Does the Endorser ’ s Nationality Matter ? An Investigation of Young Taiwanese Consumers ’ Selecting Smartphone

The use of a wrong celebrity endorser may boost the animosity of local consumers toward a specific country with a complex consumer animosity. Thus, this study aims to examine whether the match or mismatch between the endorser’s nationality and the brand’s country of origin affects consumers’ purchase intention. We use an experimental study to investigate the purchase intention of young Taiwanese consumers toward Korean smartphone brand “Samsung”. Results indicate that the endorser’s nationality does matter. While celebrity endorsement has a consistent positive effect on purchase intention, the effect of the endorser’s nationality varies. Specifically, using domestic celebrity as an endorser has more advantage in enhancing the purchase intentions of consumers who are neutral and dislike Korea than using a foreign celebrity. However, an endorser’s nationality has no significant difference among young Taiwanese consumers who like the Korean culture. Smartphone producers such as Samsung and LG can use the results to promote their new products in a foreign country such as Taiwan and Japan with a complex consumer animosity.


Introduction
Endorsement is one of the most frequently used methods in advertising to influence the attitudes and purchase intentions of consumers (e.g., Choi & Rifon, 2007;Ding et al., 2011;Kamins, 1990).Research has shown that the characteristics of advertisement endorsers effectively convey product imagery to consumers by creating a positive connection between consumers and the reputation of the product/brand represented by the endorsers (Wang & Doong, 2010;Kim et al., 2013;Kim et al., 2014).A good match between an advertisement endorser and a brand elevates brand equity by promoting brand image, improving consumer attitudes, and increasing consumer purchase intentions of the brand (Kamins, 1990;Kim et al., 2013;Koernig & Boyd, 2009;Liu et al., 2007;Marshall et al., 2008;McDaniel & Heald, 2000;Till & Busier, 1998, 2000).
By contrast, a mismatch between an endorser and a brand may produce negative effects, such as consumer mistrust generated by the association between the endorser and the brand, deterioration of the brand image, loss of customers, and loss of profitability (Koernig & Boyd, 2009;Marshall et al., 2008;Till & Busier, 1998, 2000;Kim et al., 2013).Therefore, international brand managers often spend considerable time and financial resources to select the advertisement endorsers that will have a good match between their brand and the endorser characteristics.Occasionally, endorser characteristics such as nationality may determine the success or failure of brand development in foreign markets with a complex consumer animosity (Zhang & Zhang, 2010).
For international brand managers, determining the most appropriate endorser to increase the effectiveness of an advertisement in promoting a product and brand in an international market is vital.They usually adopt either a between brand image (American brand) and the nationality of the celebrity endorsers (American celebrity endorsers).
However, using endorsers from the brand's country of origin may have serious disadvantages if a previous or ongoing political, military, economic or diplomatic conflict arouses the anger of some people in one country toward a specific country (Klein et al., 1998).The disadvantages are serious, particularly when many domestic consumers have negative attitudes or even a complex animosity toward the brand's country of origin.Although consumers who are angry toward a specific country may not believe that its brands or products are inferior, consumers may have less purchase intention toward the brands or products of that country (Shoham et al., 2006).For example, using Japanese endorsers to advertise Japanese brands, such as Sony, Toyota, and National, may not be a brilliant idea in China because of the long history of consumer animosity between Japan and China.The use of a Japanese celebrity endorser may boost the animosity of Chinese consumers toward Japan, which is separate from their assessment of product quality (Klein et al., 1998).
In this case, some international brand managers recruit domestic celebrities as endorsers to advertise their brand.The localization of advertising strategy for an international brand is common to avoid the negative effect of consumer animosity.The use of domestic celebrity endorsers in advertisements may attract more domestic customers, and products can receive a positive evaluation regardless of animosity levels.For example, Samsung, a Korean brand, used several Taiwanese celebrity endorsers, including Jolin Tsai, Dee Hsu, and Fu-Zhen Tian, to endorse its serious smartphone products (Galaxy, S4, S3, and S2) in Taiwan (Table 1).However, to promote new Galaxy S5, Samsung used Soo-Hyun Kim, a famous Korean drama actor, as the brand endorser in the global market.Thus, the effects of using a domestic celebrity as an endorser have not been discussed extensively due to its cultural complexity and changeability.Although the localization of the advertising strategy has some advantages to attract domestic customers, the mismatch between domestic endorsers and an international brand may produce negative effects, such as loss of the positive effect of the country of origin (COO) and the opportunity to attract some customers who like the COO of the brand.Currently, more and more Taiwanese teenagers are into Korean drama and pop singers.These young consumers may follow the consumption behaviour of Korean endorsers.This attitude may explain why Samsung Taiwan changed its new endorsement strategy and used Korean endorser Soo-Hyun Kim to endorse its new Galaxy S5 smartphone in Taiwan (Table 1).However, the advantage and disadvantage of the match and mismatch between brand and the endorser's nationality remains ambiguous.Therefore, understanding the role of the endorser's nationality in advertising a Korean smartphone among young Taiwanese consumers is necessary.
To resolve the globalization-localization of the advertising debate, this research investigates the responses of young Taiwanese consumers to Korean brand advertisements.This study provides a theoretical framework to understand the role of the nationality of the endorser in the purchase intention of the consumers toward a product of a foreign country.The rest of the paper is organized as follows.The succeeding section presents the research background of this study, followed by the discussion on the conceptual framework and proposed hypotheses.The research design and the result of data analysis are discussed in the subsequent two sections.Finally, the discussions, findings, implications, and limitations are given in the last section.

Celebrity Endorser
Endorsers can be of many types, including the typical consumer, the product class expert, the company president, and the celebrity (Friedman & Friedman, 1979;Kamins, 1990).In particular, the use of celebrity endorsers has always been a favourite strategy because the credibility of celebrity endorsers is effective in generating attitudes toward the advertisement and the brand (Bhatt et al., 2013).Prominent celebrity endorsers produce more positive attitude changes toward the position advocated by the consumers and induce more behavioural changes than less popular celebrities (e.g., Felix & Borges, 2014).Moreover, popular celebrities may be more effective in gaining attention and enhancing recall than less known endorsers (Goldsmith et al., 2000).
The use of celebrity endorsers in advertising also has an important role in the transfer of personality traits (Erdogan et al., 2001).For example, Shimp and his colleagues (2005) found that marketers of consumer packaged goods often feature endorsers as the centrepiece of their advertising or promotion efforts (Shimp et al., 2005).Previous researchers mainly focused on exploring the effect of image transfer from the endorser to the brand on product or service characteristics (Huber et al., 2013;McCracken, 1989).This effect may explain the reason why most international brand managers like to use celebrity endorsements than other types of endorsers.Therefore, instead of examining all types of endorsers, the current study focuses on the use of celebrity endorsers.

The Positive Effect of Celebrity Endorsement on Purchase Intention
According to Freiden (1984), a celebrity endorser is any publicly well-known individual who uses his/her reputation and charm to attract the attention of consumers.Celebrities in advertisements may generate the favourable feelings of consumers toward products and/or brands promoted by the celebrity endorsers (Felix & Borges, 2014).The celebrity endorser also contributes substantial positive effects on the financial returns of companies (Erdogan et al., 2001).Celebrity endorsement has been investigated by many previous studies as a common and effective strategy to promote brands and/or products (e.g., Freiden, 1984;Kamins, 1990;Hollensen & Schimmelpfennig, 2013).For example, Hollensen and Schimmelpfennig (2013) concluded that the use of celebrity to endorse products has a positive effect on purchase intention.However, only a few studies discussed how the endorser's nationality influences the perception of consumers on the advertised brand (e.g., Zhang & Zhang, 2010).
Nowadays, celebrity endorsement advertising has become part of daily life.Consumers see various celebrity endorsers with different nationalities on TV, the Internet, newspaper, fliers, and magazines because international brand advertisers and marketers use celebrities of different nationalities to attract different market segments.Prominent celebrity endorsers produce more positive attitude changes toward the position advocated and induce more behavioural changes than less popular celebrity (e.g., Bhatt et al., 2013;Felix & Borges, 2014).Compared with no celebrity endorsement, the use of celebrity endorsers has a positive effect on purchase intention.In fact, an increasing number of international companies, such as Nike, Coca-Cola, Samsung, and Sony, use celebrity endorsers in their own advertisements.

Consumer Animosity between Taiwan and Korea
Consumer animosity is "a previous or ongoing political, military, economic or diplomatic event" that causes the anger of some people in a country toward a specific country and influences the purchase behaviour of consumers in the international market (Klein et al., 1998).Animosity has two main dimensions: War-based animosity and economic-based animosity (Wang et al., 2013).Previous literature shows that "hostility and anger towards one specific foreign country (animosity) can adversely impact on the consumption of products from that country" (Klein et al., 1998;Klein, 2002), unrelated to consumer judgments on product quality or the valuable attributes of the product such as price and quality.For example, consumers in China avoid Japanese products not because they do not like the products but because the history of hostility between China and Japan that originated from the terrible occupation of Japan in China during World War II (Klein et al., 1998).
This study aims to investigate the effect of economic-based consumer animosity between Taiwan and Korea.Economic animosity is the result of the threat of economic advantage or aggression from one country to another and leads to an attitude of animosity (Klein et al., 1998).Like war-based animosity, economic-based animosity has a direct and negative effect on the purchase behaviour of consumers.For example, Nijssen and Douglas (2004) prove that Germany, plays a dominant role in the European monetary system and the EU market, and this heightens Dutch sensitivity to German economic power and the potential threat that it may represent.Recently, Taiwanese and Korean companies are highly competitive in terms of their respective smartphone, semiconductor, computer, textile, electronic, and consumer electronics industries.This competition may arouse the patriotism among citizens and cause them to generate negative attitudes toward the "aggressor" (i.e., Korean brand products) and their unwillingness to purchase those (Nijssen & Douglas, 2004).
Understanding the role of consumer animosity can help marketing managers to better understand the attitudes of domestic consumers toward their brand and products.In general, the negative emotions of consumers may cause them to ignore the real quality of the product from a specific country.This research investigates the effect of economic-based consumer animosity between Taiwan and Korea and selects consumers' smartphone purchase intention as the research subject.

Match-up Effect of Celebrity Endorser
The match-up effect (e.g., Koernig & Boyd, 2009;Marshall et al., 2008;Till & Busier, 1998;Kim et al., 2013) suggests that endorsers are more effective when a fit between the endorser and the endorsed brand and/or product exists.Theoretically, the globalization of advertisement strategy may be more effective than localization strategy because of the higher fit between the brand origin and the endorser's nationality.Selecting an endorser from the same origin country can increase the match between the advertisement endorsers and the brand image.
The match-up effect may even be significant for the Korean brand because of the popularity of Korean drama and pop song singers.More teenagers around the world admire Korean singers and movie stars, such as Yong-Joon Bae, Girl Generation, and PSY.Young consumers like to follow the consumption behaviour of these Korean endorsers in dramas and music videos.International Korean brands such as Samsung, LG, and Hyundai benefit from the popularity of Korean celebrities.These Korean brands have increased the use of Korean celebrity endorsement in the international market.Therefore, this study developed the following hypothesis: H1: Using a Korean endorser can increase the number of young Taiwanese consumers who have a positive attitude toward Korea to select a Korean smartphone brand than using a domestic celebrity endorser.
However, the match-up effect may become a "disadvantage" if most domestic consumers have negative attitudes toward the COO of the brand.Klein et al. (1998) revealed that economic-based animosity has a negative effect on the willingness of the consumer to purchase a product from the country where consumers have feelings of animosity.For example, consumers in the United States generally believe that Japan is unfair when it trades with the United States, and thus they express high levels of economic animosity toward Japan (Klein et al., 1998).
Therefore, the international company should take this negative attitude more seriously when the consumer animosity problem exists.
Recruiting a domestic celebrity endorser may be the best strategy for international brand managers to advertise their brand.On the one hand, domestic consumers are more familiar with domestic celebrity endorsers than with foreign ones.Using a well-known celebrity endorser may lead to higher brand attitude and purchase intentions toward the brand and/or product (Felix & Borges, 2014).On the other hand, domestic celebrity endorsers may be viewed more favourably than foreign celebrity endorsers because of their personality traits, such as dialect competence, social competence, race, and patriotism (Zhang & Zhang, 2010).Therefore, when domestic consumers have high animosity toward the COO of the brand, using domestic celebrity endorser may be more effective to increase the purchase intentions of consumers.Therefore, this study developed the following hypothesis: H2: Using a Korean endorser can increase the number of young Taiwanese consumers who have negative attitudes toward Korea to select a Korean smartphone brand than using a domestic celebrity endorser.

Pertest
This study aims to show that an external stimulus (endorser's nationality) can affect the purchase intentions of young Taiwanese consumers toward the purchase of a Korean smartphone brand.Therefore, selecting the most appropriate celebrity endorser to control the effect of the attractiveness of the endorser himself/herself is important.Moreover, the present study focuses exclusively on female endorsers to prevent any distorting effect of gender, as previous research showed that male and female subjects are perceived differently (e.g., Freiden, 1984).
To select most appropriate celebrity endorser who could be identified by nationality and associated with the smartphone advertisement, a pertest was conducted to select female celebrity endorsers who were significantly associated with three target traits, namely, (1) appropriateness, (2) trustworthiness, and (3) expertise.The celebrity endorser scored particularly well on these three personal traits (Till & Shimp, 1998).Six Taiwanese and six Korean female celebrities were first selected by the authors and then tested separately (Table 2).All these female celebrities are popular among Taiwanese college students and have no scandal or negative information recently.
In the pertest, young participants were randomly selected from an undergraduate business class at a medium-sized university in Taiwan.The use of student samples was justified because the pertest focus was on the selection of appropriate endorser rather than to generate representative estimates of the decision of consumers.A total of 30 undergraduate students were asked to view the pictures of each endorser and to indicate how much each of the target traits (appropriateness, trustworthiness, and expertise) described each endorser of the new smartphone products (e.g., 1 = not appropriate at all, 5 = very appropriate).The endorser with the highest score in the evaluation in each sub-group (Korean & Taiwanese endorsers) was selected for the final experiment.The evaluation results of the Taiwanese and Korean female endorsers are shown in Table 2.  Note.M = mean of evaluation.

Method
The current study used an experimental research design to assign subjects to different scenarios to observe how the nationality of celebrity endorsers and consumer animosity affect the intention of young Taiwanese consumers to purchase the Korean smartphone brand.An experimental design was used because our understanding of the nationality of endorsers and its effect remains unknown.To test the hypothesis, the study used a 3 × 3 between-subject design (ANOVA), with the three nationalities of endorsers (Korea vs. Taiwan vs. no country = control) and two attitudes of domestic consumers toward Korea (like vs. dislike vs. neutral = control) as the experimental factors, and the purchase intention of consumers as the dependent variable (Table 3).To test the main effect of the endorser's nationality on the purchase intention toward the Korean smartphone brand, the participants were separated into three different groups.We designed three different versions of the questionnaire.Each questionnaire included one possible scenario, such as (1) advertisement with no endorser, (2) advertisement with Korean endorser, or (3) advertisement with Taiwanese endorser.Then, three print advertisements for the Korean smartphone brand (i.e., Samsung Galaxy Note 3) were developed in cooperation with a professional advertising expert (Appendix A).To control the unpredictable influence of different brand names or product characteristics, all three advertisements used "Samsung" and "Galaxy Note 3" as the subjects.
The advertisement styles were designed as similarly as possible.

Sampling and Procedure
The experiment was based on a separate sample of business school students from two universities in the central Taiwan.Similar to all studies with a student sample (e.g., Kamins, 1990;Little et al., 2012;Till & Busier, 1998), the generalizability of the finding was questioned.However, the use of student sample was justified in this research because we focused on the attitudes of young Taiwanese consumers toward a Korean smartphone brand; thus, university students were the appropriate sample.Furthermore, the experimental research method tested the theoretical hypotheses rather than generated representative estimates of consumer response to the endorser 's nationality.Thus, the use of student sample was appropriate.
The experimental procedure was divided into two stages.In the first stage, each student participant was asked the simple question of "Do you like Korea?" by a research assistant.To increase efficiency, if the participant answered "so-so" or "no comment," then, he/she would be assigned to the control groups (Treatments 7-9).Only the participants who said "Yes, I like Korea" or "No, I hate Korea" became the targeted participants in this research and assigned into the experimental groups (Treatments 1-6).All the participants were divided into three groups (like, dislike, and neutral) and were randomly assigned one of the three different versions of the questionnaire (no endorser, Korean endorser, and Taiwanese endorser).
To increase the reliability of research and reduce the misunderstanding on the scenarios, two filter questions "From what country does the Samsung brand originate?"and "What is the endorser's nationality in this questionnaire?" were asked.If the participant misunderstood the scenarios or gave the wrong answer of any filter question, the questionnaire would be eliminated immediately from the study.Finally, a total of 486 undergraduate Taiwanese students participated in this experiment.However, 36 were excluded from the data because they did not answer the filter questions correctly.In total, 450 Taiwanese students were valid (50 per treatment) and completed the questionnaires for further analysis.The usable respondent rate was 92.6%.
In the second stage, the respondents were asked to indicate their agreement using a seven-point Likert-type scale (1 = "strongly disagree" to 7 = "strongly agree" scale).The statements on the two general constructs were (1) consumer animosity toward Korea and (2) willing to buy a Samsung smartphone.All measures were adapted from previous studies.For example, the attitude of consumers toward Korea was measure using eight items adopted from Klein et al. (1998).Purchase intention was measured using four items adopted from Dodds et al. (1991).
This research tested the reliability and validity of the variables.Table 4 shows the values of average variance extracted (AVE), composite reliabilities and the values of the coefficient α.As expected, the factor loadings (γ) of all items were greater than 0.60, and all estimates for AVE were greater than 0.77 (Fornell & Larcker, 1981).
As for convergent validity, all items had significant loadings on their respective constructs (Anderson & Gerbing, 1988).Composite reliabilities for the two constructs exceed 0.6, the generally recommended threshold (Bagozzi & Yi, 1988).The reliability test showed that the Cronbach's alpha values ranged from 0.95 to 0.96.In short, the results indicated good reliability of the measurements (Bagozzi & Yi, 1988;Hair, 2010).

Manipulation Checks
To test the correlation between consumer animosity and purchase intention, this research first test the relationship with the statistical software, AMOS.It provides standardized coefficients (S. β).Thus, coefficients between constructs can be interpreted in the same manner as standardized regression coefficients in classical ordinary least squares regression.The overall fit of the model was unsatisfactory (χ 2 = 381.88,df = 53, p < 0.01) but the relative fit indices, e.g., the comparative fit index (CFI) and normed fit index (NFI), were 0.95 and 0.94, respectively.The result shows that there is a negative relationship between consumer animosity and purchase intention (S.β = -0.69,t = -17.56,p < 0.001).
To generate accurate stimuli for the experimental manipulation, this study conducted an ANOVA, with the measured consumer attitude toward Korea as the dependent variable.The results reveal that consumers who like Korea (M Like = 5.65) are significantly more favourable than consumers who are neutral and dislike Korea (M Neutral = 4.06, M unlike = 2.82; F (2, 447) = 792.97,p < 0.01).This study also conducted a similar manipulation check on purchase intention.Consistent with our expectations, the results show that consumers who like Korea (M Like = 4.89) are the highest among the respondents compared with consumers who are neutral and dislike Korea (M Neutral = 4.24, M unlike = 2.88; F (2, 447) = 173.88,p < 0.01).Therefore, the manipulation of both independent variables is successful.

Experimental Results of Using Vs. Not Using Celebrity Endorser
To examine the positive effect of celebrity endorser, we first combined treatments 1, 2, 3, 4, 7, and 8 into group 1 (with endorser) and treatments 5, 6, and 9 into group 2 (without endorser) to test the effect of the endorser (Table 2).Then, we conducted a 2 (with endorser vs. without endorser) x 3 (like vs. dislike vs. neutral) mixed ANCOVA using the endorser as the between-subjects factor and consumer attitude toward Korea as the covariate.The analysis reveals a statistically significant effect of the covariate (F (2,444) = 168.582,p < 0.01).This finding indicates that the use of an endorser is positively associated with the attitude of consumers toward Korea.
The main effect of the use of endorser on the purchase intention of consumers who like Korea is marginally significant (F (1,148) = 3.825, p = 0.052).The same result was found for those who dislike Korea (F (1,148) = 7.648, p < 0.01).However, the main effect of using an endorser on the purchase intention of consumers who remain neutral to Korea is not significant (F (1, 148) = 1.48, p = 0.226).Consistent with the expectation, using a celebrity endorser to promote a Korean smartphone brand has the most effect on young Taiwanese consumers who like Korea (M = 4.984) and the least effect on those who dislike Korea (M = 3.076).Except for consumers with a neutral attitude toward Korea, the analysis reveals the significant effect of celebrity endorsement, particularly among consumers who have high animosity toward Korea (Figure 1).
Figure 1.The effect of using endorser on purchase intention

Experimental Results
To further understand the correlation among endorser's nationality, consumer attitude toward Korea, and purchase intention, this study conducted a 3 (with Taiwanese endorser vs. with Korean endorser vs. without endorser) x 3 (like vs. dislike vs. neutral = control) mixed ANCOVA, with the endorser's nationality as the between-subjects factor and the attitude of consumers toward Korea as the covariate.The analysis reveals the statistically significant effect of the covariate (F (2,441) = 212.364,p < 0.01), which indicates that the endorser 's nationality is positively associated with the attitude of consumers toward Korea.
Figure 2 shows that the endorser's nationality has a non-significant difference in the increase in intention of consumers who like Korea to purchase a Korean smartphone brand (F (2,147) = 2.169, p = 0.118).Although the Korean endorser has a slight advantage in increasing the intention of consumers to purchase a Korean smartphone brand (M Korea = 5.044) than the Taiwanese endorser (M Taiwan = 4.924) and without an endorser (M No = 4.704), the advantage is not significant.Therefore, H1 is rejected.
The result also reveals that the endorser's nationality has a statistically significant difference in the increase in the intention of consumers to purchase a Korean smartphone (F (2,147) = 3.567, p = 0.031).The Taiwanese endorser has a significant advantage in increasing the intention of consumers to purchase a smartphone (M Taiwan = 4.46) than without an endorser (M without endorser = 4.144) and with a Korean endorser (M Korea = 4.128).
Interestingly, the use of a Korean endorser has the disadvantage of attracting the purchase intention of neutral consumers, and the gap between using domestic and using Korean endorsers becomes larger in the group who dislikes Korea.Figure 2 shows that the endorser's nationality has a significant difference in the increase in the intentions of consumers who dislike Korea to purchase a Korean smartphone brand (F (2,147) = 34.066,p < 0.01).A Taiwanese endorser has a significant advantage in enhancing the intention of consumers to purchase a smartphone (M Taiwan = 3.852) than without an endorser (M without endorser = 2.500) and with a Korean endorser (M Korea = 2.300).Therefore, H2 is supported.
Figure 2. The effect of using different nationalities of endorser on purchase intention

Conclusion
Celebrity endorsement is one of the most frequently used methods in advertising to influence the purchase intentions of consumers (e.g., Choi & Rifon, 2007;Ding et al., 2011;Hollensen & Schimmelpfennig, 2013;Kamins, 1990).An increasing number of international companies are more likely to use celebrity endorsers than other types in their advertisements because of the efficiency and effectiveness of using celebrity endorsers.However, studies discussing the effect of the nationality of celebrity endorsers on the purchase intentions of consumers are scarce.Therefore, our study outlines the potential disadvantages of the match or mismatch between the nationality of celebrity endorsers and the endorsed brand.Overall, we discovered that the nationality of celebrity endorsers affects the intentions of young consumers to purchase a smartphone.
This result also provides new insights that add more discussions to previous research that ignored the effect of the nationality of celebrity endorsers (e.g., Bhatt et al., 2013;Felix & Borges, 2014).Contradictory to most previous research that found a positive relationship between the use of endorser and purchase intention, this

Attitude toward Korea
No endorser Taiwanese Endorser Korean Endorser research found that, in some situations, the use of animosity celebrity endorser has a negative effect on purchase intention.Consistent with the arguments of Klein et al. (1998) and Nijssen and Douglas (2004), animosity emotion generates the negative attitudes of consumers toward the products of the "aggressor" country and their unwillingness to purchase them.This study also found the same relationship between consumer animosity and purchase intention.The use of animosity celebrity celebrities in advertisements may lead to the negative feelings of some consumers toward products and/or brands endorsed.This phenomenon is particularly significant for consumers who have a strong animosity toward Korea (Figure 2).

Implications
Our findings provide some useful managerial implications for practitioners to deal with the use of celebrity endorsers.Considering that modern consumers can see various celebrities with different nationalities on TV, the Internet, newspaper, fliers, and magazines, international brand managers should not only consider the fit between the personal characteristics of the endorser and the endorsed brand and product, but they should also know the effect of the endorser's nationality, particularly when they intend to promote their products and/or brands in the countries with high consumer animosity complex in economics (e.g., Taiwan vs. Korea and US vs. Japan) and/or wars (e.g.,China vs. Japan and the UK vs. Germany).
International brand marketers should realize that the nationality of endorsers may either have positive or negative effects on the purchase intention of consumers.To reduce the negative effect of consumer animosity, international markers use domestic celebrity endorsers to reduce the negative feelings of consumers toward the COO of the brand.The result of this study shows that using a domestic endorser is more advantageous in attracting patriotic consumers and those who are neutral to the COO of the brand.This result also explains why Samsung first used a Korean endorser, Ji-Eun Lee, to endorse its smartphone in 2011 but then used Taiwanese endorsers for their new products in the subsequent years (see Table 1).
Nevertheless, international brand managers can also select advertisement endorsers from their original country to increase the match between brand image and advertisement endorsers.The use of brand original celebrity endorsers may not always be a bad idea (Little et al., 2012) because economic-based animosity is usually temporary and less likely to be passed from generation to generation.In recent years, some Taiwanese youths have increasingly developed a positive attitude toward the Korean popular culture (e.g., pop songs, dramas, and movies).Using Korean endorsers may be an effective strategy to change the attitude of young Taiwanese consumers toward Korean brand products.For a Korean brand manager, this result offers a new opportunity to recruit Korean celebrity endorsers to increase the match between brand image and advertisement endorsers.This result also explains why Samsung has recently changed its brand strategy and used Soo-Hyun Kim, a Korean drama star, as its new smartphone endorser in May 2014.

Limitations and Future Research
Similar to all pilot research, this study has several limitations.First, the context of this study was based on the discussion of two countries with economic-based animosity (i.e., Taiwan vs. Korea), and it successfully proved that the use of celebrity endorser has a positive effect on the intention of young Taiwanese consumers to purchase a smartphone.However, in the real world, consumer animosity is complex and varied (Klein et al., 1998).Different types of consumer animosities may have distinct effects.For example, Little et al. ( 2012) found that economic-based animosity (e.g., America vs. China) is temporary and less likely to be passed from generation to generation compared with war-based animosity (e.g., America vs. Vietnam).Therefore, comparing this research finding with other types of consumer animosity, such as war-based and political animosity, would be interesting.
Second, the study aims to show that an external stimulus (endorser's nationality) affects the intention of young Taiwanese consumers to purchase a smartphone.Thus, this research used an experimental design method with undergraduate business school students rather than real consumers as participants.The student consumers were not actually exposed to a genuine shopping setting, and thus they might not truly care about the product quality, price, user experience, and purchase risk as seen in the real world.Therefore, future studies should process the test with "real customers" and compare the difference with this research.
Finally, to reduce potential uncontrollable factors, this study was conducted with a specific type of high-involvement product: A smartphone.However, the use of different types of products may attract different levels of consumer attention and responses.For example, although consumers spend more time and energy searching for smartphones, they tend not to spend even more time searching for shampoo.However, these low-involvement products, such as shampoo, soft drinks, and soap, are usually endorsed by celebrities.Therefore, the replication of this research using different types of products should help verify the generalizability of the

Table 2 .
Evaluation of Taiwanese and Korean female endorsers

Table 3 .
Research design

Table 4 .
Reliability and validity of variables.The likelihood of purchasing this smartphone is.At reasonable price, I would consider buying the product.If I were going to buy a smartphone, the probability of buying this model is.The probability that I would consider buying the product is.