Beauty and Celebrity: Korean Entertainment and Its Impacts on Female Indonesian Viewers’ Consumption Intentions

: The Korean entertainment industry’s popularity and global reach has prompted private companies and destination management organizations to place Korean products and position Korea as a destination to international viewers. Korean celebrities have also been found to form positively perceived Korean beauty standards, consequently serving as an antecedent of Korean beauty product awareness and perceived beauty product quality among international audiences. Although previous studies have found that audience involvement with entertainment mediums has positively impacted product purchase and visitation, a comprehensive investigation of the audience involvement’s multi-dimensionality has yet to be conducted, speciﬁcally, on the distinction between Korean entertainment (movies, TV dramas, TV shows, etc.) involvement and celebrity worship. The present study attempts to ﬁll the research gap by comprehensively examining the role of audience involvement with Korean entertainment, celebrity worship, brand awareness, and perceived Korean beauty product quality on product purchase intentions as well as intentions to visit Korea. A total of 327 female Indonesian Korean entertainment audiences then validated the proposed conceptual framework. The structural equation modeling results found emotional involvement, referential reﬂection, and entertainment-driven celebrity worshipping tendencies to be signiﬁcant antecedents of Korean beauty product purchase and Korea visit intentions.


Introduction
The entertainment industry provides an avenue for content creators to capture their audience's attention and for brands to strategically place their products, hoping to gain the undivided attention of audiences [1]. The practice of product placement is also common in the Korean entertainment industry [2][3][4][5]. Many brands now use movies and TV dramas as a platform to launch and promote new products [6][7][8]. The entertainment industry produces media in various forms, such as movies, TV programs, radio, music, dramas or series, theater performances, and many more. Dramas have been one of the most effective media to reach consumers, as the audience tends to develop deeper involvement due to the longer viewing times [9]. The national tourism development organization, the Korea Tourism Organization (KTO), also resorted to Korean dramas as a medium to promote Korea's tourism internationally [10]. The rise in popularity of Korean pop culture, where the various mediums of the Korean entertainment industry have become well-known internationally, has allowed both Korean private companies and governmental organizations, such as the KTO, to reach the international audience more effectively [11][12][13]. The Korean entertainment industry's popularity is often referred to as the Korean wave, or "Hallyu" [14]. Studies have found that Korean entertainment media viewers have caused attitudes towards Korea to shift, and have also motivated audiences to visit Korea

Korean Entertainment Audience Involvement
Arguably the identification theory by Freud introduced in the early part of the 20th century was the first known building block to the contemporary understanding of audience involvement [32,36]. The identification theory refers to when an individual adopts the attitudes, values, beliefs, and behaviors of another individual or group [32,37]. Audience involvement has been defined and utilized in several research contexts [9,19,28,30,[38][39][40]. However, the effectiveness of audience involvement in explaining consumer preferences and behaviors can vary between the type of media consumption and research contexts [28]. Therefore, the present study adopts the definition of audience involvement developed for the context of entertainment-education, which Korean dramas, films, and TV shows generally contain [41]. Sood's conceptualization of audience involvement occurs when the audience engages in certain media programs that result in the formation of parasocial interactions and reflection [42]. Reflection refers to when the viewers think deeply about the delivered messages from the media they have consumed [42]. Parasocial interaction refers to when an individual imagines themselves forming a close relationship with a public figure in media.
The operationalization of audience involvement in the present study followed Kim's research on a Korean drama and its influence on the experiences of tourists at the filming location in Korea [28]. The study found that three dimensions of audience involvement were effective in predicting intimacy and other sensory experiences. The first dimension is termed "behavioral involvement". It refers to the degree to which an audience alters their behaviors to view more content or talk with others about the media [7,28,42]. The second dimension is emotional involvement. This is defined as the degree to which the audience identifies him or herself with the characters, locations, or storyline associated with the media [8]. Lastly, the referential reflection dimension refers to how an audience relates his or her personal experiences to the media [7].
Audience involvement was found to affect consumers in a variety of ways, such as pleasure in viewing experiences, satisfaction, and post-viewing behavioral intentions, including purchasing behavior [7]. The purchase behaviors can be either the products they see in the media or memorabilia related to the media. Moreover, audience involvement can also affect the intentions to visit places that the audience sees in the media. Filminduced travel has been a sector that relies heavily on the level of involvement by the audience [7,8,10,15,19]. A study of film-induced tourism focused on how Korean dramas can motivate viewers to visit Korea [10]. The study focused on the general viewership of Korean dramas and non-Korean celebrities. Similarly, another study also found that Japanese tourists to the UK are influenced by films and TV dramas [29]. The present study posits Korean entertainment audience involvement to measure audience involvement with the various types of media produced by the Korean entertainment industry. The different types of media include Korean dramas, Korean movies, Korean music, Korean TV shows, and more.

Korean Celebrity Worship
Although many consumer research studies operationalized involvement as a unidimensional construct, involvement consists of multiple stages in the psychology context [25,32,[38][39][40]. The main reason supporting the multidimensional construct of involvement is due to the degree of psychological absorption [23]. Psychological absorption refers to an attempt to establish a sense of identity and fulfillment connected with a person or a celebrity that can eventually lead to an addiction. Based on this principle, the level of involvement is believed to consist of four stages along a single continuum [32]. In the context of the present study, a low involvement level suggests the audience starts to grow emotionally and logically attaches to the story [30,32,43]. Eventually, the audience reaches a stage where they begin to put the celebrities into their life priorities (worship) [24,32,34].
Worship is known as the strongest form amongst other stages of audience involvement. A study found that celebrity worshippers may even trust their favorite celebrities more than their own friends [20]. Yet, the understanding of consumer research, especially in the tourism context, is limited. Additionally, the relationship between worship and the lower intensity of involvement has not been studied in a tourist behavior context. More importantly, the transition from a lower level of involvement to a higher level, or the worship level, also implies a change in context. Specifically, an audience may be involved with general Korean entertainment media in a broader sense. However, as the intensity increases, the contexts of involvement could evolve into worshipping specific celebrities. Most empirical studies found that worshippers are most apparent when the context of interest is a person [20,[23][24][25]. The importance of celebrities and the deeper connection they can cultivate with fans is fundamental to the marketing concepts of celebrity endorsements and brand ambassadorship [21,22].
Worship can be further divided into three levels, depending on the degree of worship [21,23,24]. Low worship level refers to when an audience watches, sees, or reads about the celebrity. On the other hand, a high level of worship can lead to a change in a person's social character and ultimately leads to pathological behaviors, such as erotomania, attempts to harm the celebrities, or threatening letters [23,25]. In the present study, Korean celebrity worship is operationalized using three constructs [23]. The first construct, entertainment-social, represents the mildest level of worship. It suggests that fans are attracted to the celebrity because of their entertainment abilities. Next, the intense-personal construct suggests intensive and compulsive feelings starting to form around the celebrity, such as having frequent thoughts about the celebrity even if they do not want to. The third construct, borderline pathological, reflects extreme attitudes fans may hold towards their favorite celebrity in a way that they may even engage in illegal activities on their behalf.
The current research project proposes that audience involvement in the contexts and storylines of Korean entertainment should be separately examined from fans' intense worship of individual Korean celebrities. In other words, when a fan of the Korean entertainment industry mostly focuses on the storylines and enjoys discussions related to their favorite Korean dramas, TV shows, and so forth, their level of involvement is lower than those who worship celebrities. Worship is an intense form of parasocial interaction where an imaginary interaction or relationship between the audience and the celebrity in the media is formed [30,43]. At the same time, worshipping individual celebrities implies an ample degree of involvement in the entertainment media as well [32]. Therefore, before an individual reaches the celebrity worship stage, the individual is required to exhibit a degree of Korean entertainment involvement. However, worshipping tendency is mostly studied when the celebrity is the main object of content [19,44]. For example, a study on Taiwanese film tourists focused only on the influence of celebrity involvement and celebrity worship on destination image [19]. Another study found that Korean entertainment fans from Singapore exhibit parasocial relationships with Korean celebrities primarily through social media interactions [44]. Hence, the relationship between audience involvement and celebrity worship has not been well studied. Simultaneously, according to the four levels of involvement from the concept of psychological absorption, involvement would influence celebrity worship dimensions. Consequently, the following hypothesis was developed: Hypotheses 1a-i (H1a-i): Audience involvement (behavioral involvement, emotional involvement, and referential reflection) significantly impacts celebrity worship (entertainment-social, intense-personal, and borderline pathological) of female Indonesian Korean entertainment audiences.

Brand Awareness and Perceived Product Quality
Brand awareness is critically important due to its ability to differentiate one product from another in the competitive market [27,45]. In a scenario where a buyer is shopping for a product, brand awareness has been defined as the buyer's ability to identify and distinguish a brand within a given category with adequate detail and precision before making the purchase decision [46]. Moreover, brand awareness consists of recognition and recall [47]. Recognition refers to the buyer's ability to recognize the products at the time of purchase [48]. On the other hand, recall is the ability to remember the brand and the product prior to purchase [47]. In practice, a high level of brand awareness suggests an increased likelihood of a consumer purchasing the product [49][50][51]. Additionally, brand awareness has been found to also influence consumer trust and favorable brand image [3,52]. In the context of this study, brand awareness refers to the awareness of Korean beauty products brands.
The concept of perceived product quality is closely related to consumer expectation [53]. Yet, the perceived quality is distinctively different than awareness, in that awareness can be a result of advertiser exaggeration in product promotion [54]. On the other hand, a consumer's perceived product quality is based on their own experiences with the products or other trusted sources of information [33]. Moreover, perceived quality is generally associated with a favorable attitude towards the product. Hence, perceived quality has been found to be an effective antecedent of product purchase intentions [55]. The present research focuses on Korean beauty products. Thus, the variable is specifically referred to as the perceived quality of Korean beauty products.
Similar to entertainment content from other countries, Korean brands have spent a considerable amount of funds and effort to place products in various forms of media [2,4,5,56]. Combined with the international success that Korean entertainment has achieved, product placements in Korean dramas, movies, and TV shows have the potential to reach a far wider audience around the globe [4]. The visual appearance of celebrities can become the cue for international viewers regarding their perception of Korean beauty products brands and perceived quality [11][12][13]57]. Previous studies found that celebrities' appearances enhance consciousness of one's body image at the same time. Their appearances also cause the audience to idolize celebrities [18,21,24]. Hence, this research project hypothesized that female audiences who worship Korean celebrities might exhibit a higher awareness level of Korean beauty product brands. In addition, the favorable perceived quality of Korean beauty products can be influenced by high brand awareness. The following hypotheses were proposed as a result: Hypotheses 2a-c (H2a-c). Celebrity worship (entertainment-social, intense-personal, and borderline pathological) significantly impact Korean beauty product brand awareness among female Indonesians.

Hypotheses 3 (H3).
Korean beauty product brand awareness significantly impacts the perceived product quality of Korean beauty products among female Indonesians.

Purchase and Visit Intentions
Purchase intentions can be described as an act of planning to buy products or services in the future [1]. Past research has found that consumer attitudes are one of the most critical antecedents in predicting purchase intentions [11,12,18,52,58]. Moreover, the decision-making process can be influenced by factors such as brand awareness, comparing alternatives, perceived quality, and perceived credibility of the brands and their products [33,48,50,52,54,[59][60][61][62][63]. On the other hand, destination choice is generally understood to be influenced by similar factors as product purchase intentions. Nevertheless, the nature of tourism destination choice and visit intentions can be considered as a highly involved process and usually requires longer planning [64]. Therefore, planning a travel itinerary implies higher costs, involves perceived risks, requires group consensus on destinations, and much more [65][66][67][68]. Hence, Korean beauty product purchase intentions and Korea visit intentions represent two different types of decision-making, specifically, a lower-involvement decision and a higher-involvement decision.
Past studies have found that attachment to celebrity, celebrity endorsement, and attachment to media content influence consumer attitudes, preferences, experiences, and consumption intentions [7,8,10,15,19,21,22,69]. Specifically, highly involved viewers of dramas, movies, TV dramas, and other mass media content have been found to be heavily influenced by these content on their preferences, such as travel destination, attraction visits, and they replicate their favorite celebrities' actions from the movie or drama [15,20]. The content from mass media is usually delivered in rich mediums. Thus, it can portray alluring images of products or a place. Consequently, it helps audiences form favorable attitudes towards a place or a product [70]. Hence, film-induced tourism creates curiosity that then motivates people to travel and experience the same sensation as presented in the films and dramas [15,16]. Given these empirical findings, Hypotheses 4 and 5 were developed. Additionally, the proposed conceptual model is graphically illustrated in Figure 1.

Hypothesis 4 (H4)
. Perceived product quality of Korean beauty products significantly impacts Korean beauty product purchase intentions among female Indonesians.

Hypothesis 5 (H5)
. Perceived product quality of Korean beauty products significantly impacts Korea visit intentions among female Indonesians.

Measurement Items and Survey Development
The measurement items used in the questionnaire were adapted from previous related studies. Thus, the items have all been validated by previous studies. A total of 53 items were adopted to measure ten constructs. Audience involvement consisted of three dimensions, and all items were adopted from Kim (2011) [28]. These account for behavioral involvement and were measured by six items. Emotional involvement and referential reflection were measured by six and three items, respectively. Similarly, celebrity worship also consisted of three sub-dimensions. These were entertainment-social (ten items), intense-personal (eight items), and borderline pathological (three items). Hence, a total of 22 items used by Yen and Croy (2016) [19] and Maltby et al. (2005) [24] were used to measure all three dimensions of celebrity worship. Next, the brand awareness (five items) and perceived quality (four items) constructs both used items from the study by Buil et al. (2013) [71]. The intention to purchase Korean beauty products was measured by four items from Lu et al. (2014) [3]. Lastly, Korea visit intentions were measured using three items previously used in the study of Tan and Wu (2016) [72]. The list of all measurement items can be seen in Table A1 in the Appendix A section. A five-point Likert-type scale was used. The scale labels started from 1 (strongly disagree) to 5 (strongly agree).
A brief cover letter was then added to the survey. It ensured the participants were aware of the research content before participating. Additionally, demographic questions were added to the survey. The final draft of the survey was then subjected to pre-testing. The comments and suggestions from the pre-testers were minimal. Nevertheless, their com-ments and suggestions were taken into account before the survey was finalized. The final survey was then translated into Indonesian by a fluent speaker of both languages. Another pre-test was conducted using the translated version before finalizing both the English and Indonesian versions of the survey.

Sampling and Data Collection
The target research population for the present project were female Indonesian consumers who consume Korean entertainment products, such as Korean series, Korean variety shows, K-Pop shows, Korean movies, and more. Additionally, they should also have had experience and interest in using Korean beauty products in their daily lives. A qualifying question was inserted at the start of the survey. The question was, "Have you watched at least one Korean dramas/movies/TV shows in the last two years?" Only if the answer was "yes", were they asked to complete the questionnaire. The self-administered survey was then distributed online using Google Forms. The study deployed a nonprobability convenience sampling to recruit research participants by distributing the online survey on the social media platform Instagram and instant messaging applications, such as Kakao Talk, Line, and Whatsapp. The data was collected between February and March 2020. A total of 327 completed surveys were collected.

Sample Profiles
The raw data were first screened for any irregularities. First, there were 12 cases with a large amount of missing data. These cases were excluded from further analysis. Then, potential outliers were examined. However, no outliers were identified. Possible unengaged responses were also examined by calculating the standard deviation for each case. Any cases showing a standard deviation lower than 0.5 were subjected to manual inspection [73]. No responses exhibited any traits of unengaged participants. Subsequently, skewness and kurtosis scores for all measurement items were calculated. The skewness scores ranged from −2.294 and 1.515, below the recommended threshold of +3 or −3 [74]. Kurtosis scores ranged from −1.320 to 5.766, and also fell within the range of the absolute value of 10 [75]. Given the results, the data did not violate the assumption of normality and was suitable for subsequent structural equation modeling (SEM) analysis. After screening, a total of 315 samples were subjected to further data analysis.
The full respondent profiles can be seen in Table 1. Notable characteristics are that the majority of the respondents were students and full-time employees age between 21 and 30 years old. Most were not married and had a relatively low monthly income. Most of the respondents were bachelor's degree holders. Lastly, all respondents were Indonesian apart from three Filipina respondents. However, the Filipina respondents were residing in Indonesia.

Principle Component Analysis
The 15 measurement items of audience involvement were first subjected to the principal component analysis (PCA) extraction method with varimax rotation. The PCA allowed for the items to be freely converted into factors, as previous studies using the same audience involvement and celebrity worship scales yielded varying factor structures [7,19,23,24,28]. Moreover, previous studies often conducted factor analyses prior to path analyses to determine the factor structure and reduce the number of items [7,28,58,59]. The number of factors was determined by the eigenvalue. A factor was formed if the value was greater than 1. The analysis yielded three factors, the same amount and structure as proposed. The Kaiser-Meyer-Olkin (KMO) measurement of the sampling adequacy score was 0.837, and Bartlett's test of sphericity level was at p =< 0.001. The PCA reduced the items of audience involvement to 10 and 68.84% of the total variance is explained by the factors, higher than the required 50% [74].
Factor 1, named "behavioral involvement", comprised three items. All three items were from the original behavioral involvement dimension, with the other three items removed. A total of 20.53% of the variance is explained by the factor. Factor 2, "emotional involvement", consisted of four items from the original six, and 26.20% of the variance is explained by the factor. The last factor, "referential reflection", consisted of three items and explains 22% of the variance. Hence, it retained all three items from the original scale. In terms of the measure of scale reliability, the Cronbach's Alpha scores ranged from 0.763 to 0.819, exceeding the minimum threshold of 0.70 [74]. A summary of the PCA results of audience involvement is shown in Table 2. Similar to the audience involvement dimensions, the 22 measurement items of celebrity worship were also subjected to the PCA extraction method with varimax rotation. The results of PCA produced Kaiser-Meyer-Olkin (KMO) measurement of the sampling adequacy score was 0.948, and Bartlett's test of sphericity level was at p =< 0.001. Based on the same eigenvalue greater-than-1 rule, two factors (instead of the proposed three) were produced, and 62.62% of the total variance is explained by both factors. Factor 1, entertainment-social, consisted of eight items, all from the original dimension, albeit two items did not load onto the same factor. Factor 2, intense-personal, consisted of ten items. Seven items converged into this factor came from the intense-personal items, as originally adopted. In addition, two items from the proposed entertainment-social construct loaded onto this factor. Moreover, one item from the borderline pathological construct remained and loaded onto Factor 2. The result implies that the proposed borderline pathological construct did not have any of its items converge adequately to form a construct. Hence, only two constructs were retained for further analysis. A summary of the PCA results of celebrity worship is shown in Table 3.

Confirmatory Factor Analysis
According to Anderson and Gerbing (1988), a two-step approach to SEM analysis is recommended. The first step, confirmatory factor analysis (CFA), focused on the outer model and measured the fit indices, validity, and reliability. Only if the measurement model satisfied all requirements was hypothesis testing at the structural model level suggested [76]. The data were analyzed using IBM SPSS and IBM SPSS AMOS version 23. The measurement model showed a good fit (χ 2 = 1338.811, df = 917, p < 0.001, χ 2 /df = 1.560, RMSEA = 0.46, CFI = 0.943, IFI = 0.944). All the constructs also produced satisfactory composite reliability (CR) scores. The scores ranged from 0.765 to 0.933, all higher than the minimum thresholds of 0.7 [77]. The average variance extracted (AVE) scores produced by all the constructs were higher than the required 0.5 thresholds [78]. Thus, all the constructs achieved convergence validity. Next, discriminant validity was determined by comparing the correlation between a pair of constructs against a square root of the AVE. If the correlation value was lower, discrimination validity exists. In general, all the constructs were distinctive from each other. However, three pairs of correlations had higher correlations than the square root of AVE. Thus, a further step was required for discriminant validity verification by combining items from the pair of constructs into a single factor. If the measurement model with the combined items were statistically different than the original model, discriminant validity exists. The results of the single factor tests showed all three pairs of combined items to be significantly different than the original measurement model. Namely, emotional involvement and entertainment-social (∆χ 2 = 131.082, ∆df = 8, p ≤ 0.001), emotional involvement and intense-personal (∆χ 2 = 218.883, ∆df = 8, p ≤ 0.001), and entertainment-social and intense-personal (∆χ 2 = 364.918, ∆df = 8, p ≤ 0.001). A summary of the CFA results can be seen in Table 4.
The total variance explained scores revealed that Korean beauty product purchase intentions were strongly explained by the variance at 62.1%. In comparison, the intentions to visit Korea were explained by only 22.3% of the variance. Thus, implying that the research antecedents were more effective at predicting the intentions to purchase Korean beauty products than the intentions to visit Korea. Specifically, the two constructs of celebrity worship, entertainment-social and intense-personal, produced the highest R-square value in the proposed model, with 0.736 and 0.657, respectively. Lastly, brand awareness produced the largest total impact on both purchase intentions (0.689) and visit intentions (0.431). Table 5 summarizes the SEM results.

Indirect Effect Assessment
This study also analyzed the indirect effect between constructs to help further determine the mediating effect and the type of mediation. The indirect effect was calculated using the bootstrap method with 2000 samples at a 95% confidence level. Among the three independent variables, only behavioral involvement did not produce any indirect effect on brand awareness. Thus, indicating that, in the proposed model, the final constructs were not being predicted by behavioral involvement. In comparison, both entertainment-social was found to a partial mediator from emotional involvement and referential reflection to brand awareness. Brand awareness was found to be a partial mediator between entertainmentsocial and perceived product quality (β = 0.352, p =< 0.01). Lastly, perceived product quality also played a partial mediating role between brand awareness and the two final constructs. The indirect effect from brand awareness to purchase intention (β = 0.375, p =< 0.01) and visit intention (β = 0.192, p =< 0.01) were both significant. The summary of the indirect effect assessment is included in Table 6. The proposed conceptual model with SEM results can also be seen in Figure 2.

General Discussions
The first set of results were from the factor analysis of audience involvement. The three dimensions of audience involvement were retained, albeit some items were removed due to low loadings. Nevertheless, the remaining items converged strongly on their preconceived factors. The remaining items in the behavioral involvement factor suggest that favorite characters in the storyline were important to keep viewers engaged in the program. Similarly, the remaining items in the emotional dimension were also related to the individual characters in the programs. Lastly, all three items of referential reflection were retained. The finding implies that if the audience finds the shows to be relatable to their own experience, it is likely that they will exhibit a higher level of involvement. The importance of individual characters coincides with the relatively strong convergence of the celebrity worship dimensions. Although only two dimensions were retained after factor analysis, both factors contained largely the same measurement item sets from the original scale. Specifically, the entertainment-social factor contained 8 out of 11 items from the original set. On the other hand, the intense-personal contained seven of the original nine items, as well as two items from entertainment-social and one from the borderline pathological dimension. The absence of the borderline pathological implies that obsession with Korean celebrities is not severe enough that it would lead to illegal actions. Alternatively, it can be interpreted that female Korean entertainment fans do not display traits of extreme psychopathic parasocial relationships with their favorite celebrities.
Analysis of the proposed relationships has mostly been supported apart from two hypotheses. The first unsupported relationship was between behavioral intention and entertainment-social, given that the entertainment-social dimension refers to how fans enjoy consuming and discussing all aspects of their favorite celebrities with friends, including personalized content and the same celebrity in different roles or programs. Indicating that the audience follows certain characters in a show does not imply that they would also be interested in the actors or actresses' personal lives or their characters in other programs.
The second insignificant relationship was between intense-personal and brand awareness. It could be interpreted that if celebrity worshippers are intensely focused on the celebrities, the obsessions tend to be around the details of their personal lives or to have a parasocial relationship. A parasocial relationship is different than idolizing a favorite celebrity, in that idolizing indicates a level of influence in their attitudes, preferences, and behaviors. Therefore, a lower level of worship but still highly involved audience displays a stronger likelihood to pay attention to other elements in an entertainment program than those who mostly focus on the actors or actresses.
Overall, the proposed model has yielded strong predictive ability, with 62.1% of the variance explained by the respondents' Korean beauty product purchase intentions and 22.3% of their intentions to visit Korea. The discrepancy, or more specifically, the lower total variance explained by visit intention, was expected, due to the much higher costs and involvement to visit Korea than buying cosmetic or skincare products from Korea in Indonesia. One of the barriers for Indonesians to visit Korea is the visa requirement as well as the generally higher living expenses in Korea, making a trip to Korea not realizable for all. At the same time, Korean beauty products are available in various price ranges. Korean budget beauty brands are especially popular in Indonesia, given the performance to price ratio when compared to brands from Europe and North America. The results were also consistent with previous studies in many areas, such as the significant impact of involvement and celebrity on consumer intentions [22,31,44]. The brand awareness and perceived quality of Korean beauty products relationship were also significant, given the existing reputation and the acclaimed Korean beauty standards [11,12,57]. Lastly, the role of perceived quality and intention to buy Korean products or visit Korea considered the level of involvement of the participants with Korean entertainment media and their familiarity with Korean culture and visitor attractions in Korea that have often been portrayed in the media [2,8,11,12].

Implications
The findings of the present study can be beneficial to a number of interested parties. Apart from the obvious practical implications for destination management organizations, such as the KTO or other regional tourism management organizations in Korea, and private brands that are considering product placement on Korean entertainment programs, the findings can also be useful for travel agencies in Indonesia. With the popularity of Korean celebrities and entertainment programs in the country, travel itineraries containing activities related to popular programs can help attract a certain group of travelers. However, an overemphasis on Korean wave content may also alienate tourists with other types of interests. Also, the popularity of Korean entertainment programs and celebrities are constantly changing. Thus, entertainment-induced tourism itineraries must consider changes in consumer preferences. Furthermore, product placement does not need to be restricted to only Korean brands. For example, Indonesian brands looking to increase their product awareness or create a favorable perception of their product's quality may want to consider placing their products on Korean programs as well.
Theoretically, the study made contributions in a number of ways, including, firstly, the validation of the multi-dimensionality of both the audience involvement and celebrity worship. Unlike some previous studies that measured involvement using a single construct, the benefit of operationalizing audience involvement in this manner allowed for a precise examination of which specific type of involvement was most applicable to given research content. The successful adoption of celebrity worship illustrated the significance of distinguishing the difference between the media storyline and characters in them from the actors and actresses. Measuring audience involvement using a single dimension does not distinctively identify the subject of involvement. As a result, brands and destination organizations may not be able to identify the precise source of influence. In other words, some audiences may be influenced by the fictional characters in the programs but not necessarily the actors playing those characters. At the same time, there are also audiences who would focus on all elements of their favorite celebrities' lives. The study also highlighted the emerging role of Korean beauty standards and their influence on consumers abroad.

Limitations and Recommendations for Future Research
Limitations of this study include the generalizability of the samples. The study has only sampled a small group of the overall Indonesian Korean entertainment viewers. Notably, the sample profiles showed a large group to be students at lower-income level, and generally young. However, viewers of Korean entertainment may consist of other groups of Indonesians who are avid fans of Korean entertainment content that have been omitted from this study, such as male viewers, who still represent a significant group of the audience in Indonesia. Although male viewers may not be the primary buyers of Korean beauty products, the market for male skincare products has been growing in many countries. Thus, future studies should not disregard this group of the market. Moreover, there are still other types of products, notably food, fashion, and automobiles, that are often featured in Korean media. Hence, future studies can also investigate the role of entertainment media and celebrity worship in the context of other product types. Similarly, Korean entertainment is not only popular in Indonesia but in other parts of the world. Therefore, further validation of the conceptual model would not only help to increase the robustness of the proposed theoretical framework but also help with widening the external validity of the model in other countries and cultures. Lastly, a comparison study between local entertainment media and foreign entertainment media could also yield impactful results for brands and other stakeholders alike.