The effect of online review and interaction on value co-creation in tourism virtual community

Abstract

User behavior in the tourism virtual community has become a hot research topic. Based on cognitive evaluation theory and organismic integration theory, this paper attempts to reveal the generation mechanism of user value co-creation behavior in tourism virtual community, which is of great significance for the operation and development of tourism industry. Altogether 342 valid questionnaires were collected in the survey, and structural equation modelling was employed to test the impact path. It was found that online reviews have two dimensions of characteristics, content and attributes. Online review characteristics positively influence the social interaction and information exchange dimensions of social interaction. Social interaction positively influences value co-creation behavior. The two dimensions of social interaction have a positive effect on user engagement, in turn, user engagement itself has a positive influence on recommendation behavior. Overall, the findings of the study can promote user participation in the interaction and information exchange behaviors.

Keywords:

• online review
• social interaction
• value co-creation behavior
• tourism virtual community
• online travel

1. Introduction

The development of online virtual communities has brought new ways of information exchange, interpersonal interaction, and knowledge sharing among Internet users, who can browse, access and share relevant information and knowledge through online virtual communities (Guan et al., 2018). A virtual community, also known as an online community or e-community, is an online community that exists in the internet environment and uses electronic tools as a medium for people to communicate online about certain interests and needs (Luo et al., 2021; Wang et al., 2021). According to research conducted by Wang et al. (2021) and Kim and Kim (2021), tourists frequently utilize tourism virtual communities for activities such as browsing, acquiring and sharing relevant information, and knowledge. This behavior tends to trigger knowledge sharing, travel intention and value creation among tourists.

In the tourism virtual community, online reviews serve as an interactive tool between tourists, tourism service providers, and potential tourists. Online review as an effective method for tourists to express their own opinions in terms of User-generated Content (UGC). However, online virtual community is equivalent to natural environment, the whole community should be sustainable for tourists to express and browse other opinions since the community is like a big family for netizens. They could find people with common habits and their opinions may be important for other tourists to be familiar with targeted destinations. Therefore, online reviews could reflect general knowledge of destinations, meanwhile, online reviews also act as an vital tool for tourist to understand a whole virtual community. According to Wei et al. (2013), these reviews are crucial in shaping users’ perceptions and behavior regarding tourism services. Research suggests that online reviews, whether in digital or textual format, may influence users’ perceptions of key factors such as product quality, value, and attributes in the community (Shin et al., 2020). Due to the complexity of the review text, the length, rating, and sentiment rating of the reviews can make a difference to user perceptions (Costa et al., 2019). Furthermore, online reviews are an important social tool in tourism virtual communities, where users communicate and interact with community members by posting reviews or receiving reviews from others, and the richer the content of online reviews, the more frequent the communication and interaction will be, which in turn will lead to purchase and recommendation behaviors (Marcolin et al., 2021). As the wave of online communities, people are prone to find more information through some particular moments, people with common interests will provide more effective help. However, those people are not familiar with each other, their relationships are closely related by their online communication, namely online review. People in a particular community are able to browse and outspoke their inner thoughts, which shape the atmosphere of communities. Henceforth, online review and communication are the core in understanding online communities.

Through online review browsing and expressing behavior, tourists individual value could be reinforced or altered. Our standpoint are not invariable in the whole process as our value could be shaped by our parents and surrounding environment in our childhood. Gradually, our views of the world are affected by various factors. In online tourism community, tourists are surrounded by others opinions and thoughts, so their value will be influenced by online reviews. Their value may be reinforced or co-created in the course. Furthermore, people in the community are prone to express themselves more since they could find valuable reviews which may open their mind. Moreover, those reviews may have relations with people’s interaction and exchange of some information. Tourism value co-creation is a process of co-creation of value through the interaction and collaboration of participating subjects such as tourists, enterprises, and destinations (Font et al., 2021). From the perspective of tourists, it mainly involves engagement and recommendation behavior (Yen et al., 2020). Engagement behavior is reflected in active participation and engagement in tourism activities. Recommendation behavior refers to sharing travel experiences and information with other tourists through social media.

Previous research on value co-creation behaviors in tourism virtual communities has focused on user interactions, sharing, and recommendations (Islam & Kirillova, 2021; Shen et al., 2020). However, most scholars have only focused on online reviews and interactions (Li et al., 2019; Lin et al., 2022) or the impact of destination brand communities on customer’s value change (Cheung et al., 2020). Tourists share their travel experience and experience through social platforms, which can promote positive interaction between tourists and other participants, and enhance the social interaction among them. It is also a process of improving personalized needs and tourists’ satisfaction, which has an essential impact on creating tourism value. However, previous studies have scarcely considered the combined effects of online reviews and social interaction on tourists’ value-creation behavior. Whether online comment features and social interaction can affect the value co-creation behavior of tourism virtual communities is a topic worthy of discussion.

The contributions of this study are mainly in the following three aspects. First, this paper constructs a second-order structural equation model using online reviews as a higher-order variable to reveal the influence of each feature dimension of online reviews on users’ value co-creation behavior in tourism virtual communities. Second, this study analyses the impact of online review features on each dimension of social interaction to further expand the research area of online reviews. Finally, this paper analyses in detail the mechanism of the influence of online review features and social interactions on value co-creation behavior, adding new insights to the study of value co-creation behavior in tourism virtual communities.

2. Theoretical background

2.1. Online reviews and online travel reviews

Online review is the true evaluation of products or services by consumers on the network platform, and is a common form of online word-of-mouth (Changchit et al., 2022). Through online reviews, users can express their views and opinions through text, pictures, and videos and communicate and interact with other users (Namvar & Chua, 2022; Wang et al., 2022). First, it can provide a platform for users to express their opinions directly here, regardless of the topic (Graf-Vlachy et al., 2021; Shao et al., 2021).

Second, online reviews can enable content creators to obtain feedback and suggestions from readers, so as to continuously improve and enhance the quality and value of their content (Du et al., 2021). Finally, online reviews facilitate communication and interaction, allowing users to understand other users’ perspectives more easily and to exchange experiences and knowledge (Kim & Hyun, 2021). In the tourism industry, online travel reviews are closely related to user behavior. Users can conduct social interaction through online travel reviews, share ideas and opinions, and get more feedback and responses (Moro & Stellacci, 2023). This interaction can not only increase users’ online social experience, but also help users better understand and master information. Users’ emotions and attitudes, such as feeling angry and happy, are also affected by online travel reviews (Chen et al., 2021; Zhao & Li, 2021). This influence can further cause changes in user behavior, such as engagement and recommendation. In addition, online travel reviews will also affect users’ trust and loyalty to brands and websites (Zheng et al., 2021). If there are a lot of negative reviews on a brand or website, users may lose trust in it, which will reduce user loyalty.

2.2. Tourism value co-creation behavior

Tourism value co-creation behavior refers to the specific behavior of tourists in participating in value co-creation, which is a positive and active behavior with strong purpose and autonomy (Teng & Tsai, 2020). Tourism value co-creation behavior is a multidimensional concept. According to its role in tourism services, tourism value co-creation behavior is divided into two dimensions: participation behavior and citizen behavior. Participation behavior refers to tourists participating in service production and service provision, which is mainly manifested in information collection, participation in production, and social interaction before and during tourism (Chen et al., 2022; Teng & Tsai, 2020). Citizen behavior is the out-of-role behavior that tourists show in tourism activities that is beneficial to others, mainly manifested in experience sharing, information feedback and helping others.

In the tourism virtual community, the value co-creation behavior is mainly manifested as online participation behavior, which can be understood from the perspective of environment, organization and individual. At the environmental level, the environmental characteristics and online scenes of the tourism virtual community will have a positive impact on the user’s participation behavior (Campos et al., 2018). At the organizational level, the service quality and community brand of the tourism virtual community will also affect the user’s participation behavior (Lei et al., 2020). At the personal level, online interaction, perceived value, and the quality of self-content generation are the main factors that affect the user’s participation behavior, and enable users to show specific tourism value co-creation behavior such as engagement and recommendation (Qu et al., 2022). According to the relevant knowledge of psychology, the psychological structure of attitude contains three important systems: cognition, emotion and behavior (Breckler, 1984). Individuals’ cognition and emotion significantly affect their behavioral patterns. Therefore, if a tourist becomes interested in a online community, he or she may devote time in the community. Online reviews from other tourists in this community may arise the tourist huge interests and affect their emotion and cognitive reaction, which is conducive to their further online participation behavior. Specifically, an innovative review may attract people with same minds, which is beneficial to their future participation behavior. With the development of personalized individual tourism, people are no longer one-way information receiver but information sharer, they are more eager to display and share their own experiences and stories in the course of a journey. Therefore, tourists can gain insights and supports from browsing reviews and social interactions, besides, they can experience fun, teamwork and encouragement, and thus their value could be co-created in the process of experiencing online community (Luo et al., 2019).

2.3. Cognitive evaluation theory and organismic integration theory

Cognitive evaluation theory and organismic integration theory are two important dimensions of self-determined theory (Gagné & Deci, 2005). Cognitive evaluation theory focuses on the impact of social contextual factors on individual internal motivation (Schunk & Dibenedetto, 2020). Internal motivation is the tendency of individuals to pursue interests and social interaction, and to seek and overcome challenges in the process. As a typical self-determined behavior, the maintenance of internal motivation depends on the support of social context, which can meet the basic needs of individuals (Howard et al., 2020).

The organismic integration theory focuses on the internalization and integration of individual motivation and believes that the internalization of external motivation is a positive and spontaneous process like internal motivation (Gilal et al., 2022). Motivation internalization can be understood as individuals absorbing certain values and attitudes, so that the output of behavior is no longer dependent on external rewards and punishment and is based on internal motivation. This makes it easier for individuals to comply with social rules, thus forming a more self-regulatory and self-regulated behavior output (Täuber et al., 2018). After individuals absorb specific values and attitudes in the process of internalization, they are more likely to display the behavior of creating and maintaining values (Shulga & Busser, 2021). In the tourism virtual community, users are affected by the factors of online tourism reviews, showing the internalization motivation of social interaction, and finally producing value co-creation behaviors such as engagement and recommendation. Therefore, based on cognitive evaluation theory and organismic integration theory, this paper follows the research logic of “Contextual Cognition—Internalization Motivation—Behavioral Outcomes” to reveal the generation mechanism of user value co-creation behavior in tourism virtual community.

3. Research hypotheses

3.1. The content and attribute characteristics of online reviews

The content of online reviews is the reflection of the publisher’s knowledge, which can present the microscopic characteristics of online reviews from the perspective of word granularity. The research on the content characteristics of online reviews mainly focuses on language style, emotional orientation, readability, and text semantics (Li et al., 2020). Among them, language style and emotional content are two important content features of online reviews. Linguistic style refers to the linguistic expressions used in reviews, including linguistic forms, sentence structures, and diction features (Shin et al., 2019). Language style exerts an important influence on the readability and impact of reviews (Yang et al., 2021). For example, some reviews may be more persuasive because they use powerful wording and logical structures. Emotional content refers to the emotional overtones contained in a review, including positive, negative, or neutral overtones (Huang & Liang, 2021). Mining emotional content from online reviews can identify positive or negative emotions in reviews and help consumers quickly obtain useful information (Li et al., 2020), thus assisting them in making the right purchase decisions. Both language style and emotional content are substantive expressions of consumers’ opinions about products and services, reflecting the content characteristics of online reviews.

Attribute characteristics are users’ perception and judgment of online comment information, mainly including review length, rating, and grade (Yang et al., 2021). Review length and review rating are two important attribute characteristics of online reviews. Review length refers to the number of words in a review, which is usually calculated in terms of the number of characters or words. Based on the length of the review text, online reviews can be divided into long-text reviews and short-text reviews (Wang et al., 2020). Longer reviews tend to have richer content information and may contain more details, which can better reflect the real thoughts and feelings of the reviewer (Yang et al., 2021). Shorter reviews are some simple evaluations or expressions that do not provide enough information.

Review rating refers to the reviewer’s overall assessment of a product or service, usually expressed as a star or numerical rating (Liang, 2021), with high and low ratings representing positive and negative consumer evaluations, respectively. Higher review ratings usually indicate higher reviewer satisfaction with a product or service, and higher ratings are more likely to be adopted, resulting in positive attitudes and facilitating consumer purchases, in contrast, lower review ratings indicate higher dissatisfaction (An et al., 2020), which can contribute to negative product attitudes and inhibit consumer purchases. When reading reviews, people tend to focus first on the review rating because it is an indicator of the overall evaluation. Therefore, the following hypotheses are proposed.

H1:

The four distinct, but related sub-dimensions of online reviews characteristics can be accounted for by a common underlying higher order online reviews characteristics factor model which is significantly better than a first-order service online reviews characteristics model.

3.2. The effects of online reviews on online interaction

In virtual communities, social interaction is a significant way of interaction among community members. Social interaction can be a way of interaction and communication among community members and for people to exchange their views and experiences. This paper classifies social interaction in tourism virtual communities into two types of social interaction and information exchange and analyses the impact of online reviews on the two types of social interaction as follows.

Online reviews provide a platform for users to share travel experiences, provide travel advice, share travel photos, etc., which helps to enhance communication and interaction among community members and increase the activity and cohesiveness of the community (Liu et al., 2022; Shin & Perdue, 2022). First, online reviews can improve interaction and communication among community members. Through online reviews, community members can share their travel experiences and suggestions, provide valuable information and opinions, and provide helpful guidance and reference for other users. This information sharing can promote communication and interaction among community members and increase the activity and cohesiveness of the community (Kim & Kim, 2021; Tran & Rudolf, 2022). Second, online reviews can improve the transparency of information and enhance the quality of the sharing of travel experience. Through online reviews, people can learn about others’ travel experiences and better prepare and plan their own trips. This can enhance the quality of travel experiences and contribute to the development and progress of the tourism industry (Nilashi et al., 2021). Finally, online reviews can contribute to the development and prosperity of communities. Through online reviews, community members can exchange opinions and suggestions, give feedback on the problems and shortcomings of the community, and provide valuable suggestions and support for community improvement and development (Fu & Pan, 2022; González-Rodríguez et al., 2020). This feedback can contribute to the development and prosperity of the community, improving its quality and service level and attracting more users and business opportunities.

Online reviews have a positive impact on social interactions in tourism virtual communities. Through online reviews, users can increase interaction and communication, improve the transparency of information and the quality of the travel experience, and contribute to the development and prosperity of the community. Therefore, the following hypotheses are proposed.

H2:

Online reviews positively influence social interactions in tourism virtual communities.

Online reviews play a crucial role in tourism virtual communities and have a significant impact on information exchange. Online reviews can help users acquire and share knowledge about tourism information, thus enhancing the effectiveness and quality of information exchange (Camilleri & Kozak, 2023; Wong et al., 2022).

First, online reviews can improve the transparency and reliability of the information. Users can share their travel experiences, opinions, and suggestions through online reviews so that other users can better understand the actual situation and user ratings of the destination. Such information sharing can improve the transparency and reliability of information and facilitate more informed decisions by users (Li et al., 2022). Second, online reviews can foster information exchange and share among users. Through online reviews, users can share travel photos, videos, itineraries, and other interesting travel information with each other. Such sharing can help users better understand the actual situation and user reviews of tourism destinations, promote information exchange and sharing among users, and improve the effectiveness and quality of information exchange (Cui et al., 2022; Hu, 2019). Third, online reviews can promote interaction and cooperation among users. Through online reviews, users can exchange opinions and suggestions and establish connections and collaborative relationships with other users to better plan and arrange their trips. This interaction and cooperation can promote trust and cooperation among users and improve the efficiency and effectiveness of information exchange (Jiang et al., 2021; Lee & Park, 2019). Therefore, the following hypothesis is proposed.

H3:

Online reviews positively influence the exchange of information in tourism virtual communities.

3.3. The effects of social interaction on user engagement behavior

Social interaction is the process of establishing and forming social relations through information transmission, thus generating the tendency of connection. With the diversified development of user participation in value creation and social network communication, social interaction also presents multi-dimensional characteristics. Zhao and Detlor (2023) divided social interaction into two dimensions: social exchange and information exchange. In tourism virtual communities, the higher the degree of social interaction between users, the more frequent their social exchange and information exchange will be, which will promote the generation of user sharing or recommendation behavior (Kim et al., 2022; Urbonavicius et al., 2021).

The social exchange of a tourism virtual community can increase user engagement from the perspective of attitude and behavior. Users can gain more information, knowledge, and experience through the virtual social exchange, thus, to better understand the culture, values, and expectations of the community, and accordingly adjust their attitudes and behaviors to adapt to the virtual community (Lee et al., 2018). Social exchange can also enable users to gain more sense of identity and belonging, thus further strengthening the interaction and contribution to the virtual community. From the perspective of behavior, the social exchange of a tourism virtual community will also promote the generation of user behavior (Luo et al., 2019), such as information sharing behavior, communication and cooperation behavior, so that users can better understand and comply with the rules and expectations of the community (Prentice et al., 2019). In addition, social exchange can also help users gradually cultivate behavior patterns that meet the community’s requirements during communication and cooperation, and further enhance the generation of behavior (Xie et al., 2021). Therefore, social exchange is an important means to promote user engagement in the tourism virtual communities. Thus, the following hypothesis is proposed.

The impact of tourism virtual community information exchange on user engagement lies in the following two aspects. On the one hand, information exchange in tourism virtual communities can enhance attitudinal engagement among users. Users can better understand each other’s travel experiences, travel preferences, and travel attitudes through information exchange, thus forming empathy and emotional resonance (Xue et al., 2020). This empathy and emotional resonance can promote attitudinal engagement among users and thus enhance interaction and communication among users. On the other hand, information exchange can facilitate knowledge and experience sharing among users to improve their travel literacy and cognitive level (Xi & Hamari, 2019). This knowledge and experience sharing can promote behavioral engagement between users, allowing them to understand and accept better each other’s travel behaviors and preferences, thus be more willing to engage in everyday travel actions. This behavioral engagement enhances the spirit of cooperation and sense of belonging among users, which in turn promotes social interaction and the development of tourism virtual communities. Therefore, the following hypothesis is proposed.

H4a:

Social exchange in tourism virtual communities positively influences user engagement behavior.

H4b:

Tourism virtual community information exchange positively influences user engagement behavior.

3.4. The effects of user engagement on user recommendation behavior

In tourism virtual communities, user engagement and recommendation behaviors are two important aspects of tourism value co-creation that can help travelers co-create, share and evaluate tourism value in virtual communities (Qu et al., 2022). User engagement is affected by the social media environment, brand community characteristics, user self-motivation, quality of user-generated content, and social interaction in online communities, which is the embodiment of user psychological identity (Luo et al., 2019; Rasool et al., 2020).

User recommendation behavior is a kind of information dissemination behavior, which belongs to the external performance of behavior, which is more affected by the user’s participation degree and motivation. User agreement is the synthesis of user’s cognition, emotion and behavior. It is not only the embodiment of user’s psychological level, but also the trigger of user’s behavior. Rewards, online word-of-mouth, brand reputation, and social network interaction in virtual communities will affect users’ perception, resulting in recommendation behavior (Mishra et al., 2018; Vilnai-Yavetz & Levina, 2018). In the tourism virtual community, users interact with each other by reading travel information, sharing travel videos, and publishing travel reviews, which is also the exchange process of users’ interests, views, and values (Sheng, 2019). This process includes users’ cognitive sharing, emotional communication, and behavioral interaction. After users perceive similar cognition, emotion, and behavior, they will generate forwarding or recommendation behavior. That is, user engagement produces recommendation behavior from cognitive and emotional dimensions.

H5:

User engagement behavior positively influences recommendation behavior in tourism virtual communities.

Based on the above theoretical analysis, this study constructs the following second-order theoretical hypothesis model (see Figure 1 in appendix).

Figure 1. Second-order theoretical model of this study.

4. Research design

4.1. Measurement of variables

This paper adopts the existing research to design the questionnaire. For the purpose of this study, some modifications were made to the abovementioned questionnaires to compile a new questionnaire for this study. The new questionnaire includes four parts: basic information of respondents and the measurement of online review characteristics, online interaction and value co-creation behavior. Online review characteristics include four variables: language style, affective content, review length and review star rating. Online interaction includes two variables: social interaction and information exchange. The value co-creation behavior includes two variables: engagement behavior and recommendation behavior. Specifically, the scales of language style (3 items), affective content (3 items), review length (5 items) and review star rating (3 items) were mainly adapted from the research of Luo et al. (2021), Liu et al. (2019) and Yoon et al. (2019). The scales of information exchange (5 items) and social exchange (4 items) were mainly based on the research of Liu et al. (2022), Rosário and Raimundo (2021) and Scekic et al. (2018). The scales of engagement behavior (3 items) and recommendation behavior (3 items) were taken from the research of Chen et al. (2022), Xu et al. (2020) and Li et al. (2019). These items were measured on a 1–5 Likert scale where 1 represented “strongly disagree” and 5 represented “strongly agree”.

4.2. Data collection and sample distribution

The questionnaire mainly targets people who have travel experiences and loafing experiences on online travel platforms so that the data can genuinely mirror the willingness and honest feelings about the development and management of those online travel platforms. The questionnaires were distributed on the website Sojump.com, the largest questionnaire survey platform in China. The questionnaires started with an introduction to the online travel community. We listed several well-known and popular online review platforms in mainland China, including Ctrip.com and Mafengwo.com. A filter question was set to screen out those who were not members of online review platforms to ensure the questionnaire’s validity. After designing the questionnaire, the research team first conducted a presurvey to modify the measurement items that were not clearly described. The final questionnaire was distributed through the Sojump.com platform from August to September 2022.

Four hundred and fifty-four questionnaires were collected in our survey, and 342 were valid, with an effective return rate of 75.3%. The demographic characteristics results showed that 52.35% were female, and 47.65% were male. The proportion of respondents aged 18 to 30 years was the highest (61.20%), followed by those aged 31 to 40 years (13.24%). The proportion of tertiary students and above is up to 95.33%. A total of 64.5% of the respondents showed that they are willing to use tourism virtual community platforms every time for travel planning. Demographic results show that the sample distribution is mainly composed of young participants, who are acceptors of new technologies. The specific structure of the valid sample is shown in Table 1(see in appendix)

Table 1. Description of the sample structure

Characteristics	Options	Proportion
Gender	Male	47.65%
	female	52.35%
Age	Under 18	3.97%
	18–30	63.20%
	31–40	14.24%
	41–50	12.24%
	More than 50	6.35%
Education	High school or below	4.67%
	Tertiary students	6.38%
	Undergraduates	63.45%
	Master’s and above	25.5%
Using frequency	Never used	3.16%
	Using occasionally	35.47%
	Using every time	61.37%

5. Data analysis and hypothesis testing

5.1. Multivariate normality

Multivariate normality and common method variance were examined prior to testing the measurement model. Skewness and kurtosis were used to assess the normality of the distributions of all items. The results indicated that all absolute skewness values were less than 2 and absolute kurtosis values were less than 3. Thus, a normal distribution requires no substantial deviation (Zhang et al., 2021).

5.2. Common method variance test

Common method bias refers to the artificial covariation between predictor variables and validated variables caused by the same sample data source or rater, the same measurement environment, the item context, and the item’s characteristics. This artificial covariation creates serious confusion about the study results and is potentially misleading for the conclusions, which is a kind of systematic error. Therefore, in the questionnaire design process, the questionnaire was pretested twice to ensure its quality, and the wording of the questionnaire was revised based on the respondents’ feedback. During the questionnaire survey, the method of anonymizing respondent information was used, and data were collected from people who had targeted travel and hanging out experiences on online travel platforms so that the data could truly reflect people’s willingness and true feelings about the development and management of these online travel platforms. Finally, this study uses the potential error variable control method to test for common method bias. In the structural equation model, common method bias was used as a latent variable, and the common method bias effect was tested if the significant fit of the model was better in the case of including the method bias latent variable than in the case of not including it (Luo et al., 2019).

The test results are shown in Table 2 (see in appendix). First, the validating factor analysis model M1 was constructed, followed by the model M2 with the method factor included. Comparing the main fit indices of model M1 and model M2, we obtained △χ2/df = 0.173, △RMSEA = 0.005, △SRMR = 0.002, △CFI =-0.015, and △TLI=-0.017. Thus, it can be seen that the fit indices of M2 are less significant than those of the original model. M1 did not change significantly, the changes in RMSEA and SRMR did not exceed 0.05, and the changes in CFI and TLI did not exceed 0.1. This indicates that the model was not significantly improved by adding the common method factor. Therefore, there is no significant common method bias in the measurements.

Table 2. The result of common method deviation

Model	x^2/df	RMSEA	SRMR	CFI	TLI	△x^2/df	△RMSEA	△SRMR	△CFI	△TLI
M1	2.201	0.059	0.044	0.914	0.900
M2	1.994	0.054	0.042	0.929	0.917	0.173	0.005	0.002	−0.015	−0.017

5.3. Testing the second-order factor structure

5.3.1. Psychometric properties of the first-order factors

We applied a confirmatory factor analysis to data analysis using maximum likelihood estimation. First, we tested the psychometric properties of the scales of the first-order factors of online reviews characteristics. The model displayed good fit indices (χ²/df = 2.846，RMSEA=0.066，SRMR=0.044，GFI=0.927，IFI=0.942，CFI=0.941，NFI=0. 907，TLI=0.925) and was tested further for its reliability and validity. The results were shown in Table 3 (see in appendix). The results of reliability analysis showed that the composite reliability (CR) and average variance extracted (AVE) reached the requirements of 0.7 and 0.5, respectively (Qu et al., 2022). The reliability test results meet the requirements. At the same time, AVE values also tested the convergent validity. However, the model did not pass the test of discrimination validity. Nunkoo et al. (2017) believed that there were highly correlated factors in the second-order factor model, which make up the second-order model. In this case, we only needed to test the convergence validity. Therefore, this study will test the discriminant validity in the overall model.

Table 3. Psychometric properties of the measurement scales for the first-order factors

First order factor and their indicators	SL	t-values	CR	AVE
LST. Language style			0.777	0.538
LST1. I prefer rhetorical words when I read reviews.	0.673	-
LST2. I prefer colloquial words when I read reviews.	0.739	10.381
LST3. I prefer reviews with pictures or emojis when I read reviews.	0.785	9.4151
ACO. Affective content			0.782	0.544
ACO1. I prefer more firmly positive reviews when I read reviews.	0.705	-
ACO2. I firmly believe affective content echo with my ideas.	0.768	9.6714
ACO3. I believe firmly affective expression is helpful in maintaining a sound environment.	0.739	8.9917
LEN. Review Length			0.830	0.619
LEN1. I prefer longer and more detailed reviews.	0.774	-
LEN2. I will choose to read in detail or scan quickly by the review length.	0.759	11.4917
LEN3. I will judge the effectiveness of reviews according to their length.	0.826	9.1383
SRA. Review Star rating			0.880	0.597
SRA1. I prefer high-star rating reviews when I read reviews.	0.837	-
SRA2. I prefer more like give reviews when I read reviews.	0.825	14.059
SRA3. I prefer more interactive reviews when I read reviews.	0.694	10.9249
SRA4. My choice of travel destination will be affected by reviews with high-star rating.	0.735	11.8218
SRA5. My willingness to travel will be influenced by the number of fans and likes of reviewees.	0.762	10.5361

5.3.2. Model comparison

According to testing procedures of the second-order factor model (Nunkoo et al., 2017), we developed four models using the hierarchical method to test the performance of the second-order factors of online review characteristics. The four models were: single first-order factor (M1, Figure 2,see in appendix), four first-order uncorrelated factor (M2, Figure 2, see in appendix), four first-order correlated factor (M3, Figure 2, see in appendix), four first order factors and one second order factor (M4, Figure 2, see in appendix). The confirmatory factor analysis (CFA) of the four models were shown in Table 4 (see in appendix). The results show that M2 have unacceptable model fit indices. The fit indices of M4 were slightly higher than M1 and M3. According to Koufteros et al. (2009), when the second-order factor model had a similar fit index to the first-order factor model, the second-order model was a better choice. Therefore, we adopt M4 as the most appropriate model and test its performance in the overall measurement of the structural model.

Figure 2. a) Single first-order factor (M1) b) single first-order factor (M2) c) first-order correlated factor (M3) d) first-order factors and second-order factor (M4).

Table 4. Model comparison

Fit indices	M1	M2	M3	M4
χ^2/df	2.841	11.861	2.486	2.457
RMSEA	0.074	0.178	0.066	0.065
SRMR	0.047	0.328	0.044	0.043
GFI	0.913	0.696	0.927	0.926
IFI	0.923	0.543	0.942	0.942
CFI	0.922	0.540	0.941	0.941
NFI	0.885	0.521	0.907	0.906
TLI	0.907	0.456	0.925	0.927

5.4. Testing the overall measurement and structural models

5.4.1. Reliability test

The valid data were exported into SPSS 26, and an exploratory factor analysis (EFA) was then conducted. The result indicates high reliability for the data (the reliability is 0.954). In addition, the correlation coefficients among the 29 items of the questionnaire are all above 0.4, which means that every item in the questionnaire is closely related to the others. Then, a confirmatory factor analysis (CFA) was conducted, and validity analysis revealed that the questionnaire had good quality. KMO is 0.963, and the significance is p < 0.000, which tells us that the data are adaptable to structural equation analysis. Therefore, all these data revealed that the questionnaire has sound quality and is suitable for the following study.

In this study, the above factors were measured in the form of scales, so testing the data quality of the measurement results is a prerequisite to ensure that the subsequent analysis is meaningful. First, the internal consistency of each dimension was analyzed by the Cronbach coefficient reliability test method. Among the 342 valid questionnaires, the reliability coefficients of online review characteristics, social interaction, and tourism virtual community value co-creation behavior in general and the secondary dimensions to which they belong were all in the range of 0.6–1. For exploratory factor analysis, Cronbach’s α greater than 0.5 is sufficient for further analysis (Hulland, 1999). Overall, the scale has good reliability.

5.4.2. Validity test

In this study, convergent validity analysis and discriminant validity analysis were used to jointly test the scale validity. Among them, convergent validity was confirmed by standardized factor loadings, composite reliability (CR), and average variance extracted value (AVE) methods. We conducted confirmatory factor analysis on the variables. According to the results shown in Table 5 (see in appendix), each standardized factor loading exceeded the requirement of 0.7. The AVE of each variable largely exceeded the acceptable level of 0.5, while the CR were all greater than 0.7. This indicates that the scale has good convergent validity.

Table 5. Properties of the overall measurement model

Variables and their indicators	SL	t-values	CR	AVE
ORC. Online review characteristics			0.916	0.732
LST. Language style	0.859	-
ACO. Affective content	0.846	9.505
LEN. Review Length	0.889	10.882
SRA. Review Star rating	0.828	11.434
IEX. Information exchange			0.878	0.591
IEX1. I consider that online participant shares useful information in the online travel community.	0.763	-
IEX2. I think participants shared frequently in the online travel community.	0.798	11.9802
IEX3. I think the shared information is informal at large in the online travel community.	0.773	11.7802
IEX4. I believe that online travelers will share more travel-related information in the online travel community.	0.737	11.3473
IEX5. My cognition of a destination will change during the information interaction.	0.772	11.9906
SEX. Social exchange			0.848	0.583
SEX1. I believe online travelers are able to improve their networking in online interaction.	0.771	-
SEX2. I believe online travelers are able to make friends through online interaction.	0.789	11.5282
SEX3. I believe online travelers will enhance their mutual understanding in online interaction.	0.738	10.5951
SEX4. I believe online travelers have opportunities to form common values in online interaction.	0.756	10.9678
ENG. Engagement behavior			0.800	0.572
ENG1. I will recognize the quality of the online travel community through online review and interaction.	0.803	-
ENG2. I will become more enthusiastic after reading online reviews and participating in the interactions.	0.728	12.7787
ENG3. I will provide and search for more information in this online travel community after reading online reviews and participating in interaction.	0.735	12.9793
REC. Recommendation behavior			0.830	0.619
REC1. I will share interesting information from the community with others face-to-face.	0.841	-
REC2. I will share information about the community on social platforms.	0.727	12.7824
REC3. I will recommend others to join in the online travel community.	0.789	13.617

Table 6 (see in appendix) shows the results of the discriminant validity test. In the discriminant validity test, the standardized correlation coefficients between two dimensions are less than the square root of the AVE value corresponding to the dimension, thus indicating that each dimension has good discriminant validity.

Table 6. Discriminant validity test

	LST	ACO	LEN	SRA	IEX	SEX	ENG	REC
LST	0.719
ACO	0.379	0.723
LEN	0.505	0.373	0.776
SRA	0.429	0.494	0.521	0.760
IEX	0.522	0.487	0.314	0.339	0.769
SEX	0.394	0.462	0.387	0.509	0.363	0.764
ENG	0.424	0.389	0.316	0.441	0.347	0.421	0.756
REC	0.411	0.378	0.404	0.328	0.334	0.439	0.346	0.786

The off-diagonal items are correlation coefficients; on-diagonal items are square roots of AVE.

5.4.3. Hypotheses testing

In this study, second-order structural equation modeling was performed on the data using Amos 24, and the results showed that the model had a fine goodness-of-fit. The model fit indices were χ² = 827.7947，df = 368，χ²/df = 2.249，RMSEA = 0.061，SRMR = 0.045，GFI = 0.856，IFI = 0.907，CFI = 0.906，PNFI = 0.765，TLI = 0.896.

Figure 3 (see in appendix) presents the results of the structural equation modeling path analysis. Based on the successful extraction of factors such as language style, emotional content, review length and review rating, the content features and review attribute features of users’ reviews in virtual communities can reflect online reviews. Since they reflect 2 dimensions of review content features and review attribute features, this also supports H1 that there are review content features and review attribute features in their review features in online reviews in tourism virtual communities. Moreover, since the path coefficients of these 4 factors with online reviews were 0.789 (p < 0.001), 0.653 (p < 0.001), 0.705 (p < 0.001), and 0.735 (p < 0.001), which were positive, indicating a significant positive correlation between the 4 factors and online reviews, this further verifies hypothesis H1. By the relationship between online reviews and social interaction (β = 0.655, p < 0.001) and information exchange (β = 0.556, p < 0.001), the path coefficients show that there is a significant positive correlation between online reviews and social interaction, i.e., it indicates that the gradual accumulation of online reviews has a positive effect on users’ social interaction and information exchange in the online reviews of tourism virtual communities.

Figure 3. Standardized path analysis.

***p < 0.001; **p < 0.01; and *p < 0.05

Read the detailed description of this figure

Four sub-dimensions of review can be explained from content and attribute features, and the four factors positively relate to review; Besides, review further have positive relations with social interaction. Moreover, social interaction had a significant effect on engagement, and this one has positive relation with further recommendation behaviors.

Thus, H2 and H3 were confirmed. In addition, social interaction had a significant effect on contracts (β = 0.473, p < 0.001), and hypothesis H4a was confirmed. Information exchange had a significant effect on fit (β = 0.308, p < 0.05), and hypothesis H4b was confirmed. Engagement behavior had a significant effect on recommendation behavior (β = 0.670, p < 0.001), and H5 passed validation.

6. Discussion

First, this paper explores the dimensional characteristics of online reviews and their impact on social interaction. The research found that online reviews have the content characteristics of language style and affective content, as well as the attribute characteristics of review length and review rating. These two dimensions are important characteristics of online reviews. When users browse online reviews, they will be affected by the language style, affective contents, length, and rating of the reviews, which will promote the generation of users’ reading, praising and forwarding reviews in the tourism virtual community. This also shows that users are more willing to interact and exchange information in the virtual community. That is, online reviews have a positive impact on user’ social exchange and information exchange in the tourism virtual community.

Second, this paper explores the impact of social interaction on value co-creation behavior in tourism virtual communities. The study found that social exchange and information exchange in tourism virtual communities positively affect user engagement. In this process, users have a sense of identity and belonging to similar content to achieve psychological engagement. Further, other users are more willing to exchange information on similarities, making users’ psychological and behavioral engagement. That is, social interaction has a positive impact on user engagement.

Finally, this paper also studies the two dimensions of value co-creation behavior in a tourism virtual community and their relationship. The research found that the value co-creation behavior in tourism virtual community is divided into two dimensions: user engagement behavior and user recommendation behavior. However, these two dimensions are not parallel, and the follow a sequence in the generation of value co-creation behavior. In the tourism virtual community, the frequent interaction and efficient information exchange between users lead to user agreement, which is the first level of value co-creation and the embodiment of user psychological identity. Furthermore, users generate forwarding or recommendation behavior through the engagement of cognition, emotion and behavior, which is the external manifestation of value co-creation behavior. Therefore, user engagement behavior positively affects user recommendation behavior.

7. Conclusion

7.1. Findings

As a whole, the study indicates that online review can positively affect people’s willingness to participate in online interactions including social and information exchange. Meanwhile, tourists in online travel communities who are more eager to engage in online interaction are more likely to enhance their future engagement. Furthermore, people’s engagement will exert effects on their emotion and cognition, which further influence their behavior such as recommendation. Therefore, their value can be co-created after the whole process.

7.2. Theoretical contributions

Based on the cognitive evaluation theory and organismic integration theory, this paper studies the effect of online review characteristics and social interaction on value co-creation behavior, which provides some theoretical insights for the value co-creation of tourists in tourism virtual community. First, this study provides a new perspective by introducing online reviews as a higher-order variable in the study of user value co-creation behavior in tourism virtual communities. While previous studies have focused on the impact of online reviews on tourism service quality and user satisfaction, this study focuses more on the impact of online reviews on user value co-creation behavior in tourism virtual communities, which provides a new perspective and research method for the relationship between tourism online reviews and value co-creation. Compared to previous studies, they just focus on the superficial nature of online review instead of finding out its impact from its specific characteristics (Costa et al., 2019), while by constructing a second-order structural equation model, this study can comprehensively evaluate the influence of various features of online reviews on users’ value co-creation behaviors in tourism virtual communities.

Second, instead of discussing online review and social interaction respectively (Fu & Pan, 2022), the study integrates them in different order and shows that online reviews facilitate social interaction. Theoretically, this finding reinforces the importance of social interaction in virtual communities and provides new empirical support for the development of social interaction and social network theory.

Finally, the study shows that social interaction exerts a positive impact on value co-creation behavior in tourism virtual communities. Similar to the theory of reasoned action, which posits that the individuals’ positive attitudes as well as the subjective norms and the influences from society, would have an effect on their intentions and motivations to engage in certain behaviors, this study proved surrounded influence may exert impact on people’ inner value, which further affect their action. This finding validates the mechanism of social interaction on value co-creation behavior from an empirical perspective (Gilal et al., 2022). It provides theoretical support for enterprises and managers to improve tourism service quality and user satisfaction and promote the sustainable development of the tourism industry.

7.3. Practical Implications

From a practical perspective, engaging visitors and motivating them to provide online reviews and interact with other online visitors is a critical step in their value co-creation. The findings of this study provide valuable insights into how co-created value leads to higher levels of engagement, trust, and recommendations among consumers, which is the ultimate marketing goal of any destination. The results of the study prove that online activity can improve destinations’ images. Even if some consumers know only a little about a destination, online reviews and social interactions can lead to more helpful information for various online travelers. Therefore, the content in user engagement and recommendation behavior will lead to different directions of value creation. Value shaping will directly affect their destination image and future travel plans.

Practitioners should aim to maximize the use of high-quality online content while discouraging inappropriate and offensive content, such as the use of vulgar language and extreme vocabulary. In addition, if the most active travelers in the community are rewarded by the platform, people will be motivated to participate in the communication. In addition, travel platforms may benefit if interesting and popular topics online are discussed to engage online travelers.

The findings of the study also had an impact on the development of the destination. Visitors who have visited a destination in the early stages of development can help promote that destination through UGC. Destination management organizations should encourage visitors, especially repeat visitors, to become online ambassadors who actively share their knowledge and experiences of the destination. To achieve this, reward and ranking systems can be used to instill a sense of empowerment and community. In addition, posted online review interactions for destinations may include realistic footage of the scene, so external representation of the destination is of great importance. Multimodal illustrations such as pictures, videos and 3D presentations associated with modern technology may help to impress travelers.

7.4. Limitations and future research

This study has several limitations. First, since our participants were all Chinese, the research results may not be universal for overseas travelers. Future research may test the model with more samples and may also include other variables about how and why people’s value could be co-created in the course of online activities. Second, more specific subfactors of UGC or interaction or other activities should be incorporated into the study so that the test of the model becomes more comprehensive and persuasive. Finally, under the theoretical framework of this study, other variables could be included in the research model to control for the influence of positive or negative shaping of travelers’ value. The paper provides insights into the factors that could trigger various experiences. Nevertheless, this study uses objective data and scientific quantitative research methods to provide support for demonstrating the interconnection between online review characteristics, social interactions, and value co-creation, but leaves food for thought for future related research.

Ethics statement

All projects undergo our university Research Ethics Committee review, the committee decided our study did not need ethics approval.

Acknowledgments

First of all, I would like to give my heartfelt thanks to all the people who have ever helped me in this paper. My sincere and hearty thanks and appreciations go firstly to my supervisor, Mr. Zhu Jianbin, whose suggestions and encouragement have given me much insight into these translation studies. It has been a great privilege and joy to study under his guidance and supervision. Furthermore, it is my honor to benefit from his personality and diligence, which I will treasure my whole life. My gratitude to him knows no bounds. And the paper is supported by the National Social Science Fund of China, the number is 22XKS003. In addition, many thanks go to my family for their unfailing love and unwavering support. Finally, I am really grateful to all those who devote much time to reading this thesis and give me much advice, which will benefit me in my later study.

Disclosure statement

The authors declared that they have no conflicts of interest, we do not have any commercial or associative interest that represents a conflict of interest in connection with the work submitted.

Additional information

Funding

The work was supported by the National Social Science Fund of China [22XKS003].

Notes on contributors

Yingying Gu

Ying Gu Department: Foreign Studies of College; Guilin University of Technology, Research direction: sustainable tourism; cognitive linguistics. Degree: MasterDegree

Jianbin Zhu

Jianbin Zhu Department: Foreign Studies of College; Guilin University of Technology Research direction: applied linguistics; discourse analysis; corpus study.