How Chinese Face Perception Influences Consumer ’ s Implicit and Explicit Attitude towards Brand Country of Origin

Chinese people pay much attention to their faces. Face plays a very important role in Chinese daily life. Chinese people usually do kinds of Face Works in order to earn face, enhance face, or save face. Face work influences consumer’s attitude toward brand country of origin (BCO) in China market. What’s more, face work makes it different between implicit and explicit attitudes towards a certain brand. Although existing literatures have verified some factors that affect implicit and explicit attitudes towards BCO, few research study the effect of face perception on BCO. The study uses both questionnaire and IAT technique to measure the effect of face perception on the implicit and explicit attitudes toward BCO. Results showed that: (1) consumers with high face perception have a more positive explicit attitude toward Japanese /American brands, and (2) face perception moderates the relationship between implicit and explicit attitudes towards BCO.


Introduction
The phenomenon exists that the purchase behavior of Chinese consumers deviates from their evaluation on the brands from different country.In China, for example, Chinese consumers show apparent deviation when buying and evaluating the Japanese automobile.According to the report by Sohu Automobile (http://auto.sohu.com/20110117/n278913045_6.shtml), during the period from January to December in 2010, of the Passenger Vehicle market in a narrow sense (car+MPV+SUV) in China, domestic brands accounted for 32.8%,Japanese brands 23.8%, German brands 17.9%,American brands 12.5 %, Korean brands 9.9%, and French brands 3.3%.As we can see, the sales of Japanese cars are far better than that of Germany, America, and Korea, which is really unexpected.It is because Chinese consumers always have negative evaluation on Japanese automobile.And through the auto channel of Sina and Sohu, it's easy to find a bundle of negative evaluation on Japanese automobile from Chinese consumers.Existing studies also found out that some Chinese consumers consider it a shameful behavior to buy Japanese automobile (Shi et al.2010).The deviation between consumers' behavior and attitude has attracted the interests of both scholars and marketing managers.
Many scholars study the deviation between consumers' behavior and attitude from different perspectives.Some scholars focus on the attitude or behavior process, including perspectives from the external variables, mediating variables and intensity (Arends, 2006;Fiske et al., 2002).Other scholars explore it from the attitude components (Fazio, 1984;Marlowe, 1975;Schleiche, 2004).They believe that cognition, emotion and behavior are not in the same level.Cognition and emotion affect the behavior intention, subsequently, determine the behavior itself.In the study of the deviation between consumers' behavior and attitude towards brand country of Origin (BOC), Klein et al (1998) found that Chinese animosity against Japan led to the deviation between their behavior and attitude.Some scholars found that there was also difference between Chinese consumers implicit and explicit attitudes towards Japanese brands (Li Yanjie, Huang Taohui and Shi Junqi, 2009;Yang Yangzi, Huang Taohui and Shi Junqi, 2008), which can explains the deviation.
In the collective culture, Chinese pays special attention to the face, and face has the ability of social control and norm constraint (Zhu Ruiling, 1988).Face is the social esteem obtained from others; or the public image allowed and permitted by others, also the image accepted and recognized by others in the social interaction.Face perception is the perceived self-image once someone receives some kind of social feedback (Zhu Ruiling, 1988).In the interpersonal interaction, people are used to adopting Face Work.Face Work is a variety of social skills to save face, which are developed when dealing with intercourse, especially in the situation of face to face intercourse.Face Work aims to save face, so it leads to a variety of hypocritical "formalism", which means "To be obedient seemingly but think and act on the contrary" (Jin Yaoji, 1988).In the consumer decision-making process, compared with Western consumers, Chinese consumers are more concerned about others' opinions and the social effects caused by their own consumption, and the so-called "face" is the manifestations of such social effect (Qi Haifeng, 2009).Obviously, because of the existence of face, there are difference between people's evaluation and true attitude.However, no existing studies focus on the change of the attitude from the perspective of face concept.In order to make up for the gap of existing literature, this paper studies how face perception has different effects on implicit and explicit attitudes, and tries to explain the deviation between their behavior and attitude.This paper firstly reviews existing literature about face perception, country of origin effect and attitude.Basing on experimental design and study, this paper tries to find the possible reasons to explain why face perception affects implicit and explicit attitudes towards BCO, and then analyzes how face perception affect consumer's attitude towards BCO.Finally the theoretical and managerial implications will be put forward.

Connotation of Face
Face is a social psychological construct rooted in culture, and is the social esteem obtained from others; or the public image allowed and permitted by others (Goffman, 1955;Lim andBower, 1991 Chen Zhizhao, 1982)."FACE" was firstly divided into two dimensions by Dr. Hu(1944), which was "LIAN" and "MIAN".The two interrelated and disparate concepts provide a direction for studying the connotation of Chinese face.Dr. Hu proposed that "MIAN" is kind of prestige stemming from the visible achievements or flaunting; "LIAN" is kind of respect given to someone who has a moral reputation in a group, which can be called the "moral MIAN" (Jin Yaoji,1988).Along with this view, some other Chinese scholars, such as Jin Yaoji(1988) and Cheng Zhongying(1982) etc., further enriched and developed the definition of "LIAN" and "MIAN".
Face is gained through giving "situation interpretation" to all kinds of clues in the social interactions (Mead, 1934).People usually regard these behaviors, such as, curse, exposing others' scars, wearing Rolex, as a kind of "symbol", and then take action (Face work) according to their interpretation of the symbols (known as face perception) (Zhou Meiling and He Youhui,1993).Face perception is the perceived self-image once someone receive some social feedback, and one's face perception is his/her perception of social expectation and social value (Zhu Ruiling, 1988).Individuals will form different face perception through all kinds of face events, at the same time, accumulate and integrate the experience of the past, thus develop different feelings for having face or losing face (Chen Zhizhao,1982).
The operation process of Face basically follow the three steps in the social interaction, which are social stimuli, cognitive systems (face perception) and response system (Chen Zhizhao, 1982;Huang Guangguo, 1987;Zhu Ruiling, 1988).The two parties in the social interaction evaluate one's own face and other's face according to the context information, and then adopt the corresponding face Kung Fu (Face Work).Although everyone has basic face needs (Brown and Levinson, 1978), the needs in different social situations can arouse different attention and content, while these face motives or intentions include maintaining, manifesting or hurting one's own/others' face (Zhou Meiling and He Youhui, 1993).When one's face being threatened, he tends to take action or give response to contend for saving-face, prevent face losing or enhancing-face (Bao Gongmin and Zhao Zhuojia, 2009;He Youhui, 1976).

The Mechanism of Face
In the study of "face consumption behavior", Su et al. (2007) believe that face consumption is a motivational process, in which individuals attempt to improve, maintain and keep their own face through consuming products, while showing respect to others' face.The stronger one's face perception is, the more inclined to consume initiatively and relationship-orientedly, also the more inclined to take others' evaluation as standard (Jiang Caifen, 2008)(Note 3).What's more, consumers with higher face perception are more willing to choose brands with high popularity (Shi Zhuomin et al., 2010).Some scholars believed that Chinese Face has two parts: "LIAN" and "MIAN" (also called MIANZI, MIAN is an abbreviation of MIANZI)."LIAN" is kind of respect given to someone who has a moral reputation in a group (Hu, 1944;Jin Yaoji, 1988;Cheng Zhongying, 1982), which is heavily evaluated by moral criteria or social norms.On the other hand, "MIAN" is kind of personal prestige stemming from the visible achievements or flaunting, which is mostly evaluated by a person's social achievement or social status.When study how "LIAN" and "MIAN" affect consumer behavior, Shi Zhuomin(2011) et al. discovered that the consumers' perception of "LIAN" and "MIAN" differs on the attitude and cognition towards the products in terms of country image.With positive face cognition, people believe that buying Germany cars will mostly "HAVE MIAN" (enhance their status), followed by American cars and Japanese cars, and the purchase of domestic cars was considered to "HAVE NO MIAN" (no value added for one's status); With negative face cognition, people think that buying Japanese car is highly related with "LOSE LIAN" ("shame") (Shi Zhuomin et al., 2011).In addition, a research on the difference of face cognition towards foreign cars selling in China found that, Chinese consumers perceive that they will both "HAVE LIAN" and "HAVE MIAN" when buying Germany cars, American cars and French cars; Meanwhile, Buying Chinese cars will make people feel "HAVE LIAN but LOSE MIAN", Interestingly, Buying Japanese cars makes people feel "HAVE MIAN but LOSE LIAN", while Buying South Korean cars makes people feel neither "HAVE LIAN" nor "HAVE MIAN" (Wang Fang, 2010)(Note 4).
However, the existing researches about the relationship between country of origin and face, just stay in the level where different countries of origin bring consumers different face perception.There is a lack of research about how face influences country-of-origin effect.This study aimed to clarify how face perception affect implicit and explicit attitudes on brand country of origin (BCO), helping managers as well as foreign scholars to better understand Chinese consumers on brand attitude.Schooler(1965) thought that country of origin can be explained by "Made in", that is, country of origin is where the product is made or where the company make products.Some scholars defined country of origin as the country where the product is made (Bilkey and Nes, 1982).This study defines country of origin as "the country where the headquarters marketing products or brands located" (Johansson et al., 1985).Country of origin is a multidimensional concept, which can arouse people's cognitive response and emotional reaction (Han and Terpstra,1988;Lampert and Jaffe,1996;Nebenzahl et al.,1997).When consumers lack detail information, country of origin can provide the information to help consumers to evaluate the quality, the reliability and the value of the product (Han andTerpstra,1988；Hong andWyer,1989).In addition, consumers' emotions of the country of origin also affect their evaluation on the products.Emotion factors such as patriotism and animosity (war animosity and economy animosity) have influence on consumers' evaluation of BCO (Klein et al., 1998;Shimp and Shanna, 1987).

Country-of-origin Effect
Country-of-origin effect refers to the influence that the information about original country have on consumers' evaluation, attitude to products and purchase intention (Schooler, 1965).Schooler's study (1965) confirmed the country-of-origin effect, and found that products from developed countries are more popular than that from relatively underdeveloped countries.Subsequently, numerous researches revealed some factors which influence country-of-origin effect from the perspective of country properties (Cattin,1982;Nagashirna,1977), product attributes (Kaynak et al.,2000;Wang Haizhong and Zhao Ping,2004) and consumer factors (Janda and Rao,1997;Johansson et al.,1985).
With the deepening development of the research about the country-of-origin effect, researchers gradually focused on how consumer variables influence country-of-origin effect from the perspective of the country properties, product attributes and other cognitive factors.But there is a lack of literature from the perspective of consumer variables.

Research Method of Country-of-origin Effect
In previous researches on country-of-origin effect, most researchers used self-report methods such as interviews and questionnaires to measure consumers' attitude to brands.But in fact what self-report methods measured was the explicit attitude of the subjects (Yang Yangzi, Huang Taohui and Shi Junqi, 2008).In recent years, some scholars used implicit association test (IAT) which can eliminate social desirability response bias when measure consumers' attitude to brand country-of-origin (BCO).These findings revealed that there exists inconsistency between implicit and explicit attitudes when consumers evaluate a certain brand (Li Yanjie, Huang Yunhui and Shi Junqi, 2009;Yang the Yangtze, Huang Yunhui and Shi Junqi, 2008).

Implicit and Explicit Attitude
The most popular definition of attitude can be classified into two kinds, one is monism system, and the other is ternary system.Ternary system defines attitudes as well-known ABC-attitude theory (Affective-Behavioral-Cognitive Model of Attitude), which includes affective, behavioral and cognitive elements (Eagly and Chaiken, 1993;Nesdale and Durkin, 1998).Through the in-depth study, scholars carried out an intensive discussion on whether "evaluation as a core element of the attitude" or "attitude as the connection between object and evaluation".By integrating the existing researches, Greenwald and Banaji(1995) initiatively came up with a concept of "implicit social cognition" .

Implicit Attitude
In 1995, social psychology researchers, Anthony Greenwald and Mahzarin Banaji proposed the extension of ideas already existing in cognitive psychology to social psychology.Greenwald and Banaji (1995) defined the implicit cognition as "introspectively unidentified (or inaccurately identified) traces of past experience that mediate favorable or unfavorable feeling, thought, or action toward social objects" (Greenwald & Banaji, 1995. P.8).The formation of implicit attitude is a relatively slow process, which needs accumulation of over-learned knowledge and experience.Compared with explicit attitude, implicit attitude is very stable and difficult to change (Banaji, 1993;Gregg, 2004;Smith and Decoster, 2000).In 1998 Greenwald et al. introduced IAT technique to measure the implicit attitude.The Implicit Association Test (IAT) is a measure within social psychology designed to detect the strength of a person's automatic association between mental representations of objects (concepts) in memory.The IAT procedure seeks to measure implicit attitudes by measuring their underlying automatic evaluation.The IAT is therefore similar in intent to cognitive priming procedures for measuring automatic affect or attitude (Greenwald & Banaji, 1998).The introduction of IAT technique allows researchers to understand attitudes that cannot be measured through explicit self-report methods due to lack of awareness or social desirability bias.

Explicit Attitude
Explicit attitude is the product of thinking and result of self-reflection, which affected by the factors such as self-service motivation, personal goals or the purpose of obedience to social standards (Greenwald and Banaji, 1995).In today's highly socialized environment, everyone is careful to maintain his/her image.Therefore, people are accustomed to using impression management method (impression management theory)(Note 1) to shape impression, and to conform to group norms, so as to achieve the goals or desires of the society.Although some people do it on purpose, most people have such behavior unconsciously (Depaulo, 1992;Schlenker, 2004).Individuals will think about others' feelings when he extracts and expresses his attitude during social intercourse.Therefore, consumers will change their attitude when they evaluate products or brands to keep consistent with his membership groups or the image in specific situations (Yang Yangtze, Huang Yunhui and Shi Junqi, 2008).

Relationship between Implicit and Explicit Attitude
As to the relationship between implicit and explicit attitudes, there are two kinds of views: consistency and inconsistency.The view of consistency assumed that what implicit and explicit attitudes measure is the same psychological structure, and implicit attitude measures the true attitudes, whereas explicit attitude was a distorted expression of implicit attitude after interference by other factors (Fazio and Olson, 2003;Tonherg, 1996).The view of inconsistency assumed that implicit and explicit attitudes are two different inner-psychological structures, having different psychological processing mechanism.And explicit attitude is product of thinking and result of self-reflection, while implicit attitude perhaps is a result of unconsciousness which is inaccessible through introspection, so the implicit and explicit attitudes are not relevant (Dovidio et al., 2000;Greenwald and Banaji, 1995).No matter which view has its sufficient theoretical basis and empirical evidence (Wu Mingzheng, 2005).
Through studying the correlation between the implicit and explicit measurement of attitude (use the correlation coefficient R), researchers tried to find out the relationship between implicit and explicit attitudes, and this method can be called IEC (Implicit-Explicit Correspondence) study.For example, when self-presentation levels are high, explicit attitude may change but implicit attitude may not, so self-presentation and Implicit-Explicit Correspondence are negative correlated; when self-presentation levels are low, individuals conceal less on their attitudes, so self-presentation and Implicit-Explicit Correspondence are positive correlated (Greenwald and Banaji,1995).

Hypothesis
This article, from typical Chinese cultural psychological constructs --face, attempts to study the influence of different degree of face perception on the explicit attitude to Brand Country Origin (BCO), and the effect of face perception on the relationship between implicit and explicit attitudes.
In choice of the Brand Country Origin, this study selected China, America and Japan as the research object.The main reasons why choose these three countries lie in the followings.First, China has been holding large trade volume with both America and Japan.China also has attracted much of foreign investment from the world since 1990s, among which America and Japan are all the time top 5 countries heavily invested in China, especially invested heavily in the industries of automobile, food and household appliances etc.As a result, Chinese consumers are very familiar with brands from Japan and America.Second, the World Brand Lab publishes the 500 strong world brands, of which American brands and Japanese brands always rank top of the list.Moreover, America and Japan are developed countries, but Chinese people certainly hold a special feeling for Japan rather than for America due to the war.It is interesting to study whether there is significantly different implicit/explicit attitude toward BCO between the two developed countries and why.

Effects of Face Perception on Explicit Attitude to Brands Country Origin (BCO)
When asked to evaluate brands originated from different countries, consumers can retrieve their own consumption experience with this product from memory systems, or use their bias according to the impression of the country where a product/brand is originated.(Li Yanjie et al, 2009;Yang Yang, 2008).Culture and values certainly have effect on country-of-origin (Durairaj et al., 2000), and face as a typical Chinese cultural characteristic, affect consumers' evaluation of brands originated in different countries.Subject to social expectation, explicit evaluation is susceptible to survey environment and skills.When directly interviewed, people are likely to give answers to cater for the need of others or tend to give answers for social praise, so as to hide their true thoughts and behaviors (Depaulo, 1992).Face in Chinese culture has the power of social control and social norm (Zhu Ruiling, 1988).When one's face perception is high, it often means that "his social desirability needs are strong, or considering social desirability as an important social value", and he has strong motivation to strive for gaining face or prevent losing face; while one's face perception is low, he thinks it no need to do face work or to earn face (Chen Zhizhao, 1982).Compared with consumers with low face perception, those with high face perception pay more attention to others' evaluation on them, so they are likely to consider the feelings of others when extracts and expresses their own attitudes (Wu Mingzheng, 2004).Compared with China, America and Japan are developed countries, consumers hold a more positive attitude toward products from economically developed countries (Schooler, 1965).Therefore, this paper proposed the following hypothesis: H1: Compared with consumers with low face perception, those with high face perception hold a more positive explicit attitude to brand originated from developed countries than brand originated from China.
H1a: Compared with consumers with low face perception, those with high face perception hold a more positive explicit attitude to Japanese brand than Chinese brand.
H1b: Compared with consumers with low face perception, those with high face perception hold a more positive explicit attitude to American brand than Chinese brand.

Face Perception Moderates the Relationship between Implicit attitude and Explicit Attitude towards BCO
In the implicit association test, individuals evaluated BCO in an unconscious way through their consumption experience and existing attitude (Li Yanjie et al, 2009;Yang Yang, 2008).Therefore, what affects implicit attitude is not confined to brand itself; it extends to all aspects of Country-of-Origin, especially consumer long-formed values and prejudices.Internalized moral face may affect the implicit attitude.But the binding ethics must be originated from the social supervision, that is, only when one perceives there is an audience around him, can face reflect itself (Zhu Ruiling, 1988).Implicit attitude also is a kind of unconscious attitude with stability (Greenwald and Banaji, 1995), and in the IAT test, consumers do not have too much time for conscious thought, therefore, face perception did not have a significant impact on implicit attitude.
The reflection of explicit attitude is subject to social desirability, while implicit attitude is not.It reflects the evaluation that individual is unable or unwilling to report (Greenwald and Banaji, 1995).Explicit attitude changes easily.Compared with explicit attitude, implicit attitude is difficult to be influenced by other factors.Implicit and explicit attitudes are two attitudinal levels which are different but associated.For example, the higher preference to Chinese brand in implicit attitude of Chinese consumers, the higher preference to Chinese brand in explicit attitude (Yang Yangtze, 2008).Face, as a kind of social demand motivation, does have impacts on the reflection of one's attitude in Chinese Culture.Compared with consumers with low face perception, those with high face perception perceive gaining face or losing face more easily, tend to be more consistent with the group.Face perception has a greater impact on explicit attitude, while has no significant effect on implicit attitude, which enhances the inconsistency between the implicit and explicit attitudes (Dovidio et al.,2000;Greenwald and Banaji,1995), and weakens the correlation between them.Therefore, this paper proposed the following hypothesis: H2: Face perception moderates the relationship between implicit and explicit attitudes toward BCO.
H2a: When face perception is high, the consistency between implicit and explicit attitudes towards Japanese brand is low; when face perception is low, the consistency between implicit and explicit attitudes towards Japanese brand is high.
H2b: When face perception is high, the consistency between implicit and explicit attitudes towards American brand is low; when face perception is low, the consistency between implicit and explicit attitudes to American brand is high.

Subjects
85 college students and 47 MBA students from a Guangzhou University participated in testing, valid sample is128, with 64 male and 64 female, and average age is 24, all of whom are right-handed persons.

Experiment Procedure
This research mainly adopts scales to measure subject's face perception and explicit attitude, using IAT to measure subject's implicit attitude.All scales and IAT test operate through the Windows XP Inquisit experiment system.In the study, each subject underwent three steps including scale tests, classification task to separate Chinese brands and Japanese brands, classification task to separate Chinese brands and American brands.
The whole experiment process was carried out in a lab environment, subjects were told to turn off their mobile phones and keep quiet before entering the lab, and arranged to sit by the guide members.Before the test, subjects were informed of the experiment purposes and requirements, as follows: "This is a survey about life attitude and Chinese attitude towards the products from different country, which includes 4 parts, and it will take you about 15 minutes.Please keep quiet and independent, carefully complete all tests, otherwise experimental results will be invalid." When finishing the test, only after staff checked the IAT scores, can subjects receive gifts and leave.

Measurement of Face Perception
The measurement adopted the face perception scale developed by Shi Zhuomin (2011), and the questionnaire includes 40 items, which has a high reliability in the domestic research.For the need of research, the author integrated repeat-meaning items, eventually it contained 30 items.

Measurement of Explicit Attitude
After the first step, subjects need to evaluate Chinese / Japanese / American brands.As to this part, the measurement refers to Country of Origin Image Questionnaire (country of origin image, COI).The scale was developed by Parameswaran and Pishaodi (1994) and revised by Yang Yangtze etc. ( 2008).The questionnaire was revised according to needs of this research.The final edition included 10 items, such as service, product quality, advertising and the overall national situation etc.

Measurement of Implicit Attitude
After the second test, subjects were required to complete the measurement of implicit attitude.In this part, we used the implicit association test (Implicit Association Test, IAT).Through measuring the distance of two kinds of concepts (target concept and attribute discriminations) in the neural link, the reaction time could be used as indicators to reflect the connecting tightness between a concept and an attribute of the subjects.The connecting tightness reflects the subject's attitude to a certain target, thus avoiding the subject's existing awareness and impression management strategy in the explicit attitude measurement (Yang Yangtze etc, 2008).The IAT method is widely used in social psychology, consumption psychology, and is confirmed to have good reliability and validity (Hofmann et al., 2005;Maison et al., 2004;Nosek, 2005).
In the IAT experimental procedures, Chinese words are applicable to be used as experimental materials of target concept and attribute discrimination concept.In order to select the appropriate target word, we made a pre-tested, asked 40 subjects to report the brands in their minds originated from China, Japan and America respectively.We recorded the reporting resulting and selected 5 brands standing for each country with highly brand awareness.The five Chinese brands are Haier, Lenovo, Geely, Hongqi and Midea; The five Japanese brands are Panasonic, Sony, Toyota, Honda and Hitachi; The five American brands are Microsoft, Apple, Dell, Ford and Buick.Referring to the research results from Nosek (2005), the attribute discrimination concept in this study include 5 positive words (elegant, perfect, reliable, environmental-friendly, leading) and 5 negative words (inferior, rough, heavy, cumbersome, conservative).
Referring to the experimental design which have eliminated effect of order (Greenwald, Nosek and Banaji, 2003), IAT experiment in this research could be divided into 14 stages (see Table 1), in which all words were randomly presented.
At the start of each stage, guidance shown on the screen would inform subjects of experimental requirements, and the subjects were asked to respond as soon as possible on the basis of correctness.If the response was correct, words would disappear and entered the next test; Otherwise, it would appear red" X" in the original position of the words.Red" X" would not disappear until the subjects responded correctly, then skipped into the next test.All tests completed, screen would show "Finish" and gave the IAT scores.
After the IAT test, the subjects need to fill out basic information, including gender, age, education level and monthly living expenses etc.

Scale Reliability Test
Based on retesting the scale of face perception originally developed by Shi Zhuomin (2011), we found the Cronbach alpha coefficient was 0.82.The scale of face perception had a good reliability.Then we tested impression of Country-Image scale which was originally developed by Parameswaran and Pishaodi (1994).After reversely calculating the scores of each item in explicit attitude scale, we calculated the reliability of impression scale about products from China, Japan and USA, and found the Cronbach alpha coefficient were 0.79, 0.73 and 0.82.

Results of Face Perception Measurement
The average score of 30 items in the face perception scale was regarded as the face perception score, and all subjects' face perception scored an average of 3.01(SD=0.42).The subjects with higher face perception score (which is greater than the overall average score), were regarded as the high face perception group (M=3.41,SD =0.31), otherwise were the low face perception group (M=2.69,SD=0.20).Using SPSS16.0 software for independent sample mean test(T test for independent samples), results showed the score of high face perception group was significantly higher than that of low face perception group, t = -15.90,P < 0.01.The number of subjects in high face perception group and low face perception group is respectively 60 and 68.

Influence of Face Perception on the Explicit Attitude towards BCO
The average score of 10 items in the product (from China/Japan/USA) impression scale was regarded as impression score of the products from China/Japan/USA respectively.The higher the score, the more positive evaluation of Brand Country Origin (BCO), and the higher the explicit evaluation.The explicit evaluation score of Japanese brands minus that of Chinese brands, then we would obtain the relatively explicit evaluation score of Japanese brands; Likewise, the explicit evaluation score of American brands minus that of Chinese brands, we would obtain the relatively explicit evaluation score of American brands (See Table 2).
Paired Sample T Test on the impression scale of Chinese/Japanese/American products was taken.The results showed that, on the whole, the explicit evaluation of Japanese brand (M=3.84,SD=0.50) was higher than that of Chinese brand (M=2.70,SD=0.52), t = -17.97,p< 0; the explicit evaluation of American brand (M=3.94,SD=0.43) was also higher than that of Chinese brand (M=2.70,SD=0.52), t= -20.56, p < 0 (see Table 2).
Independent T Test was used to analyze the difference of explicit evaluation of brands from three countries between the high face perception group and low face perception group.The results showed as follows.
Relatively explicit evaluation results showed that explicit evaluation of Japanese brands and American brands was higher in high face perception group than in low face perception group (see Figure 2).Specially, the relatively explicit evaluation of Japanese brands was significantly higher in high face perception group(M=1.29,SD=0.63) than in low face perception group(M=1.07,SD=0.84), t = 1.63, p<0.10, supporting the hypothesis H1a.;The relatively explicit evaluation of American brands was significantly higher in high face perception group(M=1.40,SD=0.51) than in low face perception group(M=1.11,SD=0.76), t = 2.43, p<0.05, supporting the hypothesis H1b.

IAT Data Processing
When designing the IAT procedure, we had an initial set for IAT data: (1) To ensure the fully and accurately using of data, IAT test in this research not only included the S4/S11 (Initial combined task) and S7/S14(Reversed combined task), also included S3/S10 and S6/S13 for practice; (2) To eliminate the extreme values of reaction time, we eliminated all the reaction time which is greater than the 10000ms, or less than 300ms; (3) Calculation formula(Note 2) was set to automatically calculate D. In S3/ S10 and S4/S11, subjects were asked to press the same key when the screen appeared Chinese brand--positive words, Japanese brands--negative word, and American brand--negative words, which were called compatible stage; In S6/S13 and S7/S14, subjects were asked to press the same key when the screen appeared Chinese brand--negative words, Japanese brands-positive word, and American brand--positive words, which were called incompatible stage.D value indicates difference value in compatible and incompatible reaction stage; Positive number means that subjects preferred Chinese brands, while negative number means that subjects preferred Japanese brands or American brands.And the higher the scores, the more positive implicit attitude to Chinese brands, and the more negative implicit attitude to Japanese brands or American brands.
In classification tasks of Chinese and Japanese brands, the D value (M=0.21,SD=0.95) is significantly greater than 0, t = 2.43, p < 0.05.The results showed that subjects' reaction time was significantly shorter when they combined Chinese brands with positive words (combined Japanese brands with negative words) than when they combined Chinese brands with negative words (combined Japanese brands with positive words).Thus, subjects' implicit attitude was more negative towards Japanese brands when compared with Chinese brands.In classification tasks of Chinese and American brands, the D value (M=-0.01,SD=0.95) is close to 0, t = -0.16,p >0.88.The results showed that subjects' reaction time was nearly equal no matter when they combined Chinese brands with positive words(combined American brands with negative words), or when they combined Chinese brands with negative words (combined American brands with positive words).Therefore, subjects' implicit attitude towards American brands and Chinese brands has no significant difference.
As the purpose of this study is to dig into how face perception change the relationship between consumer's implicit attitude and explicit attitude towards Japanese / American brand.Therefore, the above data was processed to get the scores of IAT, and then we used the negative value of IAT scores to be a kind of indicator, which reflects subjects' relatively implicit preference to Japan / USA.

Result of Face Perception on the Relationship between Implicit attitude and Explicit Attitude towards BCO
The results showed that face perception is not related to the implicit attitudes towards BCO.The correlation coefficients between face perception and relatively implicit attitudes towards Japanese brands/American brands were -0.13 and -0.06 respectively.However, face perception was significantly related to the explicit attitudes towards BCO.The correlation coefficients between face perception and relatively explicit attitudes towards Japanese brands/American brands were 0.27 and 0.24 respectively (See Table 3/ Table 4).
In the regression analysis of influential factors which affect explicit attitude towards Japanese brands, the independent variables of regression equation 1 included face perception and implicit attitude.The result showed that the two independent variables both have significant impacts on explicit attitude toward Japanese brands respectively, F (2,126) =7.71, p< 0.01; The independent variables of regression equation 2 included face perception, explicit attitude, and the interactive items of face perception and explicit attitude.The result showed that the three independent variables all have significant impacts on explicit attitude respectively, with the regression coefficient -0.31 (p < 0.05), F (3,125) = 6.58, p< 0.01(see Table 5).Face perception moderated the relationship between implicit and explicit attitudes towards Japanese brands.Thus H2a is supported.
In the regression analysis of influential factors which affect explicit attitude towards American brands, the independent variables of regression equation 1 included face perception and implicit attitude.The result showed that the two independent variables both have significant impacts on explicit attitude towards American brands respectively, F (2,126) =3.73, p< 0.05; The independent variables of regression equation 2 included face perception, explicit attitude, and the interactive items of face perception and explicit attitude.The result showed that the three independent variables all have significant impacts on explicit attitude respectively, with the regression coefficient -0.22 (p < 0.10), F (3,125) = 3.21, p< 0.05(see Table 5).Face perception moderated the relationship between implicit and explicit attitudes towards American brands.Thus H2b is supported.

Conclusions
Firstly, the finding of this research showed that there's clear difference on the explicit evaluation of BCO between consumers with high face perception and consumers with low face perception.Specifically, Consumers with high face perception evaluate brands from Japan/USA more positively than those with low face perception.When consumption becomes a face-saving tool, Chinese had no other choice except imitating face consumption of his/her social groups.Under the pressure of the class and referent group, individual's attitude and buying behavior must be consistent with the group.Individual with strong face perception tends to pay more attention to the extrinsic attributes of product (such as brand) rather than the intrinsic properties (such as quality) to express their social status, reputation and self image.Therefore, compared with consumers with low face perception, those with high face perception showed more positive explicit evaluation of brands from Japan/USA.As to Chinese brand, whether consumers with high face perception or with low of face perception, their explicit evaluation were of no significant difference.
Secondly, the finding of this research also showed that, compared with consumers with low face perception, those with high face perception held more positive explicit attitude towards brand from Japan and USA than that from China.Compared with Chinese brands, Japanese brands and American brands enjoy higher reputations with more expensive price and more noble designing.Therefore, consumers with high face perception have higher evaluation of Japanese brands and American brands.
Thirdly, the finding of this research revealed that consumers' implicit evaluation of Japanese brands was significantly lower than that of domestic brands, which was consistent with previous findings (Li Yanjie et al, 2009;Yang Yang, 2008).And there was no significant difference between consumers' implicit evaluation towards American brands and domestic brands.
Finally, the finding of this research also supported that face perception moderated the relationship between implicit attitude and explicit attitude toward Brands Country Origin (BCO).Specifically, when the face perception is high, the inconsistency between implicit attitude and explicit attitude towards Japanese brands/American brands is higher.When face perception is low, the inconsistency between implicit attitude and explicit attitude is lower.
In conclusion, due to the existing of face perception in Chinese culture, the inconsistency between implicit and explicit attitude has been changed, which then results in the change of the deviation between consumer attitude and consumer behavior.

Theoretical Contribution
First of all, this study not only found that for Chinese consumers there was existing inconsistency between implicit attitude and explicit attitude towards Japanese brands, but also the inconsistency between implicit attitude and explicit attitude towards American brands.This phenomenon, in a certain extent, explained why consumers held a high evaluation on some country's product, but was not willing to buy, or vise versa.In addition, previous Chinese researches on face perception and COO, only stayed in the level of consumer's face perception perceived from different brands originated from different countries.But this research went a step ahead to dig into how consumer's face perception affects his/her evaluation of BCO.The research has enriched the related research about the Country-of-Origin (COO), and provides new ideas to the study of COO Effect.
Secondly, existing previous researches mainly used qualitative method to study the concept of face perception, the difference of face perception and so on.It's the first time to use empirical method to study the impacts of different levels of face perception on consumer behaviors.And it is found that different levels of face perception have different effects on the explicit attitude towards BCO.These results has not only enriched the theoretical study of Chinese face theory, but also broadened the horizons in the research field of how Chinese face impacts Chinese consumer behavior.
Finally, this study found that face perception moderates the relationship between implicit attitude and explicit attitude.When the face perception is high, the inconsistency between explicit attitude and implicit attitude is larger.When the face perception is low, the inconsistency is smaller.The research results have enriched theoretical research about the relationship between implicit attitude and explicit attitude.At the same time, a new perspective was provided to explain the deviation between consumer attitude and consumer behavior under Chinese culture.

Managerial Implication
Through the study of how face perception affect implicit attitude and explicit attitude towards BCO, we highlighted the following two points.
First, marketing managers should pay much attention to enhance consumer's face perception in China market when making out a marketing strategy, because it will improve his/her evaluation of brand.Of course, marketing strategy should be different for brands from different nations.
For Chinese brands, enterprises should integrate various marketing resources on the base of quality, and endow products or services with higher levels of symbolism.Firstly, in terms of advertising, it is a good way to choose the appealing factors to arouse Chinese consumers' Face Collectivism, such as emphasizing the national brand and patriotism.e,g., Haier's "Haier, made in China" and Metersbonwe's "I'm a new national brand"; Another feasible way is to arouse consumers' sense of belongingness, emphasizing the harmony.Secondly, as to spokesperson, on one hand, choosing a national hero (such as Yao Ming, Yang Liwei etc.) as a spokesman can arouse people's sense of national pride.On the other hand, choosing foreign celebrities as spokespersons can shape a high-end brand image of domestic brands.Finally, as to production, domestic company can adopt information communication strategy to deliver the message of the original country of product components, making the positive national image of product components as an appealing point.In addition, Chinese enterprises can also build a plant in the country with a positive national image, or take merger and acquisition of overseas companies and so on.
For Japanese brands, marketing managers should pay attention to the fact that Chinese consumers hold a positive explicit attitude and a negative implicit attitude towards Japanese brand.Japanese companies can change Chinese consumers' brand preferences and improve their face perception to further change consumers' emotional attitude toward Japanese COO through the following strategies.First of all, the advertisement of Japanese enterprise should mainly focus on product quality and function.Secondly, Japanese enterprises need to attach importance to the event marketing, especially when China encounters natural disasters.If the enterprise can take immediate action, it will establish a good corporate image; Moreover, Japanese enterprises can improve their corporate image through charity marketing, or providing sponsorship of public events to create a positive brand association.Finally, Japanese brands can weaken the negative national image associations through establishing multiple nationality brand or brand alliance.
For American brands, marketing managers should be fully aware of the mind share and heart share of a certain American brand in China market.Try every effort to enhance brand image to meet consumer demands.First of all, continuously enhance brand image by improving product quality, value-added services, channels, advertising and marketing campaigns and so on.Secondly, enterprises can provide personalized experience, unique value-added services, through high-end sales channels, build brand network community and off-line brand clubs to meet Chinese consumer demands.Finally, enterprises should avoid cutting price greatly and avoid having quality deficit of products, which would result in Chinese consumers' antipathy.In conclusion, American brands should try to make consumers to evaluate of the brand based on the brand itself rather than the information of its original country.
Second, marketing managers should treat cautiously when hearing the saying of "I love the brand from XX country".It doesn't mean consumers will certainly buy this brand.When assessing the influence of the brand, we can not only judge from the consumers' evaluation, but also from the consumers' buying behavior.In addition, foreign brands need to enhance the face perception bringing to consumers in order to enhance consumers' brand evaluation.However, even doing so, evaluation may differ very far from the real attitude.It requires marketing managers can not only understand consumers' real attitude according to the market situation, but also need to integrate a variety of marketing methods to build consumers' brand preferences, cultivate their positive beliefs towards brand and then change their emotional attitude towards COO to be consistent with their real attitude.

Limitations and Future Research
This study has some limitations.First of all, since the IAT tests need to be performed under laboratory conditions, so we adopted convenience sample.The subjects were undergraduate students and MBA students from a Guangzhou university.There was no comparison study between the two groups, the samples size is not very big but acceptable for IAT test.
Secondly, the measurement scale of image of COO (developed by Parameswaran and Pishaodi, 1994) was used to measure explicit attitude towards Brands Country Origin (BCO).We neither ask the subject to evaluate specific brand one by one, nor match the wordings in IAT measurement.Therefore, future research can adopt a more accurate questionnaire to measure, in order to improve the validity of study results.
Finally, although this study distinguishes the difference between implicit and explicit attitude towards BCO, also reveals the impact of face perception on the difference between implicit and explicit attitude.However, the study did not examine the predictive power of implicit attitude and explicit attitude on consumer purchasing behavior.Future research could further measure the influence of implicit attitude and explicit attitude on consumer purchase intention and the mechanism of face perception in it.

Table 1 .
IAT Experimental Design

Table 2 .
Average Score of Face Perception/Impression of Products from Original Country Notes：* means significant at the level of 0. 5, ** means significant at the level of 0. 01

Table 3 .
Correlation Coefficients between Face Perception and Attitude towards Japanese BrandsNotes：* means significant at the level of 0. 5, ** means significant at the level of 0. 01

Table 4 .
Correlation Coefficients between Face Perception and Attitude towards American BrandsNotes：* means significant at the level of 0. 5, ** means significant at the level of 0. 01

Table 5 .
Stepwise Multiple Regression Analysis of Factors Influencing Explicit Attitude Notes：* means significant at the level of 0. 5, ** means significant at the level of 0. 01, *** means significant at the level of 0. 001.