1. Introduction
Korea has been one of the most popular travel destinations for Chinese tourists. For instance, in 2014 and 2015, more than six and five million Chinese tourists visited Korea. During their stay in Korea, it is common for Chinese tourists to spend a significant amount of money shopping for various products. That is why it is important to understand Chinese-tourists shopping behavior and related product categories; what they buy, why they buy, where they buy, and how much they buy. One notable thing is that more than 84% spend money on purchasing cosmetic products (Lee, 2016). This number is almost double, compared to that of money spent on purchasing clothing (Park, Ko, Gu, & Choi 2014). Specifically, for Chinese tourists, authentic Korean cosmetic brands, such as Han-bang, Sulhwasoo, and so forth, have gained significant grounds against famous global cosmetic brands and have become as the most popular brands for Chinese tourists. Not only high-priced brands, but also a low-priced cosmetic brand such as Misha, has been one of the most frequently-purchased brands by Chinese tourists, which suggests that price itself may not be an critical issue for the Chinese tourist's choice of cosmetic brands.
Chinese tourists spend on average 2,130,000 Korean-Won on shopping during their stay in Korea and interestingly, nine out of ten Chinese tourist purchased cosmetics (Yonhap News Agency, 2016). This statistics provides evidence that Chinese tourists spend a lot of their money on purchasing cosmetics for their beauty and suggests a need for a systematic examination of such shopping behaviors with underlying motivations.
That said, the current study examines the effect of Chinese tourists' expectation on Korean cosmetics and corresponding performances on shopping satisfaction and the brand attitude toward Korean cosmetics brand. As such, the current study uses expectation-performance disconfirmation framework because we believe Chinese tourists would have to rely on their expectations based on the previous direct experiences or indirect experience of other tourists and word-of-mouth (WOM). They would then compare their expectations with the corresponding performance of Korean cosmetics. In addition, a moderating effect of consumer conformity between cosmetics selection attributes and expectation-performance is analyzed. Our study is very specific and unique in that we focus on tourist's shopping satisfaction and brand attitude on the basis of the disconfirmation paradigm, using product-related factors and environmental-related factors for cosmetic brand selection in Korea.
The current study contributes to a theoretical understanding of tourist shopping behavior in general, from consumer behavior perspective and also provides practical implications for distribution strategy in that relevant stockholders such as cosmetics manufacturers, wholesalers, and retailer need to reestablish new multiple channel structures (online and offline alike) to better target Chinese tourists, who visit Korea.
2. Theoretical Background and Conceptual Model Development
2.1. Conceptual Model for Chinese Tourist Purchase of Korean Cosmetics
Shoppers are often driven by motives; whether they are internal or external. These motives shape the purchasing behavior that consumers exhibit when they are purchasing a product or service. Rohm and Swanmithan (2004) define shopping motivation as the incentives that drive individuals to engage in purchasing behavior that satisfies their wants and needs. Understanding the motivation behind the shopping experience will be vital in predicting consumers and their purchase behavior. For instance, analysis of various motives such as online convenience, physical store orientation (i.e., immediate possession and social contact), information usage, and variety seeking has led to the establishment of four distinct shopping types: convenience shoppers, variety seekers, balanced buyers, and storeoriented shoppers. Tauber's (1972) study shows that shopping has multiple dimensions that marketers must consider such as the nature of the product, the perceived risk within the product class, and the amount of information regarding potential alternatives. To understand the shopping motives of consumers, we must consider the satisfaction consumers feel from the act shopping itself as well as the utility obtained from the purchased products.
With respect to cosmetics purchase, we would like to contend that physical attractiveness is one of the most important motivations for beauty products and cosmetics because physical attractiveness is, to certain extent, a natural and universal appeal to almost everyone. Previous research streams have confirmed that physical attractiveness is positively attributed to social power and self-esteem, while eliciting positive responses from others (Cann, Siegfried, & Pearce, 1981; Dion, Berscheid, & Walster, 1972). Bloch and Richins (1992) also contended that physical attractiveness is an inherently valuable attribute to all kinds of people. Consumers are likely to relate the use of aesthetically pleasing products with desirable outcomes, which is physical attractiveness from the use of cosmetics
However, to practically link the actual purchase of cosmetics to physical attractiveness, we need to further consider what would be potential marketing-related factors, generating favorable brand attitude, and eventually leading to actual purchases. Among so many potential factors, we contend that there are two types of factors in this study: (1) product-related factors such as brand image and perceived product quality, and (2) external environmental factors such as Korean Wave marketing and word-of-mouth. These factors, in turn, would influence two dimensions of disconfirmation paradigm (expectation and performance), shopping satisfaction, and brand attitude, directly or indirectly, which is shown in the conceptual framework in [Figure 1]. We also include consumer conformity as a moderator for the effect of two types of factors on disconfirmation because we believe the purchases are likely to happen when Chinese tourists as a travel group together, visit cosmetics retail shops, and thus group's conformity can significantly would influence disconfirmation, shopping satisfaction, and brand attitude.
[Figure 1] Conceptual Framework
2.2. Two Types of Factors of Korean Cosmetics Selection by Chinese Tourists
We propose four specific factors as antecedents that Chinese tourists would use for the purchase of Korean cosmetics during their stay in Korea, which is described at the following sections. As was mentioned above, we distinguish these four factors into two categories (product-related vs. external environmental factors). Productrelated factors used in this study are brand image and perceived quality, and external environmental factors are Korean Wave and word-of-mouth effect.
2.2.1. Brand Image
Brand image can be defined within a memory context, as individuals retain memories about a particular product and brand, both good and bad, which operates as a basic attribute of brand image (Reynolds & Gutman, 1984). Studies have explored the possibility of products having more social and personal meanings associated with the product apart from utility, although no general framework has been accepted that explains where these meanings derive from or how these meanings drive consumers during the shopping process. Well-established brand images are likely to create and sustain a strong brand loyalty from consumers as well as increase brand equity. Thus brands with higher levels of brand image have a better chance to expand within the market, whereas brands with low brand image will find difficulty in expanding their brand (Edwin & Clara, 2005).
In terms of evaluating products before their purchase, a problem for Chinese tourists is the possibility that they may have to depend on brand image, instead of actual product quality, given that such tourists may not be familiar with foreign cosmetics product's actual quality. Another important note is that Chinese tourist's purchase of Korean cosmetics during his/her stay in Korea is likely to be just a transactionspecific, one-time purchase. This situational consideration may provide another evidence that foreign tourists may have to rely on brand image, rather than actual product quality when it comes to purchase decision for Korean cosmetics. Then brand image is hypothesized to influence expectation-performance framework. At the same time, brand image is likely to influence shopping satisfaction.
2.2.2. Perceived Quality
Perceived quality can be considered as consumers' impressions of a product's capability to satisfy their needs and expectations (Olshavsky, 1985). Bettman and Park (1980) defined the perceived quality as invisible emotions toward a product or brand. There are multiple factors that influence the perceived quality of a brand's product. For instance, a brand's public image can affect a consumer's perception of the brand's products without knowing the product's real quality and/or functionality. Consumers can be influenced by the opinions and suggestions of opinion leaders or even their peers. Overall, perceived quality is the expectation consumers place on the product's ability to fulfill their needs (Olshavsky, 1985). Interestingly, perceived quality may have little or nothing to do with the actual performance/ quality of the product, and could be based only on the firm's (or brand's) current public image (see corporate image) and consumer's experience with the firm's other products. Yet, the perceived quality should be one of the most important factors for purchase decision and it will significantly influence expectation and performance of disconfirmation paradigm, and eventually influence shopping satisfaction and brand attitude. Kim and Lee (2016) also propose a similar argument, contending that perceived quality affects significantly brand attitude, brand image, and price.
2.2.3. Korean Wave
The Korean Wave phenomenon refers to the increase in popularity for Korean entertainment and culture across the world. In 1990's, Korean dramas and pop songs gained a surprising amount of popularity among the Chinese population, particularly in Beijing, China. As the wave consistently stirred popularity for Korean entertainment, the enthusiasm for Korean culture expanded into food, literature, language, clothes, and beauty industries.
Over time, the Korean Wave developed from a regional occurrence to a global phenomenon. This new trend from Korea influences many countries, not only China, but also Southeast Asia, South America such as Brasil and Argentina, and even several European countries including England and France.
Its impact through Korean pop music (K-pop) expands to diverse areas including beauty industry, under the name of K-beauty. The Korean Wave significantly and substantially influences the sales volumes and market share of Korean cosmetics and personal care products in China. For instance, US Department of Commerce published a report in 2016 regarding Asian personal care and cosmetics market guide, and it shows that among imported cosmetics and personal care products in China, the market share of Korean products and brands is 30% in 2015, compared to 9% in 2013. The Korean Wave is indeed a large factor influencing the purchase of Korean cosmetics for Chinese tourists. If this is the case, we contend that most Chinese tourists would have been exposed to the Korean Wave before they visit to Korea, and we propose the Korean Wave as an external factor significantly influencing the purchase of Korean cosmetics during their visit to Korea. Thus, Korean wave marketing effort will significantly influence the two dimensions of disconfirmation framework.
2.2.4. Word-of-Mouth (WOM)
Word-of-Mouth or WOM is the spreading of knowledge or opinions on products and services by consumers through an unofficial communication method (Anderson, 1998). Numerous studies by many different researchers have examined WOM (Herr, Kardes, & Kim, 1991; Liu, 2006; Swan & Oliver, 1989). For instance, Herr, Kardes, and Kim (1991) investigated the mediating effect of WOM on persuasion with several moderating variables. They showed that WOM affects consumer judgment when participants were presented with positive attributes. Obviously, WOM effects have been proven as one of main factors influencing persuasive communication process and eventually purchase decision itself.
In this study, we also include WOM as an external factor influencing shopping satisfaction, brand attitude and disconfirmation, because we believe Chinese tourists are highly likely to be dependent on peer evaluations of travel group members, friends, or family, given that they do not know much about Korean cosmetics before they visit Korea. Even while in the Korea, it is reasonable to assume that most Chinese tourists would actively depend on the recommendations and word-of-mouths from the travel group members, from online communities or social network systems in China.
2.3. Expectancy-Disconfirmation Theory
Given the importance of customer satisfaction in marketing theory and practice, so many studies have been done to conceptualize and measure customer satisfaction. One of such efforts is the development of disconfirmation theory or paradigm. Oliver (1980) proposed a cognitive model to capture the antecedents and consequences of satisfaction decision. The proposed model defines customer satisfaction as a function of expectation and expectancy disconfirmation of perceived performance. Customer satisfaction is simply defined as the gap between customer's expectation and performance of products or services. For example, if the performance of product exceeds the customer's expectation, the customer is believed to be satisfied with the product. In other words, performance confirms or disconfirms the existing expectation that customer may have developed previously. That is why the theory is labeled as disconfirmation theory.
Oliver (1980) also shows that satisfaction based on disconfirmation influences attitude changes and purchase intentions. Tse and Wiltion (1988) examined the effects of perceived performance on consumer satisfaction literature by analyzing alternative conceptualizations of comparison standards and disconfirmation in satisfaction formations, which captures the standard formation process as well as the multiple comparison process.
We contend that the above-mentioned four factors would significantly influence these two dimensions of disconfirmation as key mediators, which will lead to shopping satisfaction and brand attitude toward Korean cosmetics.
[H 1] Brand image, perceived quality, Korean wave, and word-of-mouth, respectively, are positively related to expectation.
[H 2] Brand image, perceived quality, Korean wave, and word-of-mouth, respectively, are positively related to performance.
2.4. Shopping Satisfaction
Customer satisfaction has been a central theme in marketing and has been researched in almost every corner of marketing-related entities, such as customers, employees, stakeholders, investors, and so on. For example, Anderson and Sullivan (1993) have identified antecedents and consequences of customer satisfaction. Bearden and Teel (1983) studies customer satisfaction in relation with employee responses. Jansen-Verbeke (1991) listed factors influencing shopping satisfaction of tourists such as product and its image, store's opening time, product value, easy to approach, employee service, price, shopping environment, atmosphere, and exciting historical facts. Shin, Hwang, Lee, and Cho (2015) examined the effect of service quality on customer satisfaction for franchise coffee shop in Korea.
Shopping satisfaction is the pleasure consumers experience that results from shopping. The satisfaction can be assessed by comparing the expectation of the shopping experience and the actual shopping experience (Babin & Griffin, 1988). When consumers are satisfied with their shopping experience, the positive attitude towards a product or brand will likely elicit repurchasing behavior whereas if consumers are not satisfied, a negative attitude will form which prevents repurchase behavior (Machleit & Mantel, 2001; Turner & Reisinger, 2001). Word-of-mouth (WOM) follows a similar behavioral pattern. When consumers are satisfied by shopping, positive WOM will increase. Conversely, when consumers are not satisfied, there will be an increase of negative WOM (Anderson, 1998). In this study, we hypothesize that the four factors directly and positively influence shopping satisfaction, and additionally, via expectation and performance, indirectly influence shopping satisfaction.
[H 3] Brand image, perceived quality, Korean wave, and word-of-mouth, respectively, are positively related to shopping satisfaction.
2.5. Brand Attitude
Brand attitude is an opinion consumer has toward a product or service associated with a brand (Keller, 1993). Based off the behavioral traits of consumers, understanding brand attitude has become an integral part of marketing, considering the importance of customer-based brand equity, which can be generated by brand element, marketing mix (price, product, promotion, and place), or the leverage of secondary associations (Keller, 1993). We argue that the most influential and direct factor would be satisfaction with product itself to elicit favorable brand attitude. Thus, we hypothesize that four factors of brand image, perceived quality, Korean wave marketing, and word-of-mouth would significantly influence brand attitude. In addition, two dimensions of disconfirmation would influence shopping satisfaction and brand attitude directly. Finally, we contend that shopping satisfaction will positively influence the formation of favorable brand attitude. Based on these arguments and literature, the following hypotheses are proposed:
[H 4] Brand image, perceived quality, Korean wave, and word-of-mouth, respectively, are positively related to brand attitude.
[H 5a] Expectation is positively related to shopping satisfaction.
[H 5b] Expectation is positively related to brand attitude.
[H 6a] Performance is positively related to shopping satisfaction
[H 6b] Performance is positively related to brand attitude.
[H 7] Shopping satisfaction is positively related to brand attitude.
2.6. Consumer Conformity
In marketing, Burnkrant and Cousineau (1975) defines conformity as tendency of individuals to abide by a group or social norm (i.e., a set of group expectations on how members should behave). Conformity can be influenced by several sub-dimensions such as task, situation, brand, group, and personal characteristics (Murphy & Witt, 1975). Lascu and Zinkhan (1999) specifically provide a review and applications for marketing, treating consumer conformity as one of main factors to understand consumer's decision making process.
We propose that consumer conformity could play a big role for the purchase of Korean cosmetics. Most Chinese tourists can't correctly evaluate Korean cosmetics products and brands because they are not familiar with such brands. Therefore, it is natural for Chinese tourists to follow group consensus or evaluations about the Korean cosmetics brands they might consider to purchase. In particular, this conformity will interact with four cosmetics selection factors described above to influence expectation and performance of disconfirmation.
[H 8] Consumer conformity is moderating the effects of brand image, perceived quality, Korean wave, and word-of-mouth on expectation.
[H 9] Consumer conformity is moderating the effects of brand image, perceived quality, Korean wave, and word-of-mouth on performance.
3. Methodology
3.1. Data Collection
To test the proposed model, we conducted a survey with 250 Chinese tourists who had visited Jeju, Korea, from February 15 to February 29, 2016. Out of 250 tourists, 43 respondents' inputs were excluded from the final analysis, because of incomplete answers, missing values, or inconsistent responses. Thus, total 207 responses were used in the final analysis.
Survey participants' demographics are shown in [Table 1]. There were 40 men (19.3%) and 167 women (80.7%) who responded to the survey. 46.4% of the respondents were in their 20s, 29% in their 30s, and 24.6% of the Chinese tourist who responded to the survey were 40 years or older. More than 30% of the respondents were students, and 23.2% of them were office workers. More than half of the tourists who participated in the survey had monthly income of 2 million to 3 million Korean Won.
[Table 1] Demographic Characteristics
3.2. Measurement
Each construct are measured based on the combinations of existing multi-item scales and newly added items that we created to better fit our survey context. In addition, if necessary, we modified the original scales to fit the context for Chinese tourists purchasing Korean cosmetics. All the items are based on 1-7 point Likert scales.
Brand image was measured based on the following five items: others' evaluations for the brand, brand reputations, brand awareness, brand trust, luxurious image of the brand (Biel, 1992; Faircloth, Capella, & Alford, 2001). Perceived quality is based on five items such as functionality, properness of skin care, container design, price compared to quality, overall product quality (Aaker & Jacobson, 1994; Mitra & Golder, 2006). The Korean Wave was measured by five items: K-pop star's commercial ad, Effect of K-drama, image for Korea, sophistication of Korean cosmetics, whether the brand is actually used by K-pop stars. We measured WOM using five items such as SNS recommendation, reply to Internet positing, friend recommendation, actual purchase by peer group, general people's consumption or purchases (Trusov, Bucklin, & Pauwels, 2009).
Expectation was measured by eight items such as expected reputation, trust expectation, brand awareness, functionality, price discount expectation, quality service expectation, Korean image expectation, price/quality ratio expectation (Aaker, 1997). Performance was measured by similar items used for expectation such as exceeding reputation, better than known quality, higher credibility, superior functionality, excellence of skin care, more favorable price discount, impressive store service, and gift purchase for friends (Aaker, 1997).
Shopping satisfaction was measured based on JansenVerbeke (1991) and Heung and Cheng's (2000) studies, such as satisfaction with price, product quality, convenience, store facilities, product assortment, and customer services. Brand attitude was measured by brand familarity, differentiation, uniqueness, attribute expression, quality, and spreading words for brand (Faircloth et al., 2001). Finally, consumer conformity was measured using seven items such as group identification, compliance, internalization, opinion listening, information share, positive conformity, and behavior conformity (Bearden & Rose, 1990; Lascu & Zinkhan, 1999).
All the items were first translated into Chinese and then back-translated into English to make sure correct measures of constructs in this study. We revised and refined the final questionnaire based on this back-translation to enhance face validity of our survey questionnaire.
4. Results
4.1. Reliability and Validity Check
An exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were conducted to select test items prior to the reliability and validity check of concepts and scales presented in this research model. Cronbach‘s α was used to check the reliability of the measured variables. The results of the CFA showed an acceptable fit. To check the convergent validity, we calculate the composite reliability (CR) and average variance extracted (AVE). The results were found in [Table 2]. The Cronbach‘s α of each variable was higher than 0.70, confirming the internal consistency of measures (Hair, Black, Babin, & Anderson, 2010).
[Table 2] Reliability and Validity Check, Model Fit Indices
KMO and Bartlett's test : KMO=.862 X2 =10273.641 df=1540, p=0.000
X2 =1863.826, df=1420, X2 /df=1.313, RMR=0.065, RMSEA=0.039, GFI=0.867, NFI=0.836, CFI=0.955
The results provide another evidence of the convergent validity based on AVE of all constructs as well as composite reliability (Hair et al., 2010). The lowest composite reliabilities is 0.896 for performance construct and all others are above 0.9. In addition, all the AVEs for constructs are above 0.5, which provides evidence of convergent validity.
To check the discriminant validity of the constructs, we checked whether the square root of the AVE for each latent construct is higher than any of the bivariate correlation coefficient involving the latent construct (Fornell & Larker, 1981). [Table 3] shows descriptive statistics for each construct with correlation coefficients among constructs. Also [Table 3] shows AVE on the diagonal elements for easy comparisons between bivariate correlation coefficients and AVEs. All the square root of the AVEs are higher than bivariate correlation coefficients, which provide evidence of discriminant validity.
[Table 3] Descriptive Analysis, Correlation, and AVE
* p < .05; ** p < .01
4.2. Hypotheses Testing and Results
Measurement properties including reliability, convergent validity, and discriminant validity of the constructs were all confirmed. Now, hypotheses were tested using structural equation model (SEM). The SEM results indicated a reasonable fit (X20.05(846) = 982.514, IFI = .982, GFI = .831, AGFI = .803, NFI = .885, CFI = .982, RMR = .028, and RMSEA = .056). Even though X2 value is significantly large, but its value is sensitive to the number of respondents; it may not be a big problem at all (Bagozzi & Yi, 1988). Other indices are acceptable and appropriate.
That said, the results of the hypotheses testing were shown in [Figure 2]. The solid lines indicate significant relationships between constructs, and the dotted lines indicate non-significant relationships. Specifically, the SEM results show that brand image is positively influencing expectation, performance, and brand attitude, but not shopping satisfaction. Perceived quality showed significant positive influences on performance, shopping satisfaction, and brand attitude. The Korean Wave and related marketing efforts influences positively all the endogenous constructs in the model. In particular, the Korean Wave has a largest impact on shopping satisfaction (.373) among other constructs. Lastly, the result showed that WOM only influenced expectation, but not other constructs. It makes sense because Chinese tourist expectations are likely built on previous tourist's WOM, given that they are new to Korea or they are not familiar with Korean cosmetics.
[Figure 2] SEM results
Expectation and performance construct in turn positively and significantly influenced shopping satisfaction and brand attitude respectively. Finally, as hypothesized, shopping satisfaction is significantly influencing brand attitude.
4.3. Moderating effect of Conformity
The moderating effects of consumer conformity on expectation and performance were examined by the relationship between the four factors of Korean cosmetics selection (brand image, perceived quality, the Korean Wave, and WOM) and two subdimensions of disconfirmation (expectation and performance). Thus, there are eight different regression models with consumer conformity as a moderator. For example, using PROCESS model, we set up a regression equation as follows: expectation is a dependent variable, brand image is independent variable, consumer conformity is a moderator, and the interaction term (brand image x consumer conformity) is a covariate in the model using PROCESS software by Hayes (2013). We applied the same setup for all the remaining factors for expectation and performance as dependent variables.
For the first part of moderation analysis where expectation is a dependent variable, as is shown in [Table 4] to [Table 7], the effects of all the Korean cosmetics selection factors (brand image, perceived quality, Korean wave, and WOM) were significantly moderated by consumer conformity. Each interaction term (Brand image*conformity; perceived quality*conformity; Korea Wave*conformity; WOM*conformity) shows significant results. For example, the coefficient value of brand image*conformity terms is 0.237 (p<.05). Similarly
[Table 4] Moderation Effect of Brand Image*Conformity on Expectation
*p<0.05, **p<0.01, ***p<0.001
[Table 5] Moderation Effect of Perceived Quality*Conformity on Expectation
*p<0.05, **p<0.01, ***p<0.001
[Table 6] Moderation Effect of Korean Wave*Conformity on Expectation
*p<0.05, **p<0.01, ***p<0.001
[Table 7] Moderation Effect of WOM*Conformity on Expectation
*p<0.05, **p<0.01, ***p<0.001
For the second part of moderation analysis where performance is a dependent variable, as is shown in [Table 8] to [Table 11], the effects of brand image and word-of-mouth were significantly moderated by consumer conformity, but not for Korean Wave and WOM.
[Table 8] Moderation Effect of Brand Image*Conformity on Performance
*p<0.05, **p<0.01, ***p<0.001
[Table 9] Moderation Effect of Perceived Quality*Conformity on Performance
*p<0.05, **p<0.01, ***p<0.001
[Table 10] Moderation Effect of Korean Wave*Conformity on Performance
*p<0.05, **p<0.01, ***p<0.001
[Table 11] Moderation Effect of WOM*Conformity on Performance
*p<0.05, **p<0.01, ***p<0.001
Overall, the result showed that there is a significant interaction effect of conformity on expectation for all four selection factors. We also found a significant interaction effect of conformity on performance for two selection factors of brand image and WOM.
5. Discussions
The results of the current research are summarized as follows. First of all, we found that expectation and performance have a significantly positive effect on shopping satisfaction ([H5]) and brand attitude ([H6]), as was hypothesized. These results provide evidence that the current disconfirmation approach plays an important role for understanding shopping satisfaction and consequently brand attitude. Both are influencing on shopping satisfaction more than on brand attitude and it is no surprise because our model is based on disconfirmation. This result is consistent with previous empirical studies from Oliver (1980), Westbrook (1981), Westbrook and Reilly (1983), and Tse and Wilton (1988), and so forth. As hypothesized in ([H7]), shopping satisfaction is significantly influencing brand attitude, which is also consistent with previous research. This finding implies and confirms the importance of satisfaction on attitude and managers should strive to enhance customer's satisfaction as one of the most critical factors to monitor.
The hypotheses regarding the effects of four factors on expectation ([H1]) and performance of disconfirmation ([H2]) are shown in the [Figure 2]. The SEM results show that brand image is positively influencing expectation and performance. In addition, brand image does influence significantly brand attitude, but not shopping satisfaction, which suggests a partial mediating effect of expectation and performance on shopping satisfaction and brand attitude. That said, the results of ([H3]) and ([H4]) with respect to the effect of brand image provide an important implication in that brand attitude as a proxy for actual behavior can be inferred by existing brand image, but brand image is not directly influencing shopping satisfaction. Managers need to pur efforts on more significant and relevant factors such as perceived quality than brand image.
Perceived quality showed significant positive influences on performance, shopping satisfaction ([H3]), and brand attitude ([H4]), but not on expectation. This finding should be reexamined in the longitudinal study of the current context, because the perceived quality may not be directly related to expectation given that some or most Chinese tourists may not have direct experiences of Korean cosmetics before their visit to Korea, which suggests a possible insignificant correlation between the two. It suggests a very critical implication; when one applies disconfirmation approach with customers who may not have developed perceived quality for certain products and services, the expectation level may not correctly reflect the perceived quality.
As was described in the result section, Korean Wave as a cultural and environmental factor influences positively all the endogenous constructs in the proposed conceptual framework, including expectation ([H1]) and performance ([H2]). One thing to notice is that Korean Wave has a largest impact on shopping satisfaction ([H3]). This finding confirms that the effect of Korean Wave on Chinese is prevalently evident in many business areas such as K-pop, K-drama, K-beauty, K-fashion, and even K-food. In this sense, it is no surprise to see the impact of Korean wave on expectation, performance, shopping satisfaction, and brand attitude, which implies that Korean firms and managers should continuously put efforts on Chinese as a main target using cultural mega-trends like Korean wave.
The SEM result also showed that WOM only influenced expectation, but not other constructs, which suggests a partial mediation effect of WOM on shopping satisfaction and brand attitude through expectation and performance. It makes sense because Chinese tourist expectations are likely built on previous tourist's WOM, given that they are new to Korea or they are not familiar with Korean cosmetics. In sum, important factors related to cosmetics evaluation and selection process by Chinese tourists had positive effects on shopping satisfaction and brand attitude, directly and/or indirectly via two dimensions of disconfirmation, which are expectation and performance and via partial mediation. The results are again consistent with prior studies from Anderson and Sullivan (1993), and Heung and Cheng (2000).
Lastly, we want to address an important moderating effect of consumer conformity, providing significant implications. Specifically, the moderating effect of conformity on expectation ([H8])and performance ([H9]) are prevalent. Brand image and WOM were found to interact with consumer conformity for both of expectation and performance. The implication is that given the possible unfamiliarity with Korean cosmetics brand for Chines tourists, WOM becomes more important and relevant because it is likely to influence significantly brand image itself as well, and conformity will play a big and significant role for the effective WOM and brand image building. Considering the nature of group travels in Korea by Chinese tourists, WOM is likely to be examined within the group behavior context of Chinese tourists rather than that of individual behaviors. Therefore, the moderating effect of conformity on WOM becomes a critical factor that managers should explicitly consider and actively use it as an important marketing tool.
The overall implications of this research are as follows. First, the distribution channel should be able to specify what traits of cosmetic products need to be marketed. There are incentives to establish and expand services that cater to the Chinese tourists. For instance, Nam (2017) explored success factors of night markets in Korea and emphasized its significant influence on the related industries and potential customers. We believe such night markets would play an important role as a tour destination and as a distribution retail outlet targeting foreigners, particularly Chinese tourists.
Second, marketers need to develop a strategy that consistently increases brand image directly and/or indirectly via WOM. Corporations will have to consistently invest in raising quality of cosmetic products. Utilizing the Korean Wave trend, firms must increase positive WOMs to create more exposure of Korean culture, beauty, entertainment, clothes, and so on, to attract more Chinese tourists.
Third, it is very important to find a way to form an in-bound market in China for cosmetics, because for now, an out-bound cosmetic market is thought to be better. However, it is a matter of time when Chinese tourist will turn their attention to in-bound market for cosmetics. It is important to increase the profit in Korean market for Chinese tourists, but it is more pivotal to find a way to form a domestic market for cosmetics in China. Lastly, one notable cultural aspect of Chinese consumers is that they make their decision among the group. Most of the Chinese tourists in Korea are group tourists with around 20-30 people in each group. Thus, it is important to provide a promotion strategy that will increase group purchase intention among Chinese tourists, thus enhancing WOM and group conformity simultaneously.
6. Conclusion
The current study shows that Chinese tourist expectation and perceived performance had positive impact on their satisfaction level, which in turn positively influenced brand attitude. Four cosmetic selection factors had positive effects on shopping satisfaction and brand attitude. Therefore, using the disconfirmation paradigm, we provide a theoretical understanding of Chinese tourist's cosmetic purchase behavior in Korea, which can be a basis for generalizing the current framework to other countries. Importantly, consumer conformity was found to significantly influence expectation and performance. In particular, when tourists travel together as a group, conformity can play a big role for brand selection process and related behavioral outcomes such as shopping satisfaction and brand attitude via expectation and performance.
While the current research provides insights from a consumer behavior perspective, it also provides an important guideline for establishing multi-channel structures using both on-line and off-line channels. Korean cosmetics firms need to better understand cosmetics selection attributes and preference of Chinese tourists, which provide practical guidelines to develop retail stores and distribution outlets for Chinese tourists.
7. Limitations and Future Research
While the current study provides important managerial implications for Chines tourist's purchase pattern for Korean cosmetics, the current research is not without research limitations. The study is based only on Chinese tourists who visited Jeju, instead of Korea itself, which suggests the respondents may not be a representative sample. Thus, a future study need to collect data from different places/cities in Korea. A similar limitation can be applied to product category; we have used only cosmetics in Korea, not other product categories at all. Thus, a future research needs to include different product categories as well as service products.
Given the rapid change in economic environment and technology, and more importantly, diverse and changing characteristics of Chinese tourist, they may not find strong incentives to visit Korea, nor want to purchase Korean cosmetics brands. In other words, this kind of strong trend could be temporary, which suggests that a longitudinal research needs to be done.
For the current study, the sample is very skewed and dominated by female tourists. Thus, the study results may not be generalizable unless the sample is balanced according to important demographics such as gender, age, and education
The measurements along with operationalized definitions of constructs used in the current study may be perceived as too broad, which could be misleading. For example, brand attitude is not based on reflective measures such as likeability and preference of brand used in other studies. Thus, more thorough examinations of operationalized definitions of constructs are necessary to be more reliable and valid. In addition, brand image can be a construct representing an overall post-consumption image, but in the study, brand image was used as a pre-image influencing brand attitude through disconfirmation framework. In this light, a longitudinal study measuring brand image over time will be a good future study. With respect to the dimensions of disconfirmation paradigm, not only expectation and performance, but disconfirmation of belief may be included in the measurement to accurately reflect the true level of disconfirmation.