INFLUENCE OF PERCEIVED SERVICE QUALITY ON BEHAVIORAL INTENTIONS IN PEER-TO-PEER (P2P) ACCOMMODATION

In this study


INTRODUCTION
The need for people to move and find adequate accommodation to rest, get to know new places, research, sports, and excitement is constantly growing, especially after the two-year stoppage of this movement due to the covid-19 pandemic.New technologies and digital innovations have significantly transformed almost all spheres of business, including travel and accommodation.The emergence of the sharing economy has made travel a simpler and cheaper way of connecting people and accommodation, for which it is enough to have a mobile device, the Internet and a suitable platform.In the academic literature, the definition by Botsman and Rogers (2010) is widely accepted as the standard definition of the sharing economy.According to this definition, the sharing economy is characterized as an economic model that facilitates the sharing of underutilized goods or services, either for free or for a fee, directly between individuals.The accommodation sector of sharing economy with its platform Airbnb was a pioneer in creating and using a new model of connecting people who own unused property for accommodation with those who need such accommodation.In academic literature, the term "peer-to-peer economy" is commonly used as a more expansive expression to refer to the sharing economy in accommodation.For that reason, in this study, the term "peer-to-peer economy" will be used to refer to accommodation specifically, while the term "sharing economy" will be used more broadly to refer to all sectors.Airbnb is a peer-to-peer platform for hosting people, founded in San Francisco in 2008.It charges a fee for using the platform to the host and guest (Wanga, Nicolaub, 2017).From the time it was founded until today, Airbnb has recorded exponential growth in the rental of different types of real estate (apartments, houses, country estates, luxury apartments, etc.), accommodation area (city, village, mountain, etc.), price range, experience offered to the user (cooking, adventure, hiking and the like), thanks to a business model based on the sharing economy.This business model has enabled individual needs to take center stage, by facilitating active communication between users and service providers.It has also helped meet the guests' desire for authenticity, transparency, and flexibility, while providing an avenue for evaluating service providers.Moreover, it has encouraged service providers to respect user feedback, and has made communication and payments much more straightforward.Airbnb spread from the USA first to London, then to many European cities, South America, Russia, Australia, China, and Cuba.Although today it closed its operations in Russia and Belarus due to the invasion of Ukraine, and in China due to complicated laws, regulations, discrimination and restrictions due to covid-19, Airbnb has more than 4 million hosts who have accepted over a billion guests in almost all areas of the world.This shows that it is a very significant entity, which builds its success on the satisfaction of a growing number of users.The level of user satisfaction is a crucial variable that determines whether a user will return to a service, how they will rate it, what comments they will leave about the platform, accommodation, host, and overall atmosphere.Additionally, their comments could potentially attract other users to the service.Therefore, this paper will focus on researching perceived service quality's impact on users' behavioral intentions in peer-to-peer accommodation in the Republic of Croatia.Additionally, user engagement will be investigated as a mediating variable between service quality and behavioral intentions, as it is becoming increasingly clear that user engagement is directly related to service profitability.Therefore, it is possible to pose the following research problem: Does the perceived service quality influence the behavioral intentions of users and is this influence mediated by user engagement?

Perceived service quality
The examination of the studies emphasized the categorization of service quality determinants into various categories.Companies must grasp and fulfill customer values from a customer value creation viewpoint to attain a competitive edge in the market (Woodruff, 1997).Providing customer value has been the primary objective of businesses for many years, with the aim of delivering more value than their competitors.In recent times, service quality has been divided into various typologies from different perspectives to assess a product or service's perceived benefits.For example, Sweeney and Soutar (2001) divided the determinants of service quality into four categories: functional, emotional, economic, and social, while Sheth and Parvatlyar (1995) identified five categories of customer value in the context of service quality: functional, social, emotional, epistemic, and conditional.
The sharing economy has unique characteristics that differentiate it from other service environments, such as the absence of ownership and temporary access to or redistribution of material goods or assets such as space or time (Kathan, Matzler & Veider, 2016).As a result, the factors that contribute to the success of the sharing economy in terms of customer value may vary from other services.In the sharing economy, the widespread use of digital technologies has altered the business model, making customer perceived value even more critical in determining satisfaction and repeat usage intentions.Tussyadiah and Zach (2017) studied the quality of service in peer-to-peer (P2P) accommodation and confirmed that P2P accommodations are favored by customers who seek social and experiential incentives.The findings of the study by Liang, Schuckert, Law and Chen (2017) indicate that guests' experiences in P2P accommodations are significantly impacted by various dimensions of the experience, such as aesthetic/sense, relate/social interaction, escape, act and feel.Moreover, P2P experiences lead to the intention of guests to return to the same destination and accommodation.Zhang, Gu and Jahromi (2019) used four categories of service quality in the sharing economy: economic, social, emotional, and technical.Shao, Guo, and Ge (2019) studied the impact of the four-category mechanism on satisfaction and continued intention to use the bike-sharing system.The authors stated that the new bike-sharing business model can provide customers with a more efficient, enjoyable, and affordable mode of transportation, significantly affecting the perceived functional, emotional, and economic values.Additionally, the bikesharing service can also benefit the environment by reducing pollution and promoting sustainable development.The study used functional, emotional, economic, and ecological categories of service quality, but did not include a social component since the study focused on bike-sharing systems, where service quality does not include a social aspect.The social value lies in fulfilling customers' needs for real social interaction in a virtual environment (Zhang, Yan & Zhao, 2016), making it more appropriate to use this category for peer-to-peer accommodation.According to Shao et al. (2019), functional and economic value significantly impact satisfaction and the intention to reuse, as part of the perceived value for service users.Based on the literature analyzed, the determinants of service quality addressed in this research will be divided into three main categories: technical, economic, and social.

Technical category
The sharing economy has gained considerable attention in recent years, as it offers a flexible and convenient way of consumption for users.A growing body of research has focused on the role of platforms and technical aspects in shaping the success of sharing economy services.For instance, several studies (Zhang et al., 2019) have highlighted the importance of the application through which users book services.Specifically, users value technical aspects such as platform reliability, ease of use, response time, and appearance.Huarng and Yu (2019) argue that an effective application is crucial for achieving service quality on online platforms, which is the main factor for achieving repeat usage intention.Users seek convenience, flexibility in reservation, ease of use, and quick problem-solving features.Additionally, studies have identified other important determinants of service quality, such as trust in the platform, user support, and data privacy protection (Priporas et al., 2017, Clauss et al., 2019, Shao et al., 2019).
Moreover, tangible aspects of service, such as cleanliness and comfort, play a key role in determining the quality of accommodation service (An et al., 2019;Zhang et al., 2019).
This study aims to consolidate the technical aspects of sharing economy services into two categories: the application and the physical environment of service provision.The first category focuses on the importance of simple and efficient applications that quickly resolve users' requests.The second category highlights the significance of the physical environment, including the cleanliness and maintenance of the accommodation.By investigating the variables that proved to be significant in previous research, this study seeks to provide a more comprehensive understanding of the technical dimensions in peer-to-peer accommodation.

Economic category
The economic aspect has been identified as one of the primary drivers of participation in the sharing economy, according to numerous studies by both scientists and practitioners.The sharing economy is popular due to its ability to transform the possession of objects into sharing ownership, leading to significant economic benefits such as cost savings, expanded choice of products and services, more favorable prices, and better quality service.The monetary benefit is a prominent feature of the sharing economy, making economic profitability a key determinant of its use (Zhang et al., 2019).The price is the main factor affecting the use of sharing economy services, making sharing a more favorable option than owning (Lamberton & Rose, 2012).Based on two studies with users of Car2Go (transportation) and Airbnb (accommodation), Möhlmann (2015) concludes that cost savings positively influence satisfaction and intention to reuse the sharing economy service in the future.This is particularly evident in the accommodation sector, where users emphasize the economic benefits of sharing economy services such as Airbnb, resulting in positive experiences and a desire to participate again.Users who book accommodation at a better price of the same value experience their trip with more satisfaction, leading to the likelihood of using those services again.The advantage of price information in advance is highlighted in the literature as a significant factor affecting user experience, as it eliminates uncertainty around final costs and reduces the risk of abandonment of service usage.Sharing economy services have the added advantage of displaying the cost of the service immediately upon contracting, enhancing the financial benefit and contributing to a positive user experience and intention to reuse.In conclusion, cost savings and the possibility of price information in advance are identified as the main motives for participation in the sharing economy, with visible impacts on satisfaction and intention to reuse the service.Sharing economy services offer significant economic benefits, expanding the choice of products and services while maintaining favorable prices and quality service.

Social category
The social aspect of service quality refers to the extent to which the product or service enhances the user's social image by meeting the expectations of others (Yang & Jolly, 2006).According to Zhang et al. (2019), forming social connections or finding like-minded individuals is another common factor in the sharing economy.Acquiring social capital is highly valued as people are naturally social beings.The sharing economy has been found to provide various forms of social value, including friendship with other users or service providers and high trust through recommendations.The sharing economy can offer social value in two environments: online platforms and physical environments.Online sharing economy platforms are a form of social commerce that serve as a tool for interaction, encouraging users to continue using the sharing economy (Hamari, Sjöklint & Ukkonen, 2016).For example, Airbnb provides an opportunity for users to interact with different people, including hosts and other guests (Zekanovic-Korona and Grzunov, 2014).Hosts can share food or rides and serve as local guides, while guests can make friends and explore the destination together.A study by Clauss, Harengel, and Hock (2019) emphasized the significance of social aspects in the sharing economy that differentiate it from traditional service environments.This includes interactions with the service provider, access to other users' experiences and opinions, and a preference for trendy and modern lifestyles.Trust in the platform and service provider also has a major impact on customer loyalty.In addition to social value, the emotional aspect of service quality is also frequently mentioned in the literature.The emotional aspect refers to the customers' affective state, which is usually a result of positive feelings acquired through their experience with the product or service (Yang & Jolly, 2006).According to Zhang et al. (2019), the emotional aspect of the sharing economy is based on the idea that it provides an enjoyable and exciting alternative to traditional services.For example, studies on peer-to-peer accommodations have shown that guests feel happy and surprised by the warm welcomes they receive from hosts, while Uber riders engage in pleasant conversation during their rides.Respondents have also reported that staying in Airbnb accommodations feels more like staying in their own homes than in hotels, which exceeds their expectations.According to Yannopoulou, Moufahim and Bian ( 2013), both hosts and guests experience excitement and pleasure through friendship and hosting in the sharing economy, creating emotional value for both parties.

User engagement
User engagement is a critical aspect in the service industry, as it directly impacts the profitability of service providers.Service literature defines user engagement as the level of involvement and interaction between the user and the service, to enhance the user experience and create a positive relationship between the user and the service provider (Brodie, Hollebeek & Smith, 2011;Vivek, Beatty & Morgan, 2012;Enginkaya & Emel, 2014).Measuring user engagement in service literature can be done through various methods, such as surveys, questionnaires, and user behavior tracking.Surveys are commonly used to gather self-reported data on the level of user engagement, while questionnaires can be used to measure the level of involvement, satisfaction, and loyalty.User behavior tracking can also provide valuable insights into user engagement by analyzing user interactions with the service, such as the frequency of use, the duration of use, and the level of satisfaction with the service.
User engagement has been widely studied in service literature and has been found to impact service outcomes.According to the service-dominant logic framework proposed by Vargo and Lusch (2004), user engagement is crucial in determining the outcomes of service interactions.Schaufeli and Bakker (2004) define engagement as an emotional and mental state that encompasses the entirety of a person's experience, not just specific events or actions.Chandler and Lusch (2015) theorize that three key elements -value propositions, engagement as the alignment of connections and attitudes, and service experience as multi-party engagement -are linked and play a vital role in shaping user engagement.These studies emphasize the need for service providers to focus on enhancing and evaluating user engagement to deliver value to their clients.Also, customer satisfaction has long been studied as a concept that mediates between service quality and customer behavioral intentions (Dabholkar, Shepherd, & Thorpe, 2000;Asmi, Zhou, He & Han 2016;Shao et al., 2019).However, scholars and practitioners have realized that more than satisfaction is needed to make customers/users of services loyal and companies profitable (Hollebeek, Glynn & Brodie, 2014;Dessart, Veloutsou & Morgan-Thomas, 2016;Pansari & Kumar, 2017).Profitable loyalty and satisfaction should be developed to a higher level.Therefore, the goal of organizations has evolved from relationship marketing to encouraging customer engagement in all possible ways, which has led to an increase in the meaning of the concept of engagement.All forms of contributing customer value are placed in the conceptualization of engagement (Kumar, Aksoy, Donkers, Venkatesan, Wiesel, & Tillmanns, 2010).Pansari and Kumar (2017) identified components of engagement in terms of direct and indirect customer contributions.Numerous studies have provided empirical evidence that service quality affects customer attitudes and behavior, ultimately leading to business profitability (Prentice, 2013).These attitudes and behavioral outcomes reflect customer engagement.

Behavioral intentions
Research into users' behavioral intentions in sharing economy services is most prevalent in the area of accommodation.In doing so, the studies are divided into those that investigate intentions towards the service provider and others that deal with the issue of behavioral intentions towards a specific platform (most often Airbnb).Previous research on the determinants of behavioral intentions in the sharing economy has shown that positive emotions and experiences during use influence intentions to use the platform again (Cossío-Silva, Revilla-Camacho, Vega-Vázquez & Palacios-Florencio, 2016; Chen & Wang, 2016).Studies have predominantly been carried out in the domain of accommodation services, with a few exploring its impact in the transportation sector.Users are motivated by a series of perceived values towards the decision to use the service again.At the same time, reduced information search costs significantly impact the platform's repeat use (Wu, Chen, Chen, & Cheng, 2014).Laličić and Weismayer (2018) examined behavioral intentions in the sharing economy, using a questionnaire with statements about behavioral intentions in terms of recommendations and reuse.Answers are based on a five-point Likert scale (from strongly disagree to agree with the statement strongly).In contrast to the mentioned study, Priporas, Stylos, Rahimi, and Vedanthachari (2017) used a three-level scale proposed by Salanova, Agut, and Peiró (2005).Shuqair, Pinto and Mattila (2019) examine behavioral intentions also with a questionnaire.Participants rated their intentions using four statements on a nine-point scale, adapted from DeWitt, Nguyen and Marshall (2008).

RESEARCH MODEL
The main research goal is to examine the influence of service users' perceived quality on engagement and consequently on users' behavioral intentions in peer-to-peer accommodation.Three main variables included in this research are perceived service quality, user engagement and behavioral intentions, while the impact of perceived service quality will be observed through three aspects (technical, economic and social).Correlation between selected variables will be established with research hypotheses for the purpose of determining the research model.This research will cover various indicators for measuring the technical category of service quality that has proven to be relevant to the study.Technical indicators will include determinants: response time, trust in the platform, ease of use, functionality, the appearance of the application and accommodation condition.The economic component of the service quality of this research will include financial benefits in the form of more favorable prices compared to alternative forms of accommodation and the possibility of informing about the price in advance.In this research, the emotional category will be omitted in order to obtain clarity of results.Namely, detailed insight into the literature revealed the overlapping of service quality determinants in the social and emotional categories, according to different authors.In order to avoid the overlap of determinants in different categories, the social category will also include determinants that, according to some authors, are classified in the emotional category.Therefore, the social category of this research will deal with the variables of social values in the form of social interactions between users and service providers and the importance of insight into the comments and impressions of other usersIn this research, the social category will encompass factors that influence social interactions, inclinations towards trends, trust in the service provider, and the provider's kindness.User engagement will be measured through two dimensions.The first dimension will measure the emotional connection and feelings of the user when using the service.This includes the user's perception of pleasant fulfillment of the service, happiness, and satisfaction.Users evaluate their satisfaction by comparing their expectations with the perceived performance of the product or service.This dimension also includes the user's connection to the service, in a way that the user feels like part of the community (in sharing economy services, part of a large online community of users) and enjoys talking about their user experience.The second dimension of user engagement refers to feedback from customers.Customers often provide suggestions and act as consultants for a product or service they are using.Their feedback helps the organization recognize what is good and bad from the customers' perspective.With their advice, the organization can improve the existing product/service or contribute to the development of new ones (Hollebeek et al., 2014;Pansari & Kumar, 2017).In this research, user attitudes focused on overall satisfaction with the service and the fulfillment of expectations when using the service, emotional connections and connections of users, and the sharing of suggestions and feedback will be used as measures of user engagement.According to the analyzed literature, behavioral intentions in this research will be measured with the help of the user's attitude regarding recommendations and repeated use.

RESEARCH METHODOLOGY
The main research was conducted using a questionnaire, as a primary data source.The data collected by the questionnaire were analyzed using structural equation modeling (SEM), while data processing was performed using the statistical software SPSS.The processed data was analyzed and interpreted for the purpose of obtaining information necessary for the final consideration of the study and conclusions of the conducted research.The main purpose of the model is to explain the influence of perceived service quality on behavioral intentions through user engagement.The conceptual model includes three dimensions of service quality (technical, economic and social).Hypotheses are shown by arrows and indicate influence of service quality categories on user behavioral intentions through the mediator variable engagement (Figure 1).

Source: Authors' research
The proposed model implies that: 1) The technical category of service quality has a direct and positive influence on the behavioral intentions of users in peer-to-peer accommodation.
2) The economic category of service quality has a direct and positive influence on the behavioral intentions of users in peer-to-peer accommodation.
3) The social category of service quality has a direct and positive influence on the behavioral intentions of users of users in peer-to-peer accommodation.4) Engagement of service users is a mediator between categories of service quality and behavioral intentions of users in peer-to-peer accommodation.

Data collection
The chosen instrument for this empirical research is a questionnaire.Suitable measures were identified through a systematic literature review, and the instrument was used to measure service quality constructs (technical, economic, and social categories), user engagement, and behavioral intentions.The questionnaire was developed in two steps: creating the initial questionnaire and conducting a pilot research.During the first step, the initial questionnaire was created by harmonizing questionnaire items with the existing literature.The authors developed the questionnaire items to better adapt to the empirical research context.The questionnaire items were selected after a detailed literature review of a similar context.The authors adapted the selected items to be in accordance with the research context.During the second step, a pilot study was conducted to check the reliability and clarity of the questionnaire items.The pilot study aimed to test the reliability of the instrument and determine if the items measure a specific construct, considering they were taken from another service environment, in the absence of research in the sharing economy area.Another objective of the pilot study was to evaluate the clarity of the questions.Based on the pilot study results, changes were made to the final version of the questionnaire.The pilot study revealed that some statements were very similar to others and represented excess, so they were removed from the final version of the instrument.Additionally, some claims were found to not achieve the expected results, and their retention was unnecessary.Certain claims were also changed for several reasons.First, some claims were imprecisely formulated, and there was a possibility of their misinterpretation.Second, certain claims should have been better adapted to the context of the research, and their different wording would have more adequately clarified the intended indicator.Third, some statements only needed to be reformulated in order to adapt to the proposed ordinal scale through which respondents expressed their (dis)agreement with certain statements.The results of the analysis led to changes being made to the initial version of the instrument.This involved changing some questions and removing others.As a result of these changes, the final version of the questionnaire contained verified items that were adapted to the research context.The initial part of the final version of the questionnaire includes items relating to the demographic characteristics of the respondents.The main part of the questionnaire measures the perceptions of the respondents.For the purpose of measuring the constructs, an ordinal scale (Likert scale) was used, where respondents express their agreement/disagreement with the proposed statements in an interval of one to five: 1) never, 2) rarely, 3) sometimes, 4) often, 5) always.Each particle had equal importance for the selected variables.The study investigated users of the Airbnb platform in the Republic of Croatia.To ensure data generalizability, simple random sampling was employed, whereby each individual had an equal chance of being selected (Gelo et al., 2008).Airbnb users were selected as the research subject due to the platform's established market presence, widespread use and continuous growth.The study utilized a questionnaire administered over a three-month period, from October 2021 to January 2022.The study was originally intended to have a sample size of approximately 500 respondents, but the final sample size ended up being 327 respondents.

Structural equation model
In order to analyze the proposed model, structural equation model (SEM) is implemented.Statistical package SPSS Amos 23.0 is used to examine measurement and structural model, and the results are presented below.

Measurement model
Modeling with structural equations is usually carried out in two steps, through measurement and structural model.The purpose of a measurement model is to specify an adequate model with good suitability of indicators.After a satisfactory measurement model has defined indicators, the analysis of the structural model is examined in detail parameter estimations for the purpose of making research conclusions.In the following, a detailed analysis of the research measurement model is conducted.The measurement model examines the connections between manifest and latent variables, analyzing how good manifest variables represent latent variables.Manifest variables are measurable indicators of latent variables, which in this research refers to empirical data obtained through a questionnaire.The analysis of the measurement model is a confirmatory factor analysis (CFA).In order to examine the appropriateness of the measurement model, it is necessary to test the adjustment of the measurement model as a whole and the validity of individual latent constructs.Table 2 shows the convergent validity and reliability analysis.Factor loading values determine the correlation between factors and constructs, and they should be relatively high.Standardized factor loadings proposed by the relevant authors range between 0.5 and 0.7, and it would be ideal if the values were higher than 0.7.In this way, unidimensionality is also tested, which ensures that each indicator measures only one construct and that error components are mutually independent.Average variance extracted (AVE) measures the level at which the variance of the indicator is explained by the latent construct, or how well the indicator measures the construct.The value of AVE should be around 0.5 and higher (Hair, Ringle & Sarstedt, 2010;Zait & Bertea, 2011).Reliability is also an important indicator of validity.It is often tested by Cronbach's alpha coefficient, which shows to what extent the variables are consistent in their values.0.7 or more flexible 0.6 are usually taken as the lower limits of acceptable values, whereas a more flexible boundary is usually used for exploratory research (Tarhini, El-Masri, Ali & Serrano, 2016).As it can be noticed from Table 2, all standardized factor loadings have values above 0.5, showing that latent variables are well reflected by the indicators.Also, since all AVE values are higher than 0.5, it means that there is a strong convergent validity of the constructs.In the end, all Cronbach's alpha coefficients are above 0.7, indicating an internal consistency of the items.

Source: Authors' research
Discriminant validity results are presented in Table 3, where each value shows the correlation between latent variables.The values on the diagonal are square roots of each AVE, and since values for each construct are higher than correlation with other constructs, discriminant validity is supported.

Structural model
The structural model is derived from the measurement model and emphasizes the examination of cause and effect relationships between constructs.Structural model analysis implies testing the overall adaptation of the model and its direction, intensity and significance of parameters that show the connections between latent constructs.
Latent constructs in this model are: TCQ, ECQ, SCQ, UET and BLI.Every latent construct is measured by several indicators (Table 2), that are represented by statements in the questionnaire.Likert scale was used to measure the constructs, where respondents expressed their agreement/disagreement with the proposed statements in an interval from one to five: 1) never, 2) rarely, 3) sometimes, 4) often, 5) always.Model fit statistics is presented in Table 4 and it can be noticed that this model has an acceptable fit.Also, the p-values shown in the last column of the table indicate a positive and statistically significant influence of service quality categories on behavioral intentions, with a level of significance 5%.It means that research hypotheses 1 -3 are confirmed.
Furthermore, it is necessary to test the mediation of user engagement, which means that engagement intervenes between service quality categories and users' behavioral intentions.In order to reach a conclusion about the existence of mediation, estimates between constructs with included mediation should be checked, as shown in Table 4.It can be noted that all impacts are statistically significant.In the next step, comparing the total effect with the direct effect without mediating influence is necessary.In a model without mediator, standardized estimate between TCQ and BLI was 0.018 between ECQ and BLI 0.026 and between SCQ and BLI 0.015.In order to calculate the total effect of service quality categories on the user's behavioral intentions, it is necessary to summarize the direct and indirect effect.The indirect effect is obtained by multiplying the impact of certain quality categories on the mediator and the mediator's influence on the users' behavioral intention.This means that the indirect influence of the technical category on the behavioral intentions is obtained by multiplying the influence of technical category on engagement (0.24) and engagement on intentions behavior (0.43), which is 0.1032.In the case of the economic category, indirect impact is 0.1419, which is obtained by multiplying the influence of the economic category on engagement (0.33) and engagement on behavioral intentions (0.43).At the last, in the case of the social category, the indirect influence is 0.1075, and it is obtained by multiplying the impact of social category on engagement (0.25) and engagement on behavioral intentions (0.43).Ultimately, in order to obtain the overall effect, it is necessary to sum up the direct and indirect influence.In case of technical category, the total effect on behavioral intentions with included mediation amounts 0.121, for the economic category this effect is 0.168, while for the social category it is 0.122.In the last step it is neccesary to compare the obtained total effects with direct effects without mediation influence.
In a previously tested model without mediation, direct effect of technical category on behavioral intention was 0.02, economic category was 0.03 and social category was 0.019.Given that all direct effects (without mediation impact) are closer to zero than the total effect, and also all parameters are statistically significant, it is proven that variable user engagement has a mediator role.It means that research hypothesis 4 is also confirmed.

Source: Authors' research
The final structural model and its path diagram is presented in Figure 2. The structural model is derived from the measurement model and demonstrates cause and effect relationships between constructs.Analyzing the structural model implies testing the model's overall fit and the direction, intensity and significance of the parameters that show the connections between the latent constructs.As already discussed, all paths are statistically significant showing that the research hypotheses can be accepted.While the studies on the sharing economy did not cover all the dimensions and categories of service quality examined in the present study, some comparisons can still be made.The conclusions drawn from this research are consistent with other studies in peer-to-peer accommodation.Prior research has indicated that certain dimensions of service quality (such as ease and simplicity of application usage, customer support, data privacy, trust, response time) can enhance the user experience, leading to positive outcomes such as recommendations and repeat usage (Priporas et al., 2017;So, Oh & Min, 2018).Specifically, the social aspect of accommodation has been identified as a key factor in choosing a sharing economy platform over traditional accommodations (Guttentag, 2015;McArthur, 2015;Tussyadiah, 2016;Ert, Fleischer & Magen, 2016).It seems that the role of the host is critical in providing a comfortable and safe environment for guests, as hosts are responsible for creating opportunities for authentic experiences (Guttentag & Smith, 2017;Ariffin, 2013).While research on the sharing economy has primarily focused on the accommodation sector, studies in the transport sector also support the findings of this research (Asmi et al., 2016;Podrug, 2022).

CONCLUSIONS
The conducted research on the users' attitudes of the Airbnb service in Croatia showed that the conceptual research model was well established and that all hypotheses derived from the conceptual model were confirmed.
It was confirmed that the technical category of service quality directly and positively influences the behavioral intentions of users in peer-to-peer accommodation (H1).Technical indicators included response time, trust in the platform, ease of use, functionality, the appearance of the application and accommodation conditions.The research showed that users think application is regularly updated and informs them about the latest possibilities.Ensuring the reliability of an application is crucial for users, as it enhances their trust in the software and the security of their devices.Maintaining the correctness of the application is particularly important, as it enables users to carry out their tasks safely and with confidence.In today's era, where various forms of internet abuse exist, users tend to become skeptical, cautious and reluctant to disclose their personal information or carry out online transactions.Therefore, the trustworthiness of an application becomes primary.Users also value applications that are user-friendly, welldesigned and easily comprehensible, requiring minimal external help.This aspect is important, as it instills confidence and a sense of security in users, leading to recommendations and wider adoption of the application.
Another aspect related to the accommodation options provided through the application falls within the technical category of perceived service quality.Respondents emphasized the significance of cleanliness, orderliness, and maintenance of the accommodations.This highlights that users of accommodation services do not merely prioritize affordability while making their choices.Instead, they seek out pleasant and comfortable accommodations that meet their expectations.
The economic dimension of service quality focused on the financial benefits of using the accommodation service, such as more favorable prices compared to alternative options and the ability to access price information beforehand.The study found that the economic category of service quality had a direct and positive impact on the behavioral intentions of users in peer-to-peer accommodation (as confirmed by H2).This implies that users of this service place significant value on the affordability of accommodations, which helps them to stay within their budgets and enjoy a unique experience without having to compromise.Additionally, the availability of upfront pricing information allows users to plan their trips and expenses effectively by matching their preferences and requirements.Consequently, participating in peer-to-peer accommodation provides users with value for their money, encouraging them to continue using the application.The third category used to evaluate service quality was social, and it was found to have a direct and positive impact on the behavioral intentions of users in peer-to-peer accommodation (as confirmed by H3).This category examined the social interactions between users and service providers, as well as the importance of feedback from other users.It was found that users of the application seek insights into the comments and suggestions of other users, as it enables them to make informed decisions regarding the service.Furthermore, social interaction is a crucial aspect of the service, allowing users to meet local residents, make new friends, gain important information, and feel safe and secure in the accommodations they have rented.The friendliness of the host was also found to be a significant factor for users, despite the fact that many accommodations do not require direct interaction between the host and the user.Small gestures of kindness, conversation, advice, and attentiveness were highly valued by users.Since the use of accommodation sharing platforms is still a relatively new trend within the sharing economy, its acceptance and development have been positively influenced by all stakeholders involved.Therefore, it is not surprising that this trend continues to evolve.Additionally, users of these applications feel up-to-date and informed, which contributes to their overall satisfaction with the service.
According to the conceptual research model, shown in Figure 1, users' engagement is a variable that was a mediator between the technical, economic and social dimensions of service quality and the user's behavioral intentions in peer-topeer accommodation.The research confirmed that service user engagement is a mediator between service quality categories and sharing economy users' behavioral intentions (H4).This category examined two dimensions of user engagement: their emotional connection and feelings when using the service, and making suggestions for improvement.As for the first category, it turned out that users like using this service, because on the one hand, it's fun for them, and on the other hand they feel like members of a large online community.This gives them a security and identity with a certain group of people.They discuss their users' experience with others, share their experiences, and thereby unconsciously influence the creation of the service.The second dimension was related to giving feedback, which can also influence the improvement of the service.Users are happy to express ideas for improving the application, but also happy to share feedback.It is certain that the mentioned users' engagement actions influence the user's behavioral intentions.Users express the need to recommend this service, they plan to continue using it and at the same time, they want to use this service rather than some other forms of accommodation rental.All of the above shows that the perceived service quality impacts the user's behavioral intentions and that this impact is mediated by the users' engagement.
When interpreting the results of this research, it is important to keep in mind its limitations.One such limitation is that the study exclusively focused on the Republic of Croatia, which may limit the generalizability of its results to other geographical regions.Another limitation of this study is that it measures users' intentions to use the service again or recommend it to others, rather than their actual behavior.This may not fully capture their level of engagement with the service, as intentions do not always translate into action.To guide future research, several recommendations can be considered.Firstly, while there is extensive research in peer-to-peer accommodation, there is a lack of research on other sectors of the sharing economy, so researchers should be motivated to explore the transport sector and other areas, such as performing housework on demand.In addition to examining other sectors, future research should aim to compare and analyze the differences between various sectors of the sharing economy to determine if there are any variations in the impact of different categories or dimensions of service on specific sectors.Finally, future studies should also consider the service provider's perspective, which is less frequently examined in all areas of the sharing economy.By doing so, a more comprehensive understanding of the service could be gained, leading to higher quality services and increased provider satisfaction.

Figure 2
Figure 2 Structural model

Table 1
provides information about respondents' sociodemographic profiles.A higher percentage of women participated (60.55 %) in this research, as compared to men (39.45 %).Also, the Table presents the structure of respondents by age, showing that most of the participants are between 26 and 35 years old.Regarding the level of education, in this study the highest number of participants have bachelor degree (44.65 %), following by masters (40.06 %) and 10.70 % have high school education.Only 15 % of respondents are PhD.Furthermore, most respondents are employed or self-employed (70.64 %), while the smallest group are housewives or pensioners (3.36 %).The presented Table also shows that respondents are mostly from 2 biggest Croatian cities Split, (37.62 %) and Zagreb (33.64 %).
Source: Authors' research

Table 2
Standardized factor loadings, AVE and Cronbach's alpha values

Table 4
Structural model fit statistics