for reviews with the advanced ticket level of less than 2 (M6 to M9), only LSCN1 (Waiting time) significantly affects the review usefulness ($\beta$ = 0.614, $p$ < 0.05); similarly, for reviews with the advanced ticket level of 2 or above (M11 to M14), only LSCN1 (Waiting time) has a significant impact on the review usefulness ($\beta$ = 2.304, $p$ < 0.01). According to the model results, the number of factors that significantly impact the review usefulness in LSCN corresponding to low and high advanced ticket levels is the same. These findings are confirmed by M10 and M15, which include all variables. Therefore, H3 is not supported.

(2)

Regression analysis results of high symbolic creativity extremely negative reviews (HSCN)

Table 5. Tobit regression results for samples of high symbolic creativity extremely negative reviews.

Model	M16	M17	M18	M19	M20	M21	M22	M23	M24	M25	M26	M27	M28	M29	M30	M31	M32	M33
Hypothesis Data packet	H1/H2					H5
	All					Low advanced ticket					High advanced ticket
Independent variable
HSCN1	0.019					0.086	0.225					0.064	−0.075					−0.280
	(0.52)					(0.53)	(0.61)					(0.67)	(0.88)					(0.88)
HSCN2		0.583 *				0.551 *		−0.190				−0.379		−0.256				−0.142
		(0.26)				(0.26)		(0.67)				(0.72)		(0.32)				(0.33)
HSCN3			1.004 **			0.998 **			0.858			0.895			0.776			0.881
			(0.35)			(0.35)			(0.46)			(0.47)			(0.49)			(0.50)
HSCN4				0.086		0.061				−0.263		−0.219				0.392		0.273
				(0.41)		(0.41)				(0.48)		(0.49)				(0.70)		(0.72)
HSCN5					−0.646	−0.661					−0.282	−0.306					−1.841 *	−2.093 *
					(0.44)	(0.45)					(0.51)	(0.52)					(0.80)	(0.80)
Control variable
Text length	0.004 ***	0.004 ***	0.003 ***	0.004 ***	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.005 ***	0.005 ***	0.005 ***	0.005 ***	0.005 ***	0.004 ***
	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
Serious evaluation	−0.000	0.004	0.013	−0.001	−0.007	0.008	0.139	0.139	0.145	0.145	0.138	0.143	−0.108	−0.085	−0.087	−0.091	−0.141	−0.088
	(0.19)	(0.19)	(0.19)	(0.19)	(0.19)	(0.18)	(0.22)	(0.22)	(0.22)	(0.22)	(0.22)	(0.22)	(0.34)	(0.34)	(0.33)	(0.34)	(0.33)	(0.33)
Number of photos	0.024	0.029	0.033	0.025	0.025	0.038	0.042	0.042	0.050	0.042	0.043	0.048	−0.074	−0.079	−0.066	−0.074	−0.068	−0.060
	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)	(0.06)	(0.06)	(0.06)	(0.06)	(0.06)	(0.06)
Number of replies	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.000	0.000	0.000	0.000	0.000	0.000
	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
Release days	0.262 ***	0.250 ***	0.253 ***	0.262 ***	0.263 ***	0.244 ***	0.354 ***	0.356 ***	0.345 ***	0.355 ***	0.352 ***	0.350 ***	0.163 ***	0.159 ***	0.161 ***	0.163 ***	0.170 ***	0.165 ***
	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)
Season	−0.050	−0.050	−0.053	−0.050	−0.050	−0.053	0.028	0.028	0.027	0.027	0.026	0.032	−0.295 ***	−0.298 ***	−0.300 ***	−0.294 ***	−0.299 ***	−0.308 ***
	(0.05)	(0.05)	(0.05)	(0.05)	(0.05)	(0.05)	(0.06)	(0.06)	(0.06)	(0.06)	(0.06)	(0.06)	(0.09)	(0.09)	(0.09)	(0.09)	(0.09)	(0.09)
Fit index
Constant	−1.378 ***	−1.411 ***	−1.336 ***	−1.381 ***	−1.376 ***	−1.361 ***	−1.864 ***	−1.883 ***	−1.851 ***	−1.861 ***	−1.869 ***	−1.869 ***	−0.091	−0.004	−0.000	−0.083	−0.059	0.076
	(0.21)	(0.21)	(0.20)	(0.20)	(0.20)	(0.20)	(0.26)	(0.26)	(0.26)	(0.26)	(0.26)	(0.26)	(0.31)	(0.32)	(0.31)	(0.31)	(0.30)	(0.32)
Pseudo R Squared	0.1226	0.1258	0.1278	0.1227	0.1240	0.1323	0.1292	0.1291	0.1322	0.1293	0.1293	0.1330	0.1866	0.1881	0.1923	0.1873	0.1994	0.2094
Number of observations	709	709	709	709	709	709	531	531	531	531	531	531	178	178	178	178	178	178

Note: The dependent variable is review usefulness; standard errors are in parentheses; * p < 0.05; ** p < 0.01; *** p < 0.001.

presents the regression analysis results for the samples of high symbolic creativity extremely negative reviews (HSCN). There are 18 models (M) divided into three main groups. M16 to M21 are used to test H1 in the condition of HSCN. The results from M16 to M20 show that only HSCN2 (Ornamental items) and HSCN3 (Membership) have a significant impact on the review usefulness ($\beta$ = 0.583, $p$ < 0.05; $\beta$ = 1.004, $p$ < 0.01). M21, which includes all factors, also confirms this result. Therefore, H1 is supported in the condition of HSCN.

M22 to M33 are used to test H5, that is, whether the impact of the consistency in attribute descriptions in HSCN on the review usefulness is negatively moderated by the advanced ticket level. The results show that for reviews with the advanced ticket level of less than 2 (M22 to M26), no factors significantly affect the review usefulness; for reviews with the advanced ticket level of 2 or above (M28 to M32), only HSCN5 (Waiting time) has a significant impact on the review usefulness ($\beta$ = −1.841, $p$ < 0.05). According to the model results, the number of factors that significantly impact the review usefulness in HSCN under high advanced ticket levels is more than that under low advanced ticket levels. These findings are confirmed by M27 and M33, which include all variables. Therefore, H5 is not supported.

From the regression analysis results of LSCN and HSCN, reviews with high symbolic creativity have more consistency in attribute descriptions that significantly affect the review usefulness compared to low symbolic creativity reviews in the condition of extremely negative ratings. Therefore, H2 is supported in the condition of extremely negative reviews.

(3)

Regression analysis results of low symbolic creativity extremely positive reviews (LSCP)

Table 6. Tobit regression results for samples of low symbolic creativity extremely positive reviews.

Model	M34	M35	M36	M37	M38	M39	M40	M41	M42	M43	M44	M45	M46	M47	M48
Hypothesis Data packet	H1/H2					H6
	All					Low advanced ticket					High advanced ticket
Independent variable
LSCP1	−1.020 *				−0.940 *	−0.750				−0.568	−0.789				−1.354
	(0.42)				(0.43)	(0.46)				(0.47)	(1.04)				(1.08)
LSCP2		0.263			0.184		0.421			0.374		−1.271			−1.222
		(0.28)			(0.28)		(0.28)			(0.28)		(1.93)			(1.99)
LSCP3			−0.481		−0.470			−0.688 *		−0.682 *			−0.408		−0.514
			(0.35)		(0.35)			(0.40)		(0.41)			(0.63)		(0.66)
LSCP4				0.438	0.361				0.771 *	0.756 *				−1.084	−1.359
				(0.37)	(0.37)				(0.41)	(0.41)				(0.78)	(0.83)
Control variable
Text length	0.004 ***	0.003 ***	0.003 ***	0.003 ***	0.004 ***	0.005 ***	0.005 ***	0.005 ***	0.005 ***	0.005 ***	0.011 *	0.009	0.009	0.010	0.012
	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)
Serious evaluation	0.176	0.164	0.177	0.167	0.180	0.149	0.134	0.150	0.136	0.148	0.380	0.574	0.586	0.489	0.318
	(0.12)	(0.12)	(0.12)	(0.12)	(0.12)	(0.12)	(0.12)	(0.12)	(0.12)	(0.12)	(0.87)	(0.85)	(0.85)	(0.85)	(0.87)
Number of photos	0.092 ***	0.088 ***	0.089 ***	0.089 ***	0.092 ***	0.116 ***	0.115 ***	0.116 ***	0.115 ***	0.117 ***	0.088	0.081	0.085	0.088	0.087
	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)	(0.02)	(0.06)	(0.06)	(0.06)	(0.05)	(0.05)
Number of replies	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001	0.001	0.001	0.001	0.001
	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
Release days	0.234 ***	0.233 ***	0.234 ***	0.234 ***	0.234 ***	0.193 ***	0.192 ***	0.193 ***	0.194 ***	0.194 ***	0.231 **	0.220 *	0.222 **	0.229 **	0.247 **
	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.09)	(0.09)	(0.09)	(0.08)	(0.09)
Season	−0.102 *	−0.101 *	−0.100 *	−0.101 *	−0.101 *	−0.071	−0.069	−0.068	−0.067	−0.068	−0.226	−0.224	−0.229	−0.213	−0.199
	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)	(0.13)	(0.13)	(0.13)	(0.14)	(0.14)
Fit index
Constant	−2.575 ***	−2.516 ***	−2.513 ***	−2.503 ***	−2.565 ***	−2.919 ***	−2.900 ***	−2.905 ***	−2.870 ***	−2.916 ***	−1.540 *	−1.498 *	−1.404 *	−1.499 *	−1.737 **
	(0.22)	(0.22)	(0.22)	(0.21)	(0.22)	(0.25)	(0.25)	(0.25)	(0.25)	(0.25)	(0.62)	(0.61)	(0.60)	(0.61)	(0.65)
Pseudo R Squared	0.1449	0.1423	0.1428	0.1426	0.1465	0.1681	0.1678	0.1683	0.1687	0.1730	0.1404	0.1400	0.1399	0.1445	0.1515
Number of observations	1563	1563	1563	1563	1563	1290	1290	1290	1290	1290	273	273	273	273	273

Note: The dependent variable is review usefulness; standard errors are in parentheses; * p < 0.05; ** p < 0.01; *** p < 0.001.

presents the regression analysis results for the samples of low symbolic creativity extremely positive reviews (LSCP). There are 15 models (M) divided into three main groups. M34 to M38 are used to test H1 in the condition of LSCP. The results from M34 to M37 show that only LSCP1 (Experience item) has a significant impact on the review usefulness ($\beta$ = −1.020, $p$ < 0.05). M38, which includes all factors, also confirmed this result. Therefore, H1 is supported in the condition of LSCP.

M39 to M48 are used to test H6, that is, whether the impact of the consistency in attribute descriptions in LSCP on the review usefulness is negatively moderated by the advanced ticket level. The results show that for reviews with the advanced ticket level of less than 2 (M39 to M42), LSCP3 (Infrastructure) and LSCP4 (Membership) significantly affected the review usefulness ($\beta$ = −0.688, $p$ < 0.05; $\beta$ = 0.771, $p$ < 0.05); for reviews with the advanced ticket level of 2 or above (M44 to M47), no factors significantly affected the review usefulness. According to the model results, the number of factors that significantly impact the review usefulness in LSCP under low advanced ticket levels is more than that under high advanced ticket levels. These findings are confirmed by M43 and M48, which include all variables. Therefore, H6 is supported.

(4)

Regression analysis results of high symbolic creativity extremely positive reviews (HSCP)

Table 7. Tobit regression results for samples of high symbolic creativity extremely negative reviews.

Model	M49	M50	M51	M52	M53	M54	M55	M56	M57	M58	M59	M60	M61	M62	M63	M64	M65	M66
Hypothesis Data packet	H1/H2					H4
	All					Low advanced ticket					High advanced ticket
Independent variable
HSCP1	−0.034					0.116	0.016					0.141	−0.306					−0.756
	(0.32)					(0.33)	(0.35)					(0.36)	(0.82)					(0.89)
HSCP2		0.417 *				0.427 *		0.544				0.609		0.032				0.006
		(0.19)				(0.20)		(0.30)				(0.31)		(0.21)				(0.23)
HSCP3			0.083			0.068			−0.072			−0.157			0.403			0.270
			(0.22)			(0.22)			(0.25)			(0.25)			(0.42)			(0.49)
HSCP4				−0.534 *		−0.538 *				−0.098		−0.126				−0.890 *		−0.901 *
				(0.23)		(0.23)				(0.27)		(0.28)				(0.39)		(0.45)
HSCP5					−0.069	−0.088					0.032	−0.010					−0.275	0.079
					(0.24)	(0.24)					(0.28)	(0.28)					(0.40)	(0.45)
Control variable
Text length	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.003 ***	0.004 **	0.004 **	0.004 **	0.004 *	0.004 **	0.004 **
	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
Serious evaluation	0.014	0.017	0.013	0.016	0.013	0.018	0.077	0.078	0.078	0.078	0.078	0.077	−0.039	−0.041	−0.036	−0.022	−0.030	−0.012
	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.07)	(0.27)	(0.27)	(0.27)	(0.27)	(0.27)	(0.27)
Number of photos	0.076 ***	0.076 ***	0.075 ***	0.076 ***	0.076 ***	0.076 ***	0.092 ***	0.092 ***	0.092 ***	0.092 ***	0.092 ***	0.092 ***	0.071 *	0.070 *	0.069 *	0.073 *	0.070 *	0.074 *
	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)
Number of replies	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.001 ***	0.000	0.000	0.000	0.000	0.000	0.000
	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
Release days	0.162 ***	0.162 ***	0.163 ***	0.162 ***	0.162 ***	0.161 ***	0.144 ***	0.145 ***	0.144 ***	0.144 ***	0.144 ***	0.144 ***	0.165 ***	0.165 ***	0.167 ***	0.166 ***	0.167 ***	0.166 ***
	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.01)	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)	(0.04)
Season	−0.134 ***	−0.135 ***	−0.134 ***	−0.134 ***	−0.134 ***	−0.134 ***	−0.129 ***	−0.131 ***	−0.129 ***	−0.129 ***	−0.129 ***	−0.131 ***	−0.139	−0.140	−0.141	−0.134	−0.140	−0.129
	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.03)	(0.09)	(0.09)	(0.09)	(0.08)	(0.09)	(0.08)
Fit index
Constant	−1.963 ***	−1.974 ***	−1.961 ***	−1.967 ***	−1.962 ***	−1.975 ***	−2.385 ***	−2.388 ***	−2.386 ***	−2.386 ***	−2.385 ***	−2.381 ***	−0.417	−0.387	−0.373	−0.437	−0.382	−0.521
	(0.12)	(0.12)	(0.12)	(0.12)	(0.12)	(0.12)	(0.14)	(0.14)	(0.14)	(0.14)	(0.14)	(0.14)	(0.37)	(0.36)	(0.36)	(0.35)	(0.36)	(0.37)
Pseudo R Squared	0.1715	0.1725	0.1715	0.1727	0.1715	0.1738	0.2063	0.2072	0.2063	0.2064	0.2063	0.2074	0.0998	0.0996	0.1009	0.1071	0.1002	0.1084
Number of observations	2567	2567	2567	2567	2567	2567	2164	2164	2164	2164	2164	2164	403	403	403	403	403	403

Note: The dependent variable is review usefulness; standard errors are in parentheses; * p < 0.05; ** p < 0.01; *** p < 0.001.

presents the regression analysis results for the samples of high symbolic creativity extremely negative reviews (HSCP). There are 18 models (M) divided into three main groups. M49 to M54 are used to test H1 in the condition of HSCP. The results from M49 to M53 show that HSCP2 (Ornamental items) and HSCP4 (Playing experience) have a significant impact on the review usefulness ($\beta$ = 0.417, $p$ < 0.05; $\beta$ = −0.534, $p$ < 0.05). M54, which includes all factors, also confirms this result. Therefore, H1 is supported in the condition of HSCP.

M55 to M66 are used to test H4, that is, whether the impact of the consistency in attribute descriptions in HSCP on the review usefulness is positively moderated by the advanced ticket level. The results show that for reviews with the advanced ticket level of less than 2 (M55 to M59), no factors significantly affect the review usefulness; for reviews with the advanced ticket level of 2 or above (M61 to M65), HSCP4 (Playing experience) has significant impact on the review usefulness ($\beta$ = −0.890, $p$ < 0.05). According to the model results, the number of factors that significantly impact the review usefulness in HSCP under high advanced ticket levels is more than that under low advanced ticket levels. These findings are confirmed by M60 and M66, which include all variables. Therefore, H4 is supported.

From the regression analysis results of LSCP and HSCP, reviews with high symbolic creativity have more consistency in attribute descriptions that significantly affect the review usefulness compared to low symbolic creativity reviews in the condition of extremely positive ratings. Therefore, H2 is supported in the condition of extremely positive reviews.

As shown in

Table 8. Multicollinearity test.

Sample Group	Variable	VIF	1/VIF	Sample Group	Variable	VIF	1/VIF
Low symbolic creativity extremely negative	LSCN1	1.08	0.924	Low symbolic creativity extremely positive	LSCP1	1.20	0.836
	LSCN2	1.01	0.989		LSCP2	1.02	0.977
	LSCN3	1.02	0.979		LSCP3	1.00	0.995
	LSCN4	1.00	0.998		LSCP4	1.01	0.990
	Text length	1.22	0.818		Text length	1.59	0.628
	Serious evaluation	1.46	0.686		Serious evaluation	1.62	0.618
	Number of photos	1.46	0.686		Number of photos	1.35	0.740
	Number of replies	1.08	0.930		Number of replies	1.07	0.932
	Release days	1.01	0.991		Release days	1.02	0.981
	Season	1.01	0.994		Season	1.01	0.993
	Mean VIF	1.13	/		Mean VIF	1.19	/
High symbolic creativity extremely negative	HSCN1	1.29	0.777	High symbolic creativity extremely positive	HSCP1	1.25	0.803
	HSCN2	1.11	0.899		HSCP2	1.05	0.949
	HSCN3	1.08	0.924		HSCP3	1.01	0.995
	HSCN4	1.03	0.967		HSCP4	1.00	0.997
	HSCN5	1.04	0.963		HSCP5	1.02	0.979
	Text length	1.47	0.682		Text length	1.48	0.675
	Serious evaluation	1.69	0.593		Serious evaluation	1.52	0.656
	Number of photos	1.74	0.574		Number of photos	1.47	0.679
	Number of replies	1.10	0.910		Number of replies	1.18	0.845
	Release days	1.03	0.972		Release days	1.05	0.950
	Season	1.00	0.997		Season	1.02	0.983
	Mean VIF	1.23	/		Mean VIF	1.19	/

, we tested for multicollinearity by calculating the Variance Inflation Factor (VIF) value, which is a possible problem in regression analysis [73]. All variables have acceptable VIF values (much less than 10) and tolerance levels (closer to 1), so there is no multicollinearity problem among the variables in this study.

5.3. Summary of Research Hypotheses

Table 9. Summary of research hypotheses.

Hypothesis	Hypothetical Content	Result
H1	Not all consistency in attribute descriptions plays an important role in the review usefulness, whether they are in low symbolic creativity reviews or high symbolic creativity reviews.	Supported
H2	In terms of consistency in attribute description, there are a greater number of attributes that significantly influence the review usefulness in high symbolic creativity reviews than in low symbolic creativity reviews.	Supported
H3	Advanced ticket levels positively moderate the impact of the consistency in attribute descriptions in low symbolic creativity extremely negative reviews on review usefulness.	Not Supported
H4	Advanced ticket levels positively moderate the impact of the consistency in attribute descriptions in high symbolic creativity extremely positive reviews on review usefulness.	Supported
H5	Advanced ticket levels negatively moderate the impact of the consistency in attribute descriptions in high symbolic creativity extremely negative reviews on review usefulness.	Not Supported
H6	Advanced ticket levels negatively moderate the impact of the consistency in attribute descriptions in low symbolic creativity extremely positive reviews on review usefulness.	Supported

summarizes the main results of this study. It can be seen that the two main effect hypotheses H1 and H2 all passed the test; among the moderating effect hypotheses, H3 and H5 failed the test.

Based on the empirical analysis presented in Table 9, it is evident that consumers tend to simplify the information they consider when making decisions about the usefulness of reviews in an information overload environment. As a result, they may not pay attention to all of the service attributes. Furthermore, the attributes that consumers pay attention to and that significantly influence the evaluation of review usefulness in different types of reviews are inconsistent. For example, in LSCN, the attribute that has a significant impact on review usefulness is “Waiting time”, whereas in HSCN, the attributes that have a significant impact are “Ornamental items” and “Membership”. In LSCP, the attribute that has a significant impact is “Experience item,” whereas in HSCP, the attributes that have a significant impact are “Ornamental items” and “Playing experience”. Thus, the factors that consumers consider when evaluating review usefulness are influenced by their level of cultural adaptation to theme parks. Consumers who have a higher degree of adaptation to theme park culture, especially in high symbolic creativity reviews, are more likely to pay attention to attribute information in the reviews, which can significantly impact the review usefulness.

Another finding of this study is that advanced ticket levels do not have a moderating effect on the usefulness of extremely negative reviews. Although advanced ticket levels can raise consumer expectations and attract attention, extremely negative reviews can also raise consumer risk perception, leading them to pay attention. In such a scenario, the attention caused by high expectations is ignored. The reference point for consumers to evaluate the review usefulness dynamically changes with advanced ticket levels under the assumption of the moderating variables in this study. The research results indicate that advanced ticket levels only have a moderating effect on the consistency in attribute descriptions in consumers’ review usefulness evaluation decisions in extremely positive reviews. In high symbolic creativity extremely positive reviews, advanced ticket levels play a positive moderating role, whereas in low symbolic creativity extremely positive reviews, they play a negative moderating role. This indicates that consumers have cognitive dissonance when evaluating the online review usefulness of theme parks. When individuals have higher expectations and are more familiar with the culture, they tend to actively seek more information to eliminate their internal cognitive dissonance. Conversely, when individuals have higher expectations and are less familiar with the culture, they tend to seek less information to avoid further increasing their internal cognitive dissonance. The less familiar consumers are with the culture, the more uncertain they are about their ability to understand and process more information, leading to anxiety. Thus, they choose to avoid understanding too much information in depth to reduce their internal cognitive dissonance.

6. Research Conclusions

This study aims to explore the impact of factors influencing consumer cultural adaptation in online reviews and the interactive effects of these factors on the usefulness of theme park online reviews for the first time. Specifically, it analyzes the clarity and specificity of symbolically creative descriptions, consistency of service attribute portrayal across reviews, and the impact of extreme framing and tone of attribute descriptions on review usefulness. Additionally, this study investigates how the level of symbolic creativity and extremity of reviews affect consumers’ evaluation of attribute description consistency in review usefulness decision making, while also taking into account the moderating effect of different decision reference points (depicted by advanced ticket level).

The findings of this study reveal that, within the context of information overload, consumers tend to simplify their decision-making process when evaluating the usefulness of reviews. As a result, certain attributes may be overlooked, and the impact of this oversight varies among different types of reviews. Furthermore, cultural adaptation to theme parks significantly influences this behavior, as higher levels of adaptation lead to increased attention towards attribute information in reviews. It is important to note that the advanced ticket level does not moderate the usefulness of extremely negative reviews. Although the high expectations associated with advanced tickets may initially increase attention towards these reviews, the heightened perception of risk can sometimes override this attention boost. On the other hand, in the case of extremely positive reviews, the moderating effect of advanced ticket levels on the consistency in attribute descriptions varies based on the level of symbolic creativity. In instances where consumers are highly familiar with the theme park culture, there is a greater tendency to seek out information in order to reduce cognitive dissonance. Conversely, when consumers have low familiarity with the culture, there is a tendency to avoid gathering information due to the presence of uncertainty.

6.1. Theoretical Contributions

This study addresses two key research gaps. Firstly, it investigates the effects of high-symbol and low-symbol creative reviews on consumer evaluations of review consistency in the context of information overload. Secondly, it analyzes the impact of consistency in theme park attribute descriptions in high-symbol and low-symbol creative reviews on review usefulness. By examining these gaps, this research aims to provide insights into how tourism managers can leverage online reviews to optimize their service attributes and enhance their customers’ experiences.

This study makes a significant contribution to the academic literature by shedding light on how reviews with varying levels of symbolic creativity influence consumer evaluations of review consistency and usefulness in the context of theme park attributes. Our findings indicate that, when assessing usefulness, consumers tend to simplify their decision-making process and do not consider the consistency of all attribute descriptions mentioned in reviews. However, when confronted with reviews that exhibit high levels of symbolic creativity, consumers may pay more attention to the service attributes mentioned, thereby influencing their evaluation of review usefulness. This attention to attribute consistency is driven by cultural adaptation to the symbolic creative culture within theme parks. Unlike previous research that primarily explored how symbolic creativity impacts consumer satisfaction and loyalty [29,30,31], this study offers fresh insights by focusing on the cultural adaptation to symbolic creativity within theme parks and investigating how this adaptation mechanism influences the usefulness of online reviews.

Moreover, this study addresses the dearth of research on how extreme reviews impact consumers’ evaluation of consistency in attribute descriptions when making usefulness evaluation decisions amidst information overload in theme parks. Additionally, we investigate how different levels of advanced ticket prices impact the time and effort consumers invest in evaluating consistency in attribute descriptions during the usefulness evaluation process. Consequently, this study offers insights into the impact of extreme reviews on consumer decision making and advances the literature by examining the moderating effect of different levels of advanced ticket prices on consumers’ attention to consistency in attribute descriptions in reviews.

Our study has revealed that advanced ticket levels do not influence how consumers evaluate extremely negative reviews, indicating that advanced ticket levels do not change consumers’ perceptions of such reviews. This emphasizes the significant impact of extremely negative reviews in triggering consumers’ perception of risk and shaping their attention towards potential risks associated with a service. Additionally, our observation that risk perception can influence consumer attention suggests that when consumers are highly focused on risks, they may disregard the attention drawn by high expectations. The moderating effect of advanced ticket levels on the consistency in attribute descriptions in consumer evaluations of reviews further highlights that the influence of advanced ticket levels may vary depending on the specific context. This implies that the impact of advanced ticket levels is not consistent across all situations. Moreover, our research shows that consumer expectations and familiarity with a particular culture can impact their evaluation of review usefulness. Consumers with higher expectations and greater cultural familiarity are more likely to seek additional information in order to resolve any cognitive dissonance they may experience. On the other hand, consumer discomfort with unfamiliar cultures may discourage them from seeking in-depth information as a means to reduce cognitive dissonance when they feel uncertain or anxious about unfamiliar cultural contexts. While previous studies have mainly focused on the influence of cultural adaptation effects on consumer satisfaction or loyalty [12,56], our research goes further by examining the moderating effect of advanced ticket levels on the usefulness evaluation of reviews based on cultural adaptation effects. This provides valuable insights into the role of advanced ticket levels in shaping consumers’ assessment of attribute consistency during the decision-making process in various situations involving cultural adaptation dynamics.

6.2. Management Implications

Our research on theme park service attributes can provide valuable insights for theme park managers to understand the specific service attributes that consumers value the most when evaluating online reviews from other consumers. These attributes include experiential projects, waiting time, viewing projects, and overall playing experience. Additionally, our study found that low-symbolic creative reviews are more likely to mention experiential projects and waiting time, while high-symbolic creative reviews are more likely to mention viewing projects and overall playing experience. This information can help theme park managers understand the differences in service attributes that consumers focus on in different types of reviews.

This study further finds that the type of advanced ticket can also impact consumers’ expectations of theme park services. Theme park managers need to be careful in providing services that align with the value of the premium ticket types to avoid any negative impact on consumer satisfaction caused by gaps between expectations and reality. This study recommends that theme parks promote attention towards creative-related attributes and enhance symbolic creativity to improve consumer loyalty. This can be achieved by cultivating word-of-mouth marketing, promoting the symbolic creative culture of the theme park, and increasing the level of advanced ticket pricing to enhance consumers’ attention to the attributes of the theme park and, in turn, boost consumer loyalty.

For new users who are unfamiliar with the symbolic creative culture of the theme park, this study recommends the introduction of low-priced basic advanced tickets to attract these consumers’ attention to the symbolic creative culture of the theme park and enhance consumer stickiness.

6.3. Limitations and Future Directions

However, this study has certain limitations. Firstly, this research focuses only on Disneyland, limiting the application and promotion of research results to other tourism products that offer different attributes and experiences. Future research should include various types of tourism products to broaden the scope of the research. Secondly, this study only considers online review data from Chinese websites and excludes reviews from foreign language websites. Consumer characteristics vary across different language and cultural backgrounds. Therefore, multilingual data samples must be collected and analyzed to expand the applicability of the research results.