Do Gender and Prior Experience Moderate the Factors Influencing Attitude toward Using Social Media for Festival Attendance ?

Male and female consumers differ in terms of decision-making. This is also true for inexperienced and experienced consumers. Although the extant research delineates studies that have centered on the interrelationships between perceived ease of use (PEU), perceived enjoyment (PE), perceived usefulness (PU), attitude toward using social media (ATUSM), and users’ behavioral intentions, there is still a dearth of empirical research about gender and prior experience as the two potential moderators. With this recognition, our study examines gender and prior experience as moderators of the effects of PEU, PE, and PU on ATUSM. Our study also investigates ATUSM as a full mediator of the impacts of PEU, PE, and PU on users’ intentions to attend festivals (IAF). Data came from the students of the Korean Government Scholarship Program. The results from structural equation modeling reveal that the positive influence of PE on ATUSM is stronger among female users. As expected, the positive impact of PU on ATUSM is stronger for male users. As hypothesized, prior experience strengthens the positive effects of PE and PU on ATUSM. The results further reveal that ATUSM fully mediates the effects of PE and PU on IAF. Based on the aforesaid findings, theoretical implications are discussed and managerial implications to enhance business sustainability are offered.


Introduction
Today's astute consumers take advantage of social media, where they exchange ideas, convey their experience about a product and/or a company to other consumers, and have a strong influence on electronic word of mouth.As one of the most powerful social networking sites, Facebook had 1.28 billion active users on average for March 2017 [1].Twitter and Sina Weibo (widely used by Chinese consumers) are also among the most powerful online communication platforms [2].Both TripAdvisor and Airbnb have become important mechanisms that enable consumers to leave their online reviews, share their ideas, and seek information generated by other consumers, cf.[3][4][5].This is not surprising, because recent research reveals that both argument quality and source credibility engender information seeking and entertainment motives [6].Recent research also indicates that customer value creation links customer engagement to customer stickiness [7].These changes or advances in the utilization of social media have strengthened the bargaining power of consumers [8].According to Schuckert et al. [9], 71% of independent travel-related bookings are carried out online and 36% of all package tour bookings are done online.In short, consumers utilize social media as an important platform to make comparisons and leave their reviews online, seek credible information about a product and/or a company, and choose the product based on what they have found and read in different social networking sites, cf.[9][10][11].
Social media also helps companies and/or organizers promote their events/festivals.Consumers with pleasant social media experiences have positive attitudes toward utilizing social media and therefore are more likely to attend festivals [10].When there are festivals, the management of companies can use web link messages and create pages in the relevant social networking sites to show the attractiveness of these events, cf.[12,13].However, consumers should be able to trust these social networking sites, find them useful when seeking information about festivals, perceive them to be easy to use, and entertain themselves while searching the information needed [10,14,15].This is important because research suggests that there are discrepancies in the representation of companies (e.g., hotels) in terms of various criteria such as language characteristics, ratings, and usefulness [16].In short, managers' full understanding of consumers' perceived ease of use (PEU), perceived enjoyment (PE), perceived usefulness (PU), attitude toward using social media (ATUSM), and (continuance) intention to use social media can promote business sustainability, cf.[17].
There is evidence about the interrelationships between PEU, PE, PU, ATUSM, and intentions to attend festivals (IAF)/intentions to use social media [10,15].However, what is missing in the relevant literature is related to whether the effects of PEU, PE, and PU on ATUSM are moderated by gender and prior experience [18,19].

Purpose
Against the above backdrop, our study investigates gender and prior experience as moderators of the impacts of PEU, PE, and PU on ATUSM for festival attendance.Our study also examines ATUSM as a full mediator of the impacts of PEU, PE, and PU on IAF.Specifically, the objectives of our study are to address the following questions: (1) What is the nature of the relationship between PEU, PE, and PU and ATUSM?(2) What is the nature of the relationship between ATUSM and IAF? (3) Does ATUSM act as a full mediator of the impacts of PEU, PE, and PU on IAF? (4) Does gender function as a moderator of the effects of PEU, PE, and PU on ATUSM?(5) Does prior experience moderate the influences of PEU, PE, and PU on ATUSM?

Contribution of the Empirical Investigation to Current Knowledge
Examination of the aforesaid relationships contributes to the research on social media and business sustainability and expands current understanding in the following ways.First, almost a decade ago, Sun and Zhang [19] underscored the need for gender as a moderator in information technology behavior research.The current literature also presented several empirical studies which considered gender and experience as the two potential moderators [20,21].However, no empirical study has gauged gender differences in relationships between PEU, PE, PU, and ATUSM in the context of hospitality and tourism so far.This is surprising, because women and men differ in terms of decision-making [22,23].As stated by Karatepe et al. [22] (p.1088), "This distinction between the two sexes impacts how each gender observes the environment, processes, evaluates and retrieves information, and makes judgments".Therefore, investigating gender differences is relevant and important.Without sound findings, management of companies is likely to make poor decisions.Failing to consider gender differences may give rise to an accumulation of inappropriate decisions if there are gender differences.The absence of differences between female and male users may also engender an accumulation of inappropriate decisions and problematic situations if management adopts a gender-sensitive approach, cf.[24].Under these circumstances, management is unable to achieve a sustainable competitive advantage and promote business sustainability.
In addition, experienced users or consumers take advantage of their prior experience in making decisions and forming attitudes and intentions [19].However, the relevant literature is still devoid of empirical research that treats prior experience as a moderator of the effects of PEU, PE, and PU simultaneously on ATUSM.Although there are empirical studies on social media in hospitality and tourism, Sotiriadis's [13] (p. 216) recent examination of the literature on social media states, " . . . the research agenda is still long".Our discussion about the lack of empirical research concerning gender and prior experience as moderators proposed above is in this research agenda.
Second, recent studies still underscore the need for more research about social media in the context of hospitality and tourism [6,13,25].Sotiriadis's [13] work shows that there is little empirical research about social media associated with events/festivals.This actually echoes Leung et al.'s [8] conclusion that empirical research on social media in hospitality and tourism is still in its development stage.With this realization, our study responds to such a call for research by using a sample of students from the Korean Government Scholarship Program and tests gender and prior experience differences in relationships between PEU, PE, PU, and ATUSM, as well as the underlying mechanism that links PEU, PE, and PU to IAF.In short, our study is among the first to propose and examine the mediating and moderating effects mentioned above by using TAM and agentic and communal perspectives.
Lastly, the results of our study are likely to help managers ascertain gender-and prior experience-based differences and strategies that may promote festivals in various social networking sites.This would contribute to the accomplishment of a sustainable competitive advantage and enhancement of business sustainability.

Theoretical Focus and Hypotheses
Social media or social media marketing is an important tool for establishing and maintaining a sustainable brand image and enhancing long-term relationships with customers, cf.[26].Not surprisingly, social media serves as a significant platform on which consumer-generated content enables potential and actual buyers to share and obtain information and exchange ideas prior to purchase, re-purchase, and/or recommendation to others [27,28].A recent study focusing on a literature review of sharing tourism experiences in social media highlights that online reviews, as well as experience sharing by tourists, can make it possible for management to respond to several negative online reviews and manage e-complaints [13].This is important because consumers who plan to buy a product or attend a festival are interested in learning what ideas past and current buyers have and how the company or the organizer handles or has already handled critical service encounters.
Our study uses a technology acceptance model (TAM) to develop hypotheses that pertain to the underlying mechanism that links PEU, PE, and PU to IAF [10,15,29,30].An examination of the relevant literature also underscores the utilization of TAM [31][32][33][34].TAM specifically explains computer usage behavior [35].When applied to the use of social media in the context of our study, TAM proposes that individuals' PEU and PU of social media, as well as PE, predict their ATUSM, which leads to their propensity to attend festivals [10,15].In TAM, attitude toward using is defined as "the degree of evaluative affect that an individual associates with using the target system in his or her job" [36] (p.476).In the context of our study, 'attitude toward using' represents individuals' ATUSM for festival attendance.
PEU refers to " . . . the degree to which the user expects the target system to be free of efforts" [35] (p.985).Users' PEU refers to the degree which they find the website easy to manage, comprehend its compositions and operations with ease, and are able to obtain data and information when needed [36].Hsu and Lin's [34] study provided support for a positive relationship between PEU and attitude toward utilizing blog.Nasri and Charfeddine [37] showed that PEU increased attitude toward utilizing Facebook.Likewise, Morosan and Jeong [38] found a positive linkage between PE and attitude toward using web sites for hotel reservations.In short, PEU refers to the degree to which users find it easy to seek information about festivals on social media.
According to Davis et al. [39], users' behaviors will vary based on whether their motivation is intrinsic or extrinsic.As cogently discussed by Ayeh et al. [15], PE captures intrinsic motivation in lieu of extrinsic motivation because users are more interested in using social media to view photographs and videos, read comments, leave comments about restaurants, hotels, museums, events, and airlines online, and review destinations.Hsu and Lin [34] documented that PE enhanced attitude toward using blogs.Morosan and Jeong [38] also demonstrated a similar finding between PE and attitude toward using web sites for hotel reservations.
PU refers to the degree to which individuals believe utilizing an application or a particular system will enable them to perform the job better [29].Using social media to find information about festivals enhances their planning for festival attendance.As a result, consumers find social media useful when they are able to access new or newest data and information, seek information quickly, and obtain reliable information, cf.[10,15,40].In empirical terms, Nasri and Charfeddine [37] reported that PU heightened attitudes toward using Facebook.A similar finding between PU and attitude toward using web sites for hotel reservations was also reported in Morosan and Jeong's [38] study.
It seems that there are empirical findings that have focused on the effects of PEU, PE, and PU on consumers' ATUSM.These relationships (e.g., the effects of PEU and PU on attitude) are also supported in the framework of TAM, e.g., [23,36,[41][42][43][44].For example, Lee et al. [10] reported that PE triggered consumers' attitude toward using Facebook for festival searches.In a study of leisure travelers, Ayeh et al. [15] demonstrated that consumers' ATUSM for travel planning, as well as their intentions to use social media, increased due to PEU, PE and PU.In addition, a study done by Cheung and To [45] indicated that PU was a predictor of Chinese consumers' attitudes toward co-creation in social media.An observation that can be readily made based on the empirical findings presented above is that little is known about the effects of PEU, PE, and PU on users' ATUSM for festival attendance.This observation is consistent with Sotiriadis's [13] literature review on social media in festival settings.Accordingly, we predict: Hypothesis 1 (H1).PEU will exert a positive influence on ATUSM for festival attendance.
Hypothesis 2 (H2).PE will exert a positive influence on ATUSM for festival attendance.

Hypothesis 3 (H3). PU will exert a positive influence on ATUSM for festival attendance.
The empirical studies consistently reveal that attitude is positively associated with intention.Specifically, prior research indicated that attitude was positively associated with intention [35,38,46].The writings on social media also consistently demonstrate a positive association between attitude and intention.For instance, Ayeh et al.'s [15] study showed a positive linkage between the two constructs.Lee et al. [10] documented that attitude toward using Facebook exerted a strong positive influence (0.86) on users' IAF.Based on the above information, we predict: Hypothesis 4 (H4).ATUSM will exert a positive impact on users' IAF.
The hypotheses we have proposed so far implicitly refer to ATUSM as a full mediator.Consistent with TAM, there are studies supporting the PEU, PE, and PU → attitude → behavioral intentions relationships, e.g., [37].In the theoretical framework of TAM, our study contends that users' PEU and PU of social media as well as their entertaining and pleasant experiences associated with the use of social media predict their attitude that in turn governs their behavioral intentions.In empirical terms, Lee et al. [10] reported that PE was linked to IAF through the mediating role of attitude toward using Facebook.Ayeh et al. [15] found that ATUSM mediated the impacts of PEU, PE, and PU on proclivity to use social media.In light of the tenets of TAM and the limited evidence delineated above, we predict: Hypothesis 5 (H5).ATUSM will fully mediate the impacts of (a) PEU, (b) PE, and (c) PU on users' IAF.
Our study proposes that the positive impacts of PEU, PE, and PU on ATUSM are moderated by gender.Understanding gender-based differences is significant for user acceptance research [19].To develop such hypotheses, our study borrows the idea from agentic and communal perspectives.Specifically, men are more task-and goal-oriented (agentic perspective), while women are more relationship-oriented (communal perspective), e.g., [47].The 'I can find it myself' approach that is associated with higher assertiveness and task mastery makes PEU less important for males, cf.[48].PEU seems to be more important for female users because they exhibit computer anxiety at elevated levels, and such computer anxiety is negatively associated with computer self-efficacy [23].Not surprisingly, women who are more relationship-oriented can use an avoidance approach when confronted with high levels of stress that may reduce their self-efficacy, cf.[22,48].Accordingly, it is more important for female users who perceive social media easy to use for finding out about festivals or events.Hu et al. [49] argue that Arab female individuals possess higher anxiety about the use of computer technology than their male counterparts.However, their study does not provide any empirical support for gender as a moderator between PEU and attitude toward computer technology.
Zhang et al. [50] argue that women are inclined to care more about interpersonal and emotional aspects, while men are inclined to focus more on instrumental values.Being more relationship-oriented causes women be influenced by others [19].A recent study reveals that the majority of travel blog entries about Istanbul, Turkey were shared by female bloggers [51].This may be due to the fact that they enjoy sharing ideas and photographs about a destination.This is directly related to PE. Accordingly, women pay more attention to entertaining and pleasant experiences while seeking information about festivals.In short, this suggests that women are more sensitive to entertaining and pleasant experiences when they use social media to receive information about festivals.
Past research suggests that men are more task-oriented, e.g., [47], focus more on the achievement of instrumental goals [52], and tend to have higher assertiveness and task mastery [48].Under these circumstances, males are more interested in achievement needs [23] and the outcome of the information search process, and this is directly associated with PU [19].Accordingly, it is more important for males to use social media that enables them to obtain new, newest, and accurate information about festivals.Though limited, there is evidence supporting the premise that gender is a moderator between PU and attitude.Hu et al. [49] demonstrated that male users paid greater attention to PU that would result in attitude toward using computer technology at work.The findings reported above and our discussion based on agentic and communal perspectives implicitly suggest gender differences in relationships between PEU, PE, PU, and ATUSM.Therefore, we predict: Hypothesis 6a (H6a).Gender moderates the positive impact of PEU on ATUSM such that the positive impact is stronger among female users than among male users.Hypothesis 6b (H6b).Gender moderates the positive impact of PE on ATUSM such that the positive impact is stronger among female users than among male users.Hypothesis 6c (H6c).Gender moderates the positive impact of PU on ATUSM such that the positive impact is stronger among male users than among female users.
In addition to gender-based differences, our study contends that prior experience can play a moderating role on the association between PEU, PE, PU, and ATUSM for festival attendance.Users' prior experience refers to being " . . .more familiar and more knowledgeable about the technology of interest" [19] (p.69).Knowledge obtained from past behavior/experience will help to form users' attitude because knowledge is more accessible in memory due to past experience [53].Therefore, users' prior experience may show differences in relationships among the variables (i.e., PEU, PE, PU, and attitude) in TAM.
For experts or experienced users, PEU is likely to be less important.This is because of the fact that users may have gained sufficient knowledge through past experience with similar social media or social networking sites.Under these circumstances, PEU does not appear to be important for experienced users.This discussion is also consistent with Sun and Zhang's [19] proposition that PEU is less important for experienced users.In empirical terms, Taylor and Todd [53] found that the positive influence of PEU on attitude toward information technology use was stronger for inexperienced users.
As stated earlier, PE captures intrinsic motivation [15,39].Individuals are more interested in using social media to view photographs and videos, read comments, and review events/festivals.The actual process of doing this on social media should be interesting and entertaining [10].The ones with prior experience use their experience to shape their ATUSM for festival attendance.However, it appears that PE is more important for inexperienced users because such users expect to possess entertaining and pleasant experiences while using social media to find out about festivals.That is, the more attention inexperienced users pay to PE, the higher their ATUSM is.Otherwise, the absence of such experiences erodes their motivation to use social media or social networking sites.Our search of the extant tourism research demonstrates a lack of evidence about prior experience as a moderator of the effect of PE on ATUSM.
Experienced users are more interested in the outcome of the information search process.This is not surprising, because such users focus on usefulness and accessing new, newest, and accurate information about festivals quickly.In empirical terms, Taylor and Todd [53] hypothesized that the positive impact of PU on attitude toward using information technology was stronger for experienced users than for inexperienced users.However, their empirical study failed to provide support for this.Venkatesh and Bala [20] reported that the positive linkage between PEU and behavioral intention was weaker among the users with increasing experience.Venkatesh et al. [21] also reported that the positive association between behavioral intention and use behavior was weaker among users with increasing experience.In short, our study proposes that prior experience helps shape ATUSM when users perceive that using social media enables them to obtain useful information about festivals promptly.In light of the discussion presented above, we predict: Hypothesis 7a (H7a).Prior experience moderates the positive influence of PEU on ATUSM such that the positive influence is stronger among inexperienced users than among experienced users.

Hypothesis 7b (H7b).
Prior experience moderates the positive influence of PE on ATUSM such that the positive influence is stronger among inexperienced users than among experienced users.

Hypothesis 7c (H7c).
Prior experience moderates the positive influence of PU on ATUSM such that the positive influence is stronger among experienced users than among inexperienced users.

Research Model
The hypothesized relationships developed based on the precepts of TAM, as well as agentic and communal perspectives, are presented in the research model in Figure 1.The model posits that ATUSM is a full mediator of the impacts of PEU, PE, and PU on users' IAF.Broadly speaking, users' favorable perceptions of ease of use and usefulness of social media about festivals, as well as their enjoyment experiences while seeking information about festivals, activate their attitude toward using social media.Such users in turn display intentions to attend festivals.
The model also posits that gender is a moderator of the effects of PEU, PE, and PU on ATUSM for festival attendance.Specifically, the positive association between ease of use and enjoyment and attitude toward utilizing social media is stronger among female users than among male users.However, the positive association between usefulness and attitude toward utilizing social media is stronger among male users than female users.Prior experience functions as a moderator of the effects of PEU, PE, and PU on ATUSM for festival attendance.That is, the positive impacts of ease of use and enjoyment on attitude toward utilizing social media are stronger among inexperienced users than among experienced users.On the other hand, the positive relationship between usefulness and attitude toward utilizing social media is stronger among experienced users than among inexperienced users.

Sample and Data Collection
The sample used in our study consisted of undergraduate, graduate, and doctoral students of the Korean Government Scholarship Program.There are at least three reasons for selecting these students.First, students who study abroad are treated as buyers in a lucrative niche market.However, such a group of buyers is still relatively understudied, and little is known about their travel behaviors.Second, these students stay in Korea for at least three years for their higher education.They visit various destinations in the country.Therefore, they are more familiar with the Korean culture than those who are considered short-stay tourists.Third, these students spend significant time seeking information about leisure activities (including local events and festivals) through social media.However, those who are considered short-stay tourists do not seem to spend such time seeking information about these activities via social media.
In addition to the scale items, several questions about social media usage were added to the main survey.The questionnaire also included items about respondents' demographic profile regarding gender, age, education level, and ethnicity.On top of that, all of the subjects had prior experience with the use of social media.
Data were collected using both face-to-face and e-mail surveys by the researchers.Four hundred and thirty-seven questionnaires were distributed to the potential respondents and a total of 262 questionnaires were gathered.Of the returned questionnaires, 42 were discarded because they were unreliable or incomplete.In addition, the dataset was checked for outliers using the Mahalanobis distance [54] and for normality using skewness and kurtosis prior to the analysis [55].Among the 220 questionnaires, 29 questionnaires were discarded.Hence, the final sample of this study consisted of 191 students.The response rate was 43.7% (191/437).Table 1 presents the subject profile.Table 2 provides the social media usage characteristics and festival attendance.

Sample and Data Collection
The sample used in our study consisted of undergraduate, graduate, and doctoral students of the Korean Government Scholarship Program.There are at least three reasons for selecting these students.First, students who study abroad are treated as buyers in a lucrative niche market.However, such a group of buyers is still relatively understudied, and little is known about their travel behaviors.Second, these students stay in Korea for at least three years for their higher education.They visit various destinations in the country.Therefore, they are more familiar with the Korean culture than those who are considered short-stay tourists.Third, these students spend significant time seeking information about leisure activities (including local events and festivals) through social media.However, those who are considered short-stay tourists do not seem to spend such time seeking information about these activities via social media.
In addition to the scale items, several questions about social media usage were added to the main survey.The questionnaire also included items about respondents' demographic profile regarding gender, age, education level, and ethnicity.On top of that, all of the subjects had prior experience with the use of social media.
Data were collected using both face-to-face and e-mail surveys by the researchers.Four hundred and thirty-seven questionnaires were distributed to the potential respondents and a total of 262 questionnaires were gathered.Of the returned questionnaires, 42 were discarded because they were unreliable or incomplete.In addition, the dataset was checked for outliers using the Mahalanobis distance [54] and for normality using skewness and kurtosis prior to the analysis [55].Among the 220 questionnaires, 29 questionnaires were discarded.Hence, the final sample of this study consisted of 191 students.The response rate was 43.7% (191/437).Table 1 presents the subject profile.Table 2 provides the social media usage characteristics and festival attendance.In addition to the scale items, several questions about social media usage were added to the main survey.The questionnaire also included items about respondents' demographic profile regarding gender, age, education level, and ethnicity.On top of that, all of the subjects had prior experience with the use of social media.
Data were collected using both face-to-face and e-mail surveys by the researchers.Four hundred and thirty-seven questionnaires were distributed to the potential respondents and a total of 262 questionnaires were gathered.Of the returned questionnaires, 42 were discarded because they were unreliable or incomplete.In addition, the dataset was checked for outliers using the Mahalanobis distance [54] and for normality using skewness and kurtosis prior to the analysis [55].Among the 220 questionnaires, 29 questionnaires were discarded.Hence, the final sample of this study consisted of 191 students.The response rate was 43.7% (191/437).Table 1 presents the subject profile.Table 2 provides the social media usage characteristics and festival attendance.

Measurement Development
The study constructs were operationalized adopting or adapting items from previous empirical studies.The items are given in Table A1.Broadly speaking, four items were used to measure PEU, e.g., [10,56], while PU was assessed via six items, e.g., [30,57].PE was measured through five items, e.g., [30,58].Three items were used to operationalize ATUSM, e.g., [34].Lastly, a three-item scale [10,59] was utilized to measure IAF.Each item in the aforesaid constructs was rated on a seven-point scale ranging from 1 (strongly disagree) to 7 (strongly agree).
Face validity refers to "extent to which the content of the items is consistent with the construct definition based solely on the researcher's judgment" [60] (p.689).This was achieved in light of the researchers' judgment.

Assessment of the Measurement Model
The measurement model was tested using confirmatory factor analysis.As a result of the initial assessment of the measurement model, one item (PU4) from the PU construct was removed based on correlation measurement errors.The final five-factor measurement model fit the data well based on the following statistics: χ 2 [160] = 284.324,Q = 1.777 (≤5; [61]); Standardized root mean square residual (SRMR) = 0.035 (≤0.08 [55,60]); Root mean square error of approximation (RMSEA) [90% confidence interval (CI)] = 0.064 (≤0.08 [55,60]) [0.052; 0.076]; Tucker-Lewis index (TLI) = 0.958 (≥0.90 [55,60]); and Comparative fit index (CFI) = 0.965 (≥0.90 [55,60]).As reported in Table 3, all items loaded onto their underlying latent variables and the loadings were significant.The average variance extracted for each latent construct was also greater than the recommended cut-off value of 0.50 [60,62].The aforesaid results collectively demonstrated that convergent validity was achieved [60].The findings further illustrated that all measures were reliable due to the fact that each composite reliability score was greater than the recommended cut-off value of 0.60 [63], and coefficient alphas for the observed variables exceeded the 0.70 threshold [60].
Table 4 shows that the average variances extracted by latent constructs were greater than any squared correlations among constructs.This implied that the constructs were empirically distinct from each other [63].In summary, the measurement model's discriminant validity was satisfactory.

Main Effects and Hypotheses Testing
H1 predicts that PEU will exert a positive influence on ATUSM for festival attendance.As shown in Figure 2, the empirical findings do not support H1 because PEU (γ 11 = 0.147, t = 1.237, p > 0.05) has no significant influence on ATUSM.That is, users' perceptions of ease of use do not influence their attitude toward utilizing social media.
H2 proposes that PE will have a positive influence on ATUSM for festival attendance.Similarly, H3 suggests that PU will have a positive influence on ATUSM for festival attendance.The results provided empirical support for H2 and H3 because PE (γ 12 = 0.262, t = 3.463, p < 0.01) and PU (γ 13 = 0.461, t = 4.066, p < 0.01) had a positive impact on ATUSM.That is, users' favorable perceptions of enjoyment and usefulness influence their attitude toward utilizing social media.
H4 predicts that ATUSM will exert a positive effect on IAF.The results showed that ATUSM (β 21 = 0.822, t = 10.925,p < 0.01) positively affected IAF.Hence, the empirical data supported H4.This finding highlights the importance of ATUSM, which fosters IAF.That is, users who report ATUSM at elevated levels are more inclined to attend festivals.The findings accounted for 60.4% of the variance in ATUSM and 67.5% in IAF.
H5 suggests that ATUSM will act as a full mediator of the impacts of (a) PEU, (b) PE, and (c) PU on IAF.The Aroian version of the Sobel test showed the significant and positive findings for the paths from PE (z = 3.448, p < 0.01) and PU (z = 3.807, p < 0.01) to IAF via ATUSM (Figure 2).When using the bootstrapped CI, mediation was demonstrated by the exclusion of zero from the CI around the indirect effect.The PE → ATUSM → IAF (unstandardized indirect point estimate: 0.180; 95% CI [0.080; 0.317]) and PU → ATUSM → IAF (unstandardized indirect point estimate: 0.337; 95% CI [0.173; 0.527]) relationships did not contain zero.The results demonstrated that ATUSM acted as a full mediator in the association between PE and PU and IAF.Thus, the aforesaid results collectively supported H5b and H5c.In short, users who entertain themselves while seeking information via social media and find the relevant information useful display higher ATUSM, which in turn gives rise to higher IAF.
No mediating effect could be found regarding the PEU → ATUSM → IAF because PEU was not significantly related to ATUSM.

Mediating Effects and Hypotheses Testing
H5 suggests that ATUSM will act as a full mediator of the impacts of (a) PEU, (b) PE, and (c) PU on IAF.The Aroian version of the Sobel test showed the significant and positive findings for the paths from PE (z = 3.448, p < 0.01) and PU (z = 3.807, p < 0.01) to IAF via ATUSM (Figure 2).When using the bootstrapped CI, mediation was demonstrated by the exclusion of zero from the CI around the indirect effect.The PE → ATUSM → IAF (unstandardized indirect point estimate: 0.180; 95% CI [0.080; 0.317]) and PU → ATUSM → IAF (unstandardized indirect point estimate: 0.337; 95% CI [0.173; 0.527]) relationships did not contain zero.The results demonstrated that ATUSM acted as a full mediator in the association between PE and PU and IAF.Thus, the aforesaid results collectively supported H5b and H5c.In short, users who entertain themselves while seeking information via social media and find the relevant information useful display higher ATUSM, which in turn gives rise to higher IAF.
No mediating effect could be found regarding the PEU → ATUSM → IAF because PEU was not significantly related to ATUSM.

Moderating Effects and Hypotheses Testing
To test the moderating roles of gender (MALE vs. FEMALE) and prior experience (EXPERT vs. NOVICE) of using social media on the relationship between PEU/PE/PU and ATUSM based on a multi-group approach, we tested the equivalence of a specific path by constraining the particular parameter of interest in the nested model to be equal in sequence, and all paths in the baseline model were allowed to be freely estimated.Using a multi-group analysis, we used a mean split to divide the dataset into two sub-groups for prior experience of using social media.With the mean at 6.05 years, the novice group includes 114 cases while the expert group includes 77 ones.Also, the female group included 113 cases, while the male group included 78 cases.The χ 2 difference test, by comparing the baseline model with the nested model (constrained to be equal) calculated for 1 degree of freedom, allows equivalence to be gauged for the specific parameter of interest across two groups [63].The results for the equivalence tests of the moderating hypotheses paths are given in Table 5. [1] = 4.718 * (H7c: Supported) Notes: PEU = perceived ease of use, PE = perceived enjoyment, PU = perceived usefulness, ATUSM = attitude toward suing social media, SRMR = Standardized root mean square residual, RMSEA = Root mean square error of approximation, TLI = Tucker-Lewis index, CFI = Comparative fit index, ns = not significant.t-values (one-tailed test): t > 1.65, * p < 0.05; t > 2.33, **p < 0.01.χ 2 values: χ 2 > 3.84, * p < 0.05; χ 2 > 6.64, ** p < 0.01.
H6b predicts that the positive impact of PE on ATUSM is stronger among female users than among male users.For the path from PE to ATUSM, the estimate across groups was different, and a significant χ 2 difference was observed (χ 2 [1] = 5.272, p < 0.05).Specifically, the effect was higher for the female group than for the male group (MALE: γ 12 = 0.094, t = 0.713, p > 0.05; FEMALE: γ 12 = 0.392, t = 2.955, p < 0.01).Thus, this supported the moderating role of gender on the association between PE and ATUSM.
H6c proposes that the positive impact of PU on ATUSM is stronger among male users than among female users.The estimates for the path coefficients from PU to ATUSM across groups were compared.There was a significant difference.That is, the path was different across gender groups (χ 2 [1] = 6.316, p < 0.05).In particular, the effect was greater for the male group than for the female group (MALE: γ 13 = 0.685, t = 5.258, p < 0.01; FEMALE: γ 13 = 0.224, t = 2.473, p < 0.01).This result was thus consistent with H6c.H6a, which referred to gender as a moderator between PEU and ATUSM, was not supported.This was because of the fact that the difference in χ 2 between the unconstrained model and the constrained model was not significant (χ 2 [1] = 0.085, p > 0.05).
We asked a question in the title of our paper: 'Do gender and prior experience moderate the factors influencing attitude toward using social media for festival attendance?' The answer is 'yes'.Specifically, the findings presented in our study suggest that PE exerts a stronger positive impact on female users' ATUSM for festival attendance relative to male users.Female users are more sensitive to entertaining and pleasant experiences when seeking information about festivals through social media or social networking sites.This may be due to the fact that women are more relationship-oriented, e.g., [22], are likely to care more about interpersonal and emotional aspects [50], and enjoy sharing ideas, feedback, videos and photographs, cf.[51].The findings further suggest that PU has a stronger positive influence on male users' ATUSM for festival attendance.Male users focus more on new, newest, and correct information when they spend time seeking information about festivals.This refers to the achievement of instrumental goals, cf.[52] and the outcome of the information search process [19].The finding pertaining to gender as a moderator between PU and ATUSM for festival attendance is concordant with limited research in the existing knowledge base, cf.[49].
Consistent with the study hypotheses, prior experience functions as a moderator of the impacts of PE and PU on ATUSM for festival attendance.Broadly speaking, the positive influence of PE on ATUSM for festival attendance is stronger for inexperienced users (NOVICE) than for experienced users (EXPERT).As discussed in other studies [5,10], using social media to view photographs and videos, read online reviews, review events, and exchange ideas and information with other users should be interesting and entertaining.This is especially important for inexperienced users (NOVICE), because such users expect to have interesting and entertaining experiences while using social media to obtain information about festivals.Consequently, this will motivate these users to exhibit a continued usage of social media.
In addition, the positive effect of PU on ATUSM for festival attendance is stronger for experienced users (EXPERT) than for inexperienced users (NOVICE).Experienced users (EXPERT) are more familiar with technology or social media they have utilized so far.When they reach the information needed about festivals through social media, their attitude toward using it will increase.Our finding regarding prior experience as a moderator of the influence of PU on ATUSM is in line with past research investigating the moderating role of prior experience on the association between PU and attitude toward using information technology [53].
The findings suggest that gender and prior experience do not significantly moderate the effect of PEU on ATUSM for festival attendance.Today's female users are accustomed to using recent technology and know how to manage social networking sites despite some difficulties.These sites offer friendly web pages and services and enable such users to control their accounts in their native language or in English.This explanation is also valid for the absence of prior experience as a moderator because it seems that inexperienced users are not in need for much experience about social networking sites.When they encounter any specific problems, they are likely to receive assistance from another user to manage the situation.
The findings further suggest that ATUSM is a full mediator between PE and PU and users' IAF.The results presented here are concordant with TAM [10,15].It appears that users seek entertaining and pleasant experiences while dealing with social media, seeking information about festivals, reading feedback, and viewing videos and photographs.It also seems that users are more interested in reaching useful information about festivals.Under these circumstances, both PE and PU trigger users' ATUSM that in turn gives rise to IAF.However, PEU does not have any bearing on ATUSM.This may be due to the fact that users can obtain assistance from other users to solve specific problems emerging from a relevant web page or its link.
By assessing the aforementioned direct, mediating, and moderating effects, our study contributes to current knowledge and delineates several useful implications for the accomplishment of a sustainable competitive advantage and enhancement of business sustainability.Specifically, the extant literature lacks empirical evidence pertaining to gender and prior experience that can moderate the impacts of PEU, PE, and PU on ATUSM for festival attendance.Gender and prior experience moderate the impacts of users' perceptions of entertaining and pleasant experiences and usefulness of social media in their decision-making on their attitude toward using social media for festival attendance.Management can design and implement several useful strategies by considering gender and prior experience as the two moderators.These gaps are evident, especially in Sotiriadis's [13] literature review, and Leung et al.'s [8] and Venkatesh et al.'s [21] studies.
Our study also expands current knowledge pertaining to ATUSM as a full mediator of the impacts of PE and PU on IAF.Users' perceptions of usefulness of social media as well as their enjoyment associated with the utilization of social media positively influence attitude toward using social media that in turn directs their proclivity to attend festivals.The voids regarding the abovementioned linkages are evident in Sotiriadis's [13] literature review and Leung et al.'s [8] study.

Management Implications
The findings of our study recommend several useful guidelines for managerial action to accomplish a sustainable competitive advantage and foster business sustainability.First, management of companies should make sure that users do not have any difficulty in finding information about festivals in the relevant social networking sites.What is promoted about the festival should make it possible to reach new, newest, and correct information about the attractive features of the festival.While doing so, users should have the opportunity to download videos that include promotion of the festival, view photographs about the festival, and possess entertaining and pleasant experiences.This is critical, since PE and PU predict users' ATUSM for festival attendance and therefore influence their IAF.Users also need to find the opportunity to have an interaction with the organization via pictures posted on company news, brand commercials, and company images [12].
Second, opinion leaders who are highly respected in a society can play a significant role in making users read the content of the information about the festival.Then such users can share the information via social networking sites.As the study findings suggest, the positive effect of PE on ATUSM for festival attendance is stronger for female users, while the positive influence of PU on ATUSM for festival attendance is stronger for male users.Male users who are more task-and goal-oriented attach more importance to usefulness of information as a result of their search in the social networking sites.Female users who are more relationship-oriented attach more importance to entertaining and pleasant experiences.To increase the number of festival attendees, management should work with the social networking sites that create a platform for enabling male users to reach new, newest, and correct information and making it possible for female users to enjoy reading online reviews about past organizations and sharing information and other relevant issues about the festival.
Third, prior experience is a moderator between PE and PU and ATUSM for festival attendance.The positive linkage between PE and ATUSM for festival attendance is stronger among inexperienced users (NOVICE), while the positive association between PU and ATUSM for festival attendance is stronger among experienced users (EXPERT).The presence of photographs, videos, and/or stories about past festivals organized in different settings and posted in the social networking sites should motivate inexperienced users to have fun and make them entertain themselves and obtain pleasant experiences.On the other hand, experienced users' perceptions of usefulness of information should enable them to acquire sufficient, timely, and correct information about the festival from Internet, cf.[7].

Limitations and Avenues for Future Research
There are several limitations which, far from detracting from its value, highlight avenues for future research.First, cross-sectional data were gathered from the students of the Korean Government Scholarship Program.This prevents us from making generalizations about causality.To overcome such a limitation, in future studies collecting data from the relevant parties via a temporal separation between the variables would be a potential remedy.
Second, future research could benefit from a cross-national study.For example, data could be collected from different countries such as the United States, China, and Turkey to make comparisons about the direct, mediating, and moderating effects tested in this study.
Third, future research could take into consideration the effects of the potential advertising efforts of companies on tourists' proclivity to use social media to find about tourism activities, cf.[64].In closing, our study considered the 'ATUSM for festival attendance' and 'IAF' constructs.To enhance our understanding, future research could utilize user satisfaction, linking PEU, PE, and PU to 'electronic word of mouth'.

Table 3 .
Reliability and convergent validity tests.

Table 4 .
Descriptive statistics and discriminant validity test.
Notes: Correlations are presented in the lower off diagonal, squared correlations (i.e., shared variances) are presented in parentheses [ ] in the lower off diagonal, and average variances extracted are presented along the diagonal.All correlations (one-tailed) are significant at p < 0.01.

Table 5 .
Moderating role gender and prior experience and hypotheses test results.