Trust Discrimination Tendency in Average Citizens at In-nation and Out-nation Levels in Canada , China and the United States

The aim of the study was to explore discrimination tendency expressed in trust towards average citizens at in-nation and out-nation levels. A total of 370 Americans, Canadians and Chinese people, who were not university students, participated in the study. A self-developed questionnaire was used in the research, and the results from this sample showed a trust discrimination tendency. Specifically, trust towards the average citizens of one’s own nation tends to discriminate: the level of Americans’ trust toward Americans was measured as the lowest, the level of trust between Chinese and Chinese was measured as the highest, and Canadians’ in-nation trust was intermediate. People’s trust towards the citizens of other nations tends to discriminate as well: Americans trust Canadians more than they trust Chinese, Canadians trust Americans less than they trust Chinese and Chinese trust Americans slightly less than they trust Canadians. In-nation trust tends to discriminate with out-nation trust: People’s trust towards the citizens of other nations is likely to be either more or less than their trust towards citizens of their own nation. The study found an interesting phenomenon of heterogeneity of in-nation trust and homogeneity of out-nation trust: the variances of in-nation trust tend to be greater than those of out-nation trust. Furthermore, when participants were asked how they trust the average citizens of other nations, they seem to have relatively more similar opinions for those citizens in generally highly trusted nations; however, they seem to have relatively more diverse perspectives for those citizens in generally low trusted nations.


Introduction
The purpose of the study was to explore average citizen trust tendency of Americans, Canadians and Chinese, including how they trust their own average citizens and how they trust other nation's average citizens, as well as related similarity and dissimilarity."Citizen" is a collective term that, in practice, means residents at large, the public, or constituents (Whitener, 1997).For the sake of convenience, and brevity, the terms "Americans" (the people who live in the United States of America), "Canadians" and "Chinese" are used, rather than "American citizens", "Canadian citizens" and "Chinese citizens".When both those trusting (trustors) and those being trusted (trustees) are within the same nation, this form of trust is called in-nation trust, e.g.Americans trust Americans, Canadians trust Canadians, and Chinese trust Chinese.When the trustors and trustees are in different nations (e.g.Americans trust Canadians and Chinese, Canadians trust Americans and Chinese, Chinese trust Americans and Canadians), this form of trust is called out-nation trust.
Trust is a rather broad concept that has been defined in many disciplines, with various focuses and from different perspectives.In general, the majority of scholars are of the opinion that trust is a belief or confidence about another party's integrity (including reliability, predictability, and dependability) and/or benevolence (including goodwill, motives, intentions and caring) (Mayer, Davis, & Schoorman, 1995;Nooteboom, Berger, & Noorderhaven, 1997;Ross & LaCroix, 1996).Based on the aims and methods of this study, and according to the basic ideas of the majority of trust concepts, trust is defined as a positive expectation, confidence and belief about other people, other groups and other nations' attempts, attitudes and behaviours.
Canada is the neighbour and firm political ally of the United States.These two countries are bound together by countless ties (Carroll, 1992).It was felt that it would be interesting to investigate whether the level of trust by Canadians towards Americans reflects their unique relationship with each other, as well as whether the level of trust by Americans toward Canadians reflects the ally relationship.China, as an emerging power, is often considered within American society as the main rival of the Unites States and appears to be threatening America's top power position (Zhang, 2005).At this point, the level of trust of Chinese towards Americans may be an objective frame of reference to determine general Chinese attitudes toward the USA and vice versa.
We were of the opinion that how citizens of each of the three nations trust their own people could provide not only in-nation trust information, but also a frame of reference for comparing out-nation trust.That is, the degree to which Americans trust Americans could be used as a basis to infer whether Americans trust other nations less or more.The same applies to Canadians and Chinese.In addition, the study also examine how similar or how dissimilar the out-nation trust is between Americans, Canadians and Chinese.
A study conducted by Brewer, Aday and Gross (2005) answered the question of whether Americans trust other nations.They analysed original pane survey data to determine Americans' generalised beliefs about how much the United States can trust other nations.In their results, only about one-fourth of Americans in their sample thought that the United States could trust other nations.This indicates that most Americans tend to place little trust in other nations (Brewer, Aday & Gross, 2005).In their study, "other nations" was a general concept, rather than referring to a special country.But it is reasonable to hypothesize that Americans' may have different trust in different nations.So, Unlike Brewer et al.'s study, our study used American in-nation trust to identify the level of Americans trusting other nations, and also compared two other specific nations (e.g.Canada and China), rather than the general reference to "other nations".
In reviewing the related research literature on trust at national level, we found that most of the studies did not cover "each other trust", even though some of them included five or more nations (Huang & Dastmalchian, 2006;Huff & Kelley, 2005).In those studies, findings on trust were given from the point of view of individuals within a nation, and did not specify how a nation trusts other nations.Huff and Kelley (2003) selected mid-level managers from large banks in Japan, Korea, Hong Kong, Taiwan, China, Malaysia and the United States to test internal trust (trust within the organisation) and external trust (an organisation's trust towards suppliers, customers, etc.).The study compared the trust levels within each nation, not cross-national levels of trust.Similarly, Yoshino (2002) conducted a study in Britain, France, Italy and the United States in which participants were asked whether they felt that most people could be trusted, or that they could not be too careful in dealing with people.It was found that rates of trust in the United States and Great Britain were high, whereas in Italy and France, they were low.The higher trust levels obviously did not imply a higher level of trust by Americans towards the British, French and Italians; neither did the lower level of trust mean a lower level of trust by Italians towards Americans, British and French, and so on.A study conducted by Yuki, Maddux, Brewer and Takemura (2005) examined group-based trust in America and Japan.Even though different nations were studies, the type of trust explored was in-nation trust (within their own country), and not out-nation trust.Furthermore, in many studies which focused on trust regarding domestic issues, those conducted on the trust of average citizens are rare.Subjects of these studies are very diverse, such as decision makers (Smith, Larimer, Littvay, & Hibbing, 2007), supervisees (Atuahene-Gima, & Li, 2002), patients (Sohler, Fitzpatrick, Lindsay, Anastos, & Cunningham, 2007), younger and older people (Rossotti, Winter, & Watts, 2006) and strangers (Axelrod, 2004).A few trust studies can be found whose subjects are citizens.One was conducted by Yang (2005), and another by Carter and Bélanger (2005).Based on a survey of 320 public administrators, Yang's study found that public administrators generally have a neutral (neither trustful nor distrustful) view of citizens.Here the trust is administrator towards citizen, rather than citizen towards citizen.As for the study by Carter and Bélanger, they explored citizen trust, but only the trust towards utilisation of e-government services, not towards other citizens.Thus, some direct and clear information on in-nation and out-nation trust by citizens is lacking.This study is original in that it not only explored trust between average citizens at in-nation level, but also at out-nation level.Moreover, the out-nation trust is specified and targeted.
Hominoid social systems are predisposed to organise themselves within some range of group-based inequality; these groups may be defined on the basis of race, sex, nationality, ethnicity, social class, skin color, or any other group distinction that the human mind is capable of constructing (Sidanius & Pratto, 1999).This is true at intranational as well as international level.People are placed in non-egalitarian and hierarchically structured relationships among social groups (Sidanius & Pratto, 1999).The group/nation-based hierarchies both within and between Canada, China and the United States differ given their different political and economic systems, histories, cultures and traditions, as well as relationships to others.The general framework which seems most reasonable to assume is that trust in such situations and conditions tends to be discriminated, because as a cultural and historical phenomenon, trust behavior may be substantially different in various social and cultural systems (Fukuyama, 1995).
There are generally three parts to a trust process to form a trust set: trustor, trustee and "with respect to" (Hardin, 2006).In theory, the trustor, trustee and "with respect to" have unlimited possibilities.Even if only one part of a trust set differs from only one part of other trust sets, the constructs of trust are likely to differ from each other.In our study, trustors are general citizens of Canada, China and the United States, trustees are their own citizens of the three nations and the citizens of other two nations, and "with respect to" refers to international issues.It is believed that as long as trustors and trustees in the trust process are different from each other at in-nation level and also at out-nation level, a trust discrimination tendency is expected.
Trust, as the property of a relationship (Sydow, 1998), is based on relationships (Boisot & Child, 1996;Child, 1998;Tyler, 1998).Our relationships have become both more varied and more numerous than ever, as described by Hardin (2006).Under these preconditions, trust discrimination tendency is more likely to match a discriminate trust premise than a trust unanimity tendency.In other words, the probability of discriminate preconditions corresponding to discriminate trust is much greater than the probability of discriminate preconditions corresponding to unanimous trust.The satisfaction and importance of relationships also affect trust (Holmes & Rempel, 1989).Among the three nations, the three relationships between each pair (e.g.Canada vs. China, Canada vs. the United States, and China vs. the United States) differ considerably.The different relationships, in general, may elicit different levels of trust, so out-nation trust is more likely to be revealed in a discriminate manner.
Many previous research studies on trust have pointed out discriminating trust, rather than trust unanimity, in all kinds of situations.For example, Fukuyama (1995) identified Japan, Germany and the United States as high-trust societies, and France, Italy and China as low-trust societies.Huff and Kelley (2003) suggested that the trust propensity and organisational external trust in the United States is higher than in Japan, Korea, Malaysia and China.Vishwanath (2004) reports that Canadian auction participants tend to trust the sources more and are therefore more willing to participate in an online auction.German auction participants exhibit modest levels of interpersonal trust when they choose a product online, and France exhibits the lowest levels.Brewer and Campbell (1976) also found that the in-group was favoured over all out-groups in dimensions such as trustworthiness, obedience, friendliness and honesty in their 10-year study.These studies covered cross-national level (Buchan, Croson, & Dawes, 2002;Yuki, Maddux, Brewer, & Takemura, 2005), cross-group level (Sutter & Kocher, 2007;Wang & Yamagishi, 2005;Weyman, Pidgeon, Walls, & Horlick-Jones, 2006), cross-individual level (Hiraishi, Ando, Ono, & Hasegawa, 2004;Van Overwalle & Heylighen, 2006), and so on.
Given the above, the plentiful research, various trustors and trustees, different issues covered by "with respect to", the hierarchy group-based unequal societies, the changing factors and the discriminating relationships provide trust discrimination preconditions.People therefore seem to trust someone, some groups, some nations, some things more, and others less.Some researchers even believe that trust varies with the "up and down of daily life" (Uslaner, 2003).
We refer to this phenomenon as trust discrimination -the difference in trust level for one specific social group versus another social group with which it is naturally coupled (i.e.trust in males versus females, wealthy versus poor people, etc.).
We thus posit the following trust discrimination tendency hypotheses: 1. In-nation trust has discrimination tendency, specifically, the manner in which Americans trust Americans, Canadians trust Canadians and Chinese trust Chinese tend to differ.
2. Out-nation trust has discrimination tendency too, specifically, the manner in which Canadians trust Chinese is unlike the manner in which they trust Americans, the manner in which Chinese trust Americans is unlike the manner in which they trust Canadians, and the manner in which Americans trust Canadians is unlike the manner in which they trust Chinese.
3. In-nation and out-nation differences tend to discriminate.For instance, the manner in which Chinese trust Chinese is likely to be distinct from the manner in which they trust citizens of other nations, the manner in which Americans trust Americans is likely to be distinct from the manner in which they trust citizens of other nations, and the manner in which Canadians trust Canadians is likely to be distinct from the manner in which they trust citizens of other nations.
4. I trust you and you trust me has discrimination tendency, which means the trust of nation A toward nation B would be different from the trust of nation B toward nation A.

Participants
Quota sampling method.To ensure the comparability of the samples across the three countries, our subjects in each country were recruited by means of the same quota sampling technique.An average distribution for the sample was in the same categories, namely age (19-30 years, 31-50 years, 51+), gender (male and female), and socio-economic status (SES) (higher and lower SES).It was observed that the quota sampling for age and gender could be done easily, but it was not easy to determine SES.We therefore tried to achieve a balance between higher and lower status.
Participant recruiting.Unlike the majority of studies in psychology, our subject pool was made up of ordinary citizens, rather than highly pre-selected university students.These participants not only provided greater variance in individual levels of exposure to trust, but also more accurately reflected the influence of culture on society as a whole.This study recruited a total of 370 participants with a mean age of 33.43 in Canada, China and the United States.A total of 127 were Americans (63 male and 64 female) with a mean age of 34.46; 123 were Canadians (63 male and 60 female) with a mean age of 32.30; 120 participants were Chinese (60 male and 60 female) living in the People's Republic of China with a mean age of 33.54.
The participants were recruited in the greater Boston area (the United States), greater Ottawa area (Canada) and Kunming city as well as surrounding areas (People's Republic of China).
The principle of recruiting participants, namely choosing as many places as possible and at the same time, making sure that average citizens had public access to them, was complied with.These places included cafeterias, cafés, tea-shops, public libraries, bus and train stations, newspaper viewing boards, food courts in shopping malls, and so on.

Materials
We believe and agree that trust is a three-part relation: A trusts B to do, or with respect to, C (Hardin, 2006).In our questions, we therefore needed to mention something people could think about as a respect to.We first tried to find an event/thing/issue which would have similar importance and relevance across the three nations.However, it was found that some issues are more important for one nation, but not for others.For example, the Iraq war, even though involving the three countries in some way, still seemed more related to America.More attention is paid to the 2009 Olympic Sports Games in China than in North America, even though the games are international.Due to the difficulty of finding such a neutral event/thing/issue equal in importance for the three countries, it was decided to use the general term "in regard to international issues", rather than mention something specific.
Our questionnaire needed to cover both in-nation and out-nation trust.Using the questionnaire for Americans as an example, our questions were posed in the following ways: "How likely would you trust the average American citizen to do what is right in regard to international issues?","How likely would you trust the average Canadian citizen to do what is right in regard to international issues?",and "How likely would you trust the average Chinese citizen to do what is right in regard to international issues?"In the Canadian and Chinese versions of the questionnaire, questions were first asked about the trust of Canadians and Chinese, respectively, and then about the other two nations' average citizens.
A seven-point response scale was used, with 1= strongly distrust, 2 = distrust, 3 = slightly distrust, 4 = neutral, 5 = slightly trust, 6 = trust, to 7 = strongly trust.After each question, participants were asked to circle a number (any one of the 7 numbers) that best expresses their degree of trust.
The questionnaire was written in both English and Chinese.The Chinese version was created by the first author, who is fluent in both languages.A great effort was made to ensure that the two versions of the questionnaire were equivalent in terms of the meanings and relevance in the two cultures.Moreover, the Chinese version was also checked by a professional linguist.The English version was used for both the American and Canadian participants.
In the instructions for the questionnaires, it was made particularly clear that the information participants provided would be kept totally anonymous.There were no right or wrong answers, or good or bad answers.Participants were encouraged to express their own thoughts and ideas, not what they thought would be generally desirable.It was believed that this information would control influence on participants' response preferences.

Procedure
A researcher was placed in different randomly selected public locations.When a potential participant appeared, the researcher, who can speak both English and Chinese, approached him/her.After the person listened to a short explanation of the aim, time and context of this research and expressed willingness to take part voluntarily, a questionnaire in the appropriate language was given to the participant, participation response rate was about 65% in American and Canada, 80% in China.While the participant filled in the questionnaire, the researcher waited nearby, ready to answer questions.

Results
Results regarding both in-nation and out-nation trust are shown in Figure 1 and Table 1.

In-nation trust
In Figure 1, the darker colour columns represent in-nation trust, in other words trust by Americans towards Americans (left), by Canadians towards Canadians (middle) and by Chinese towards Chinese (right).The trust mean of Americans trusting Americans was 3.96 (SD = 1.43),Canadians trusting Canadians was 4.77 (SD = 1.43), and Chinese trusting Chinese was 4.86 (SD = 1.32).Among the three trust means, there was no significant difference between the one for Chinese and the one for the Canadians, t(241) = .5, p > .05.However, the trust mean for Americans was the smallest with significant difference from that for Chinese and Canadians (t(245) = 5.11, p < .001;t(248) = 4.74, p < .001).

Out-nation trust
The lighter colour columns of Figure 1 indicate trust means for out-nation trust, in other words, trust by Americans towards Canadians and Chinese (left), by Canadians toward Americans and Chinese (middle) and by Chinese toward Americans and Canadians (right).The trust mean of Americans trusting Canadians (M = 4.55, SD = 1.07) is significantly higher than that of Americans trusting Chinese (M = 4.02, SD = 1.14), t (126) = 4.83, p < .001.The trust mean of Canadians trusting Americans (M = 3.69, SD = 1.42) is significantly lower than that of Canadians trusting Chinese (M = 4.05, SD = 1.18), t(122) = 2.50, p < .05.The trust mean of Chinese trusting Americans (M = 4.51, SD = 1.10) is not significantly different from that of Chinese trusting Canadians (M = 4.64, SD = 1.04), t(119) = 1.42, p > .05.
When we averaged each pair of out-nation trust means, it was found that the Chinese out-nation trust level (average level of trust towards Americans and Canadians) was the highest (M = 4.58, SD = .95).The Canadian out-nation trust level (average level of trust towards Americans and Chinese) was the lowest (M = 3.87, SD = 1.04).The American out-nation trust level (average level of trust towards Canadians and Chinese) was in the middle (M = 4.28, SD = .916).The three trust extents differ significantly from each other (ts are 5.515, 2.45 and 3.346 respectively for comparing Chinese and Canadians, Chinese and Americans, and Canadians and Americans, ps <.05).
However, the trust means for Americans showed a different pattern.Specifically, the trust mean of Americans trusting Americans was 3.96 (SD = 1.43), which is significantly lower (t (126) = 4.11, p < .001)than that of Americans trusting Canadians (M = 4.55, SD = 1.07).But, the in-nation trust mean for Americans is not significantly different from the trust mean of Americans trusting Chinese (M = 4.02, SD = 1.14), t(126) = .41,p > .05.

I trust you and you trust me
It showed us that Americans trust Canadians more than Canadians trust Americans (t(248) = 5.49, p < .001),and Americans trust Chinese less than Chinese trust Americans (t(245) = 3.451 , p < .001).Similarly, Canadians trust Chinese less than Chinese trust Canadians (t (241) = 4.16, p < .001).All the differences of the three reciprocal trusts were significant.

Variances of in-nation trust and out-nation trust
This issue was not originally included in the study, but its significance became apparent during the process of analyzing data.We found two interesting phenomena worthy of mention.
Firstly, according to the box boundaries which are to the 25 th and 75 th percentile of each trust distribution in Figures 2, 3 and 4, the variation spread for in-nation trust is larger.
Secondly, for out-nation trust chosen number distributions, it seems that the lower trust mean is associated with the larger ranges of trust chosen number distribution.For instance, the trust means of Americans trusting Chinese and Canadians trusting Americans are smaller, and their trust box plot spread is larger (see Figures 3 and 4, respectively).Using Pearson correlation to test the relationship between out-nation trust mean and corresponding standard deviations, it was found that the relationship is negative at a significant level (p = .014,< .05),indicating that the smaller the trust mean, the larger the trust number variances.

In-nation trust
Our result, that Americans display the lowest level of trust towards their own citizens, and Chinese and Canadians have a higher level of trust towards their own citizens, supports the first hypothesis, namely that in-nation trust has a discrimination tendency.It shows that even though the targets of trust are the same, trusting one's own nation's citizens, the trust levels in different nations differ.At the same time, this fact of in-nation trust also provides a frame of reference for comparing out-nation trust.
Our result that Americans display the lowest level of trust towards their own citizens, Chinese and Canadians have higher level of trust towards their own citizens supports the first hypothesis, namely that in-nation trust has discrimination tendency.We can see that although trusting their own citizens, people of each nation display different trust level.At the same time, this fact of in-nation trust also provides a frame of reference for comparing out-nation trust.

Out-nation trust
The second hypothesis of the study is also supported by the study data.Out-nation trust in our study has discrimination tendency.From each nation's perspective, average citizens of a nation tend to trust citizens from other nations differently.For example, Americans trust Canadians more than they trust Chinese.This decreased amount of trust towards Chinese parallels with certain debates among scholars that China is the country most frequently named as the US's main adversary after the cold war, making it the most likely object for the transference of hostility (Murray & Meyer, 1999).Future US-China relations are likely to be tested by more conflicts, which means an even more bumpy road ahead (Dumbaugh, 2005;Jian, 2001).
Canadians are likely to trust Americans less than Chinese.This result may be a surprising one.It has been opined that Canada and the US are practically synonymous, both share the same values and ideals at home and abroad, their economies are inextricably linked, the cultures and people are indistinguishable for the most part, and the two countries depend on one another for their mutual security (David, 2002).Future Canada-US relations are considered very optimistic, as the relationship between the two nations has always been very good (King& Blanchard, 2006).Our results, however, seem to sound some contradictory voice in term of reciprocal citizen trust.
Chinese trust Americans and Canadians similarly without statistically significant difference.This finding is not unexpected, since the relationship between China and the US seems more conflicting and more likely to be considered as threatening compared to China's relationship with Canada (Dumbaugh, 2005;Harding, 1992).
In addition, if the mean of the average amount of trust towards other nations is considered (e.g. the level of trust of the average American towards Canadians and Chinese as a general American out-nation trust level), it is found that the level of trust towards other nations differs in each nation.In this study, Canadians are the least likely to trust other nations' citizens and Chinese are most likely to trust other nations' citizens.This means that the trust between national citizens tends to discriminate.

In-nation and out-nation trust
The third hypothesis, namely that in-nation and out-nation differences tend to discriminate, is also supported by the results of this study.People are less likely to trust their own nation's citizens and other nations' citizens in the same way.Americans, Canadians and Chineses all express discrimination tendencies in in-nation and out-nation trusts.The discrimination is displayed in two different ways: Firstly, Canadians in the sample trust other Canadians more than they trust Americans and Chinese.Similarly, Chinese in the sample trust other Chinese more than they trust Americans and Canadians.These in-nation trust and out-nation distrust tendencies do indeed reflect "in-group love and out-group hate" (Brewer, 1999).Secondly, the level of trust by Americans towards Americans is lower than that by Americans towards Canadians, and the same or at least not higher than that by Americans towards Chinese.Americans' in-nation trust does not express "love" and out-nation distrust does not express "hate"; on the contrary, it seems to be in-nation distrust and out-nation trust.
This contrasts with the opinion that discrimination between in-group and out-group is a matter of relative favouritism toward the in-group and absence of equivalent favouritism towards out-groups (Brewer, 1999).Another in-group trust study among Australians and Japanese conducted by Kiyonari, Foddy and Yamagishi (2007) also reports that 27 of the 82 participants trusted the in-group trustee, while 29 trusted the out-group trustee.There was no difference between in-and out-groups.In the current study, neither "in-nation trust and out-nation distrust" nor "in-nation distrust and out-nation trust" can express the whole trust situation, since the American model does not fit the former notion, and the Canadian and Chinese models do not fit the latter notion.In such situations, the idea of trust discrimination tendency (TDT) can fit all models.

I trust you and you trust me
It is rather clear to see the different levels for the trust of each other.In our sample, the degrees to which Americans trust Canadians and Canadians trust Americans differ.The degrees to which Americans trust Chinese and Chinese trust Americans also differ.Canadians' trust of Chinese is not at the same level as the Chinese trust of Canadians.

Are the Chinese a less trust nation?
Our results tell us that for in-nation trust, the trust level of the Chinese of our sample is the highest of the three countries, and at the same time, no other nation in our study showed more out-nation trust than China.Figure 1 clearly illustrates this point.From our samples, and based on responses to our specific questions of in-nation and out-nation trust, we would say that our results challenge Fukuyama's (1995) classification of China as representing a low-trust society, because in our study Chinese not only trust their own citizens the most, but also trust other nations' citizens the most.More explorations are needed to better explain the contradiction between Fukuyama's conclusion and our results.

Are Americans less likely to trust other nations?
As mentioned previously, some research studies indicate that Americans seem to place little trust in other nations (Brewer et al., 2005).However, our results show that Americans are likely to place more trust in Canadians, and at least place the same trust in Chinese as they do in their own citizens.By contrast, both Canadians and Chinese put less trust in other nations than their own.Furthermore, the absolute values of trust towards other nations indicate that levels of trust by Americans are not any less than those of Canadians.So, one could infer that Americans are not likely to trust other nations less.

Trust criteria
What would determine whether one nation trusts other nations less or more?And what would decide whether a nation is a higher or lower trusted nation?To answer these two questions, we suggest that an objective reference is needed.There may be more than one criterion for the reference.We would like to suggest that a logical and feasible one is using in-nation trust level as a judgement reference for out-nation trust level, for instance using the American in-nation trust level to judge the trust level of Americans towards other nations.In addition, the levels of both in-nation and out-nation trust can be used to determine whether a nation is a higher or lower trust nation, for instance using Chinese in-nation and out-nation trust levels to judge whether China is a lower trust nation among the three nations.

In-nation heterogeneity and out-nation homogeneity
It is a little surprising to see that people's opinions regarding how likely they would trust their own people are rather variable.People seem less likely to have a similar opinion regarding trust towards their own citizens.The opinions have different trust tendencies.Comparatively, out-nation trust tends to be less variable.People, in general, are more likely to have a comparatively similar trust pattern towards people of other nations.We prefer to use the terms "in-nation heterogeneity" and "out-nation homogeneity" to indicate this degree of trust.It is likely that regarding to trust own nation's citizen, the ideas of people tend to be more diverse, variable and different, while regarding to trust other nations' citizens, the ideas of people tend to be more identical, similar and consistent.

Higher trust homogeneity and lower trust heterogeneity
When considering out-nation trust in our study, our results show that the largest trust distribution range appears in Americans trusting Chinese and Canadians trusting Americans.These are also the two lowest trust levels among all six out-nation trusts.These facts together with the statistically significant test suggest that the level of being trusted is likely to be negatively associated with the diverse degree of trust judgement and opinion.The higher the level of being trusted, the smaller the trust diversity, and vice versa.
We are hoping that the crude and original opinions of "in-nation heterogeneity and out-nation homogeneity" and "higher-trust homogeneity and lower-trust heterogeneity" may elicit both new and mature ideas, comments to further explore the interesting phenomena.
It is believed that trust discrimination is not always and automatically an absolutely bad thing.Langfred (2004) conducted research entitled "Too much of a good thing?Negative effects of high trust and individual autonomy in self-managing teams", and pointed out that "under some conditions, too much trust in a self-managing team can be harmful" (Langfred, 2004, p.390).The idea posted by Sztompka (1999) is also appreciated: It is functional to trust the trustworthy, and it is equally functional to distrust the untrustworthy.Furthermore, according to social Darwinism, social inequality is justified as "survival of the fittest" (Greene, 2003).Trust discrimination can be considered as a sort of social inequality that insures survival of the fittest.In the complex and various situations arising in non-egalitarian hierarchically structured relationships (Hardin, 2006;Sidanius & Pratto, 1999), unanimous trust may not be expected; rather, various, hierarchical and discriminating trust can embody and reflect corresponding complex and various characteristics.

Conclusion
The current results from the sample of the average citizens of the three nations in this study suggest that as a whole, trust possesses a discrimination tendency.This discrimination tendency is reflected in in-nation trust.The level of each nation's trust of its own citizens tends to be different.This trust discrimination tendency is also reflected in out-nation trust.People usually tend not to trust other nations' people similarly (e.g., Americans trust Canadians more than Chinese, and Canadians trust Chinese more than Americans).In-nation and out-nation trust also tend to differ by either in-nation trust being higher than out-nation trust, such as we see in Canada and in China, or in-nation not being higher than out-nation trust, e.g.Americans.Last, trust of each other tends to discriminate too: for example, Canadians trust Americans less than Americans trust them, and Chinese trust Americans more than Americans trust Chinese.
The greatest meaning of our results is that trust should not be considered stable and unchangeable any more.On the contrary, trust is changeable, and it has a discrimination tendency.This characteristic of trust has not been discussed previously and we believe it may discourage understanding of trust.The discrimination tendency of trust will help us realize the inevitability of varied trust levels in different cultures and towards different targets, and therefore help us understand relationships better.It should also improve readiness for resolving potential conflicts and contradictions among various nations and groups.
The current study also found the phenomena of heterogeneity of in-nation trust and homogeneity of out-nation trust, which means that the variances of in-nation trust are likely to be greater than those of out-nation trust.People of a nation usually have more diverse and variable trust attitudes toward their own citizens, and tend to have less diverse and more similar trust attitudes toward other nations' citizens.Furthermore, people seem to have relatively more similar opinions about those in generally highly trusted nations.However, people seem to have relatively more diverse opinions about those in generally less-trusted nations.

Figure 1 .
Figure 1.Trust means of in-nation trust and out-nation trust for Americans (left), Canadians (middle) and Chinese (right)

Table 1 .
. Trust means, standard deviations and paired differences of trust means for the three nations