In-group and out-group perspectives A cross-cultural comparison of four countries

As the interrelationships among countries are increasing in various social, political, educational, and commercial contexts, it would be beneﬁcial to understand how people of different cultures perceive the links of human relationships and how they present different behaviors toward in-group and out-group individuals. By conducting questionnaire, the present study aims to explore the differences in in-group and out-group perceptions and behavioral patterns of private organization employees in South Korea, Japan, Canada, and the United States. Statistical analysis of the collected data yielded cultural differences of human relationships based on in-group and out-group distinctions. To obtain more generally applicable ﬁndings, however, further investigation should examine other factors such as age, geographical locations, education, and occupations.

With the increasing amount, frequency, and variety of international interactions among countries, scholars have examined various cultural subjects to recommend the benefits of better communication skills and strategies for successful outcomes.To provide general and detailed guidelines of how to function effectively in intercultural environments, researchers explored topics such as individualism and collectivism dimension (e.g., Hofstede, 1980;Oyserman, Coon, & Kemmelmeier, 2002), conflict

In-group and out-group
In-group and out-group perspectives are closely tied with individualism and collectivism dimension of cultures (Triandis & Gelfand, 1998).Individualism is known to be more prevalent in Western countries including the United States (Vandello & Cohen, 1999) and Canada (Li, 2001) where people place more importance on individual needs and identity over group needs and harmony (Ting-Toomey & Chung, 2012).On the other hand, collectivists "identify themselves as embedded in groups and relationships rather than as separate from others" (Coon & Kemmelmeier, 2001, p. 349).Collectivism is a common value in East Asian countries such as South Korea (Lee, 2012) and Japan (Kobayashi, Kerbo, & Sharp, 2010) where harmony among in-group members including family, friends (Samovar et al., 2017), and colleagues (Jandt, 1998) is very important.
Behavioral biases of in-group and out-group have been explained by in-group virtues and out-group vices (Schaefer, 1998), in-group love and out-group hate (Weisel & Bӧhm, 2015), in-group favoritism (Balliet, Wu, & De Dreu, 2014) and out-group rejection (Nawata & Yamaguchi, 2014) or out-group prejudice (Aboud, 2003).That is, certain behaviors of the in-group can be perceived as virtuous by in-group members while the same practices can be recognized as unacceptable if done by out-group members (Neuliep, 2000).For instance, eating food with the hands is taken as a natural cultural practice by people in some countries while people in other countries might see it as unsanitary and unhealthy.
While North Americans do not tend to make a clear distinction between in-group and out-group members (Yum, 2000), East Asians are inclined to draw explicit boundaries between themselves (Chen & Starosta, 1998) and keep social and work group relationships long-lasting and unchanging (Lustig & Koester, 2006).That is, individualists are very good at entering and leaving groups (Triandis et al., 1993) and have skills in making relatively superficial (Triandis et al., 1988) and non-intimate relationships (Triandis, 1991) with other members of a group.On the other hand, collectivists spend much time establishing relationships (Triandis, 1991) and mutual dependence (Yum, 2000) on interpersonal obligations (Triandis et al., 1988) that takes precedence over responsibility to oneself (Lustig & Koester, 2006).This behavioral perspective stems from the idea that the members of collectivistic cultures hold the belief that the basic unit of life is the collective or group, while individualists believe the basic unit of survival is the self (Hui & Villareal, 1989).

Importance of understanding in-group and out-group perceptions
The present study began with an idea that most people, if not all, want to enjoy effective, successful, constructive, and healthy intercultural encounters.The next step was to carry out research on how people of different cultures shape relational boundaries and apply behavioral patterns to people of different levels of intimacy.Because "definitions of in-groups and out-groups differ widely across cultures" (Neuliep, 2000, p. 176) and distinguishing in-group from out-group is a universal human tendency (Brewer & Campbell, 1976) of which "strangers and foreigners have the lowest weight for grouping" (Chang & Chang, 1994, p. 55), it is imperative to see how people of different cultures categorize others in various social interactions.Acknowledging these arrangements could help people better prepare for more positive cultural encounters.Moreover, a huge challenge associated with increased multicultural encounters requires accurate understanding of cultural perceptions including relationships with co-workers, friends, social members, and even intercultural relatives.If clear interpretations of in-group and out-group relationships with social and cultural counterparts are recognized, fewer misunderstandings and more productive mutual outcomes would be possible.
Previous researchers have explored in-group and out-group identity and behaviors mainly from the perspectives of social identity theory (e.g., Marques, Robalo & Rocha, 1992), self-categorization theory (e.g., Hogg & Reid, 2006), intergroup contact theory (Pettigrew & Tropp, 2006), and expectancyviolation theory (Bettencourt et al., 1997).These theories were examined by using a meta-analytic test (e.g., Pettigrew & Tropp, 2006), normative-ethical analysis (e.g., Horak & Klein, 2016), stereotype content model (Lee & Fiske, 2006) and other methods of analysis.With these efforts, scholars have offered diverse characteristics (e.g., Weisel & Bӧhm, 2015), cultural roots (e.g., Yum, 2000), and behavioral patterns (e.g., Brewer, 2007) of in-group and out-group perspectives to help people better understand cultural variabilities.Not many studies, however, examine who might be included as in-group members and who might be treated as out-group members by people of different cultural backgrounds.
those companies or who know the executives with study participation decision authority for those organizations.Participating organizations were selected on the basis of size and geographical locations.
Two hundred fifty (250) copies of the questionnaires were sent to each organization for a total of 2,000 copies to eight organizations.One thousand three hundred ninety seven (1,397) questionnaires were completed and returned, and 1,352 were statistically analyzed.Thirty-one questionnaires were excluded from the analysis because of missing demographic information.In addition to those 31 individuals, 14 other participants were excluded from the analysis due to nationality issue.All South Korean participants were South Korean citizens, but there was one American citizen in a Japanese company.Also, there were two Germans, one American, one Japanese, and one Indian in Canadian organizations.In the United States, there were three Chinese, two Germans, two Indian, and one Canadian in the participating organizations.People working or residing in foreign countries might have adapted to a new culture and their original cultural perspectives and behavioral patterns could have changed.Thus, all 14 people working in foreign countries were excluded from the statistical analysis to prevent possible contaminating effects.
Table 1 displays the specific number of participants by countries, organizations, and gender.Among the 1,352 respondents included in the statistical analysis (N = 842, males; N = 510, females), the number of respondents representing each country and organization varied.* City with less than 300,000 people in population ** City with more than 1,000,000 people in population Japanese participants were the oldest group (M = 38.8years; SD = 15.7)followed by Canadian respondents (M = 35.6 years; SD = 14.2) and Korean participants (M = 34.8years; SD = 15.2).American respondents were the youngest group (M = 31.4years; SD = 13.6).The oldest respondent was a South Korean director of a company in a small city (56.5 years), and the youngest was an interning college student in a bank in a small city in the United State (21.2 years).Among all participants from four countries, 784 (58%) have four-year college degrees in various majors.Social sciences (N = 342), natural sciences (N = 228), and engineering (N = 119) were the three most popular fields of study for participants.Ninety-five remaining respondents indicated different academic majors.

Procedure
Before the completion of the questionnaire for the present study, a pilot study was conducted with 148 participants who were not included in the main study (South Korean, N = 43; Japanese, N = 34; Canadian, N = 32, American, N = 39).These participants completed the questionnaire in Korean, Japanese, or English versions of the questionnaire before the conduction of back translation procedure.
Following the pilot study, the main questionnaire was composed with translation and the back translation method (Brislin, 1970) to maximize translation fidelity.
Because the participants might apply different inclusions and definitions to groups, each of the eight groups was described in the questionnaire.This was done to eliminate the confusions the participants might face during the selection process in the first part of the questionnaire.Immediate family included parents and spouse's parents of the participants, the spouse and children.If the respondents were not married, parents and unmarried siblings were used as their only immediate family.Relatives meant bloodtied members and in-laws who are not immediate family including married siblings, cousins, uncles, aunts, nephews, and nieces.Because the concepts of "siblings" differ widely among cultures (Adams, 1999) and it is advisable to use clear definitions of siblings in cultural studies (Cicirrelli, 1994) including marital status of siblings (White, 2001), this issue was considered and specified when groups are suggested for selections.School alums are those, whether they maintain contacts or not, who have graduated from same schools during the participants' education.Regional members are people who are born or reside in the same city or province as the respondents.Co-workers include those who work together in the same company and meet regularly for work assignments and accomplishments.Religious members are those who attend the same church, temple, or synagogue for worship and religious activities.Social members are those who spend time together for special interest such as volunteer work, sports activities, and learning groups.Finally, friends are those people the respondents might have met through school, work, military or religion who sustain close ties and friendship.
Although the second part of the questionnaire included four sub-divisions, items were not indicated or grouped with headings.This was done to hinder the participants from being influenced by possible response biases based on the categories.That way, it was expected to collect more accurate responses of what the respondents actually think rather than what they were anticipated to present as more socially acceptable answers.Also, the consent form of the questionnaire included information to ask the participants not to interact, share ideas, ask for other respondents' thoughts, or discuss the questions with other participants.This was done to prevent any influence among participants while completing the questionnaire.

Method of Analysis
Two sets of one-way analysis of variance (ANOVA) were used to analyze the data for the present study.
The first set was used to compare the selection of in-group and out-group members indicated by the participants.The second set was utilized to measure the behavioral patterns and perceptual viewpoints of the respondents toward in-group and out-group individuals.In the second section of the questionnaire, thirty-two items were the dependent variables (criterion variables), and each country was the independent variable (explanatory variable).As the ANOVA is an adequate method to measure the significance of responses "involving three or more groups" (Stacks & Hocking, 1999, p. 382), this statistical test was used to analyze cultural differences of four countries.An average score for each item selected by the participants in each nation was the unit of analysis.
Significant F-ratios were conducted before the Duncan Multiple Range test to compute the significance of items by each country.That way, it was anticipated to see how the responses are attributed to the cultural differences of the participants.Finding the F values to measure the ratio of mean regression in the present study was done to draw conclusions by interpreting the significant differences of collected data.With the following use of the Duncan Multiple Range test, significant differences among the sample mean scores of four participating countries were determined.
When analyzing the data for the thirty-two items, a factor analysis was conducted again to confirm that the items were grouped as a reliable method to measure the intended cultural differences.This was done to verify that such scores could be explained as indicators of the underlying constructs.Five-point scale Likert format was used to collect data on the participants' in-group and out-group behavioral perspectives.

Results
Table 2 displays the findings of the first section of the questionnaire.Although not all eight suggested groups yielded significant differences among countries, the results substantiate the presence of cultural differences in in-group and out-group categorizations.Family, religious members, and social members are the three groups that did not present significant differences among all four participating countries.The five remaining groups presented significant differences among some countries.Among eight suggested groups, Japanese respondents presented the highest selection of percentages in six groups as their ingroups followed by South Korean participants with two remaining groups, while Canadian and American groups did not choose any top percentage selections.Japanese people's selection of co-workers as in-group members was significantly higher than that of the Korean group F (1, 705) = 2.38; p < 0.05, Canadian participants F (1, 608) = 3.12; p < 0.001, and American participants F (1, 659) = 3.24; p < 0.001.However, no significant value differences were found among South Korean, Canadian, and American respondents for this item.The percentage of perceiving co-workers as in-group members for South Korean was not significantly higher than that of Canadian group F (1, 693) = 6.95; p = 0.05, and American group F (1, 744) = 8.14; p = 0.05 in the present study.Also, no significant selection differences were found between Canadian participants and American respondents F (1, 659) = 7.24; p = 0.05.Because Japanese organizations are fundamentally social organization (Grainger & Miyamoto, 2003) in which harmony of the group members (Beamer & Varner, 2001) and closeness and friendship between employees is considered very important (Wolf, 2013), most institutes and organizations view employees as one family unit (Wolf, 2013), where communication among members is often used as a tool of harmony establishing or maintaining (Okabe, 1983).This unique approach is different from the United States where maximizing the profits (Husted & Salazar, 2006) and performance goals (Shinkle, 2012) are more important, and possibly from Canada that has heavy cultural influence from the United States (Martin & Nakayama, 2011).

South
An interesting finding arose from the participants' perceptual categorization of regional members.Between South Korean respondents and Japanese participants, there were no significantly different statistical presentations of this group as an in-group F (1, 705) = 8.69; p = 0.235.Also, Canadian group and American participants did not present significant difference perception toward regional members F (1, 647) = 9.84; p = 0.229.Twenty-nine percent of South Korean respondents view regional members as their in-group and behave accordingly.This selection was significantly higher than that of Canadian respondents F (1, 693) = 9.27; df = 1; p < 0.05 and that of American participants F (1, 744) = 9.83; df = 1; p < 0.05.Twenty-eight percent of Japanese participants selected regional members as their in-group which was significantly higher than that of Canadian respondents F (1, 608) = 7.94; p < 0.05 and that of American participants F (1, 659) = 8.92; df = 1; p < 0.05.Because Korean (Kang & Lee, 2007;Yee, 2000) and Japanese (Yum, 1987) people tend to keep particularly close ties with regional members, the results of this in-group selection are consistent with previous studies (e.g., Chang & Chang, 1994).
Besides relatives and friends with higher selecting percentage as in-group by South Korean and Japanese participants compared with Canadian and American respondents, school alumnus is another group that yielded significant differences among four countries.Again, no significant selection differences were found between South Korean respondents and Japanese participants F (1, 705) = 7.38; p > 0.05, and between Canadian participants and American group F (1, 647) = 8.84; p > 0.05.South Korean participants' selection of school alumnus as in-group was significantly higher than that of Canadian respondents F (1, 693) = 7.69; df = 1; p < 0.05 and that of American group in the present study F (1, 744) = 10.22; df = 1; p < 0.05.Close to a half of all Japanese participants (49%) perceived school alumnus as in-group members which was significantly higher than the selection presented by Canadian participants F (1, 608) = 9.82; p < 0.05 and by American participants F (1, 659) = 10.12;p < 0.05.South Korean (Lee, 2016) and Japanese (Iwama, 1989) youths spend more time at school than youths in most other countries.For example, schools in Japan are in session from 240 to 250 days while American schools offer about 180 school days a year (Iwama, 1989).Friendships developed at school would be important regardless of cultures, but American friendship is known to show tendency to decline in satisfaction, commitment, and investments as friends grow older (Oswald & Clark, 2003).During school, however, intimacy and developing friendship are the most important positive features for Japanese adolescence (Schneider, 2000) and the ties of friendship are extremely strong with fellow alumni in South Korea that will continue as closest friends through their adult life (Chang & Chang, 1994).Previous research supports the data from the present study that close friendships developed in youths is recognized as a means of maintaining harmony and social cohesion in Japan (Schneider, Lee & Alvarez-Valdivia, 2012) and with high trust level among alumni in South Korea (Chang & Chang, 1994).
Table 3 displays actual behavioral and perceptual patterns of participants in the present study.Japanese participants showed highest selection percentages with six groups among eight suggested groups, and also showed behavioral patterns that are more positively ingroup-oriented in both small and large cities than three other countries in the present study.The participants from Ichihara, Japan presented the highest total score for 32 items (N = 147; 119.34/160;SD = 13.56) and the respondents from Houston presented the lowest total score for the same items (N = 176; 104.78/160;SD = 10.27)F (1, 323) = 16.67;p = < 0.005.That is, positive and more favorable behavioral and perceptual presentations toward in-group members were significantly higher for the small city Japanese group than the respondents from a large city in the United States.For example, in the problem-solving approaches section of the questionnaire, one item asked participants: "When solving problems, my group members' needs, interests, and goals are more important than needs, interests, and goals of my own."For this item, the mean score for small city Japanese participants (M = 3.77; SD = 0.84) was higher than any other groups in the present study followed by respondents from large city Japanese respondents (M = 3.68; SD = 0.79).These behavioral differences might indicate various degrees of interpersonal perceptions and interactions among people of different cultures.American respondents from a small city (Cincinnati) showed the lowest score for this item (M = 2.74; SD = 0.76).With a great emphasis on group harmony in Japanese society in general (Wolf, 2013) when compared with Western societies in which individualistic concerns preceding the groups' interests (Vandello & Cohen, 1999), this finding shows existing perceptual differences among cultures.* City with less than 300,000 people in the population ** City with more than 1,000,000 people in the population Among male respondents in all eight groups, Japanese respondents from a small city (Ichihara) presented highest overall scores (N = 94; 114.77/160;SD = 12.78) followed closely by Korean male respondents from a small city (Gumi; N = 103; 113.83/160;SD = 13.58) with no significant differences between two groups F (1, 197) = 6.32; p = 0.05.The male group from Edmonton, Canada scored the lowest (N = 99; 106.39/160;SD = 14.24) followed closely by the male group from Houston (N = 121; 106.44/160;SD = 15.48).No significant differences were presented between these two groups, either F (1, 220) = 6.92; p = 0.05.However, overall mean score variation yielded significant differences between the male group from a small city in Japan and the male group from a large city in Canada F (1, 193) = 9.83; p < 0.05.One item with the largest score difference between these two groups came from the material aspects with in-group members.This item asked respondents: "If a group member were to ask me for financial help, even without guaranteeing to pay me back, I would still help within my means."For this item, the Japanese group from a small city scored the highest (M = 3.59; SD = 0.88) followed by South Korean group from a large city (Daegu; M = 3.38; SD = 0.84) and South Korean group from a small city (Gumi; M = 3.32; SD = 0.91).Male group from a large city in Canada scored the lowest for this item (Edmonton; M = 2.35; SD = 0.76) followed by male group from a large city (Houston; M = 2.42; SD = 0.68) and by male group from a small city (Cincinnati; M = 2.51; SD = 0.66) in the United States.
Among female respondents in all four countries, Japanese respondents from a small city (Ichihara) presented highest overall scores (N = 53; 120.92/160;SD = 11.54) which was the highest among all groups across gender.This score was significantly higher than the female group from a small city (Cincinnati) in the United States F (1, 145) = 8.62; p < 0.005., of which the participants' overall score was the lowest among all groups including male groups (N = 92; 102.66; SD = 11.56).The female group with the second highest overall score was respondents from a large city (Sapporo) in Japan (N = 60; 119.88/160;SD = 11.83) and the female group with the second lowest overall score was participants from a large city (Edmonton) in Canada (N = 55;110.29/160;SD = 13.34).The item with the largest scoring difference came from an interpersonal issue statement that asked respondents: "If I had a job, I would think that maintaining good relationships with my co-workers would be more important than an increase in salary."For this item, the Japanese female group from a small city presented the highest average score (M = 3.74; SD = 0.79) followed by the Japanese female group from a large city (M = 3.65; SD = 0.77) and South Korean female group from a small city (M = 3.46; SD = 0.80).The American female group from a small city showed the lowest score for this item (M = 2.49; SD = 0.67) followed by Canadian female group from a large city (M = 2.74; SD = 0.63).Although the changing role of women in modern Japan has been described as a dramatic phenomenon (Jandt, 2013) with the increasing availability of meaningful career and employment (Holloway, 2010), Japanese children growing up with the parents' emphasis on consideration for others and sociability (Power, Kobayashi-Winata & Kelley, 1992) might be the cause of maintaining cultural difference with females in western society where, for example, U.S. parents emphasize independence and individual achievement during child development period (Befu, 1986;Yi, 2001).
After modification of the items from the pilot study, Cronbach's alpha coefficients of reliability for the main study were α = 0.89 for cooperative and competitive behaviors items with inter-item correlations ranged from -0.01 to 0.42, and α = 0.87 for interpersonal issues items, with inter-item correlations ranged from -0.01 to 0.38.For the remaining two groups, α = 0.91 for problem-solving approaches with interitem correlations ranged from -0.03 to 0.30, and α = 0.92 for material aspects items with inter-item correlations ranged from -0.01 to 0.41.Cronbach's alpha coefficients for the total 32 items in the second part of the questionnaire were α = 0.90.In addition to commercial and educational exchanges, 39,000 U.S. troops in Japan and 23,000 American soldiers in South Korea (Price, 2017) are protecting tight relationships among alliances.

Discussion and Implications
Most intercultural and cross-cultural studies, if not all, explore cultural phenomena to suggest better communication skills or more accurate perceptual guidelines about subjected cultures.More recent studies have argued that cultural differences among countries might be less pronounced (e.g., McCann, Honeycutt, & Keaton, 2010) than previously, or even, currently presumed.Culture is not a stagnant but a fluid phenomenon (Matsumoto, Kudoh, & Takeuchi, 1996) and actually, "many cultures are in a state of great transition" (Neuliep, 2015, p. 49).However, the results of the present study suggest that deeply imbedded cultural variations still exist and cannot be ignored when members of these cultures are interacting.In addition, the results indicate significant statistical differences between Asian and non-Asians in selections of in-group members and their behavioral patterns.If the gaps of cultural discrepancies are narrower in the modern world with the influences of transportations, media, and more frequent engagements among countries, it is even more crucial to accurately examine the cultural variabilities to offer more precise guidelines to people who are exposed to different cultures.
All participants in the present study were private organizational employees and this group did not constitute a representative sample of the populations involved.Thus, the findings could only help us understand the perspectives of organizational members in four countries regarding in-group and outgroup perspectives.To draw more generalizable findings of this topic, future studies should consider including people with different professions, age, education, socio-economic status, and geographical locations.Also, to measure different in-group and outgroup categorizations and behavioral patterns, this study relied exclusively on participants' self-reports.Although surveys tend to yield strong reliability (Babbie, 2001), there is no opportunity to evaluate the actual behavior of the respondents (Monette, Sullivan, & DeJong, 1986) and thus, there is no assurance that the attitudes and behavior of the participants described in the questionnaire are an accurate reflection of their action.To improve the validity and applicability of the findings, future studies should consider using other methods such as experimental approaches that "allow observation of actual behavior rather than relying upon people's statements" (Persell, 1984, p. 42).
Despite its limitations with participant sampling and data collection, the present study delivers two main messages.First, cultural differences among countries exist in various degrees and contexts.By including more groups such as people with military affiliations or political orientations, this study would provide more pragmatic data to various social groups.Second, with changing aspects of cultures, proper understanding and preparations without underestimating the values, status, and degrees of cultural differences are necessary.If people are well equipped with cultural sensitivity, encounters among members of different cultures could produce more promising outcomes than what people might anticipate.
Positively or negatively, every person belongs to various familial and social groups and often establishes different degrees of attachment with members of these groups.People shape and develop perceptual and interactional group behavior not only by personal characteristics but also by cultural influences.The results of the present study support the existence of in-group and out-group distinctions among cultures in terms of categorizations and behavioral patterns toward members of various groups.It would be a mistake, however, to engage in any cultural encounters by assuming all noticeable differences are due to cultural variabilities.Every culture in both individualistic and collectivistic category is different, as is every individual within any given cultural boundary.With more specific factors and accurate measurement in future studies, more objective recommendations and guidelines of in-group and outgroup behaviors among people of different cultures could be made.

*
Percentage calculated of the participants who claim to adhere to a religion.** Number of the respondents who attend religious establishments.

Table 1 :
Number of Respondents by Countries, Organizations, and Gender

Table 3
Behavioral Characteristics toward In-group and Out-group Members by Nation and Gender (ICEF Monitor, 2016)and encounters among South Korea, Japan, Canada, and the United States have grown steadily with no signs of fading in the foreseeable future.For example, billions of dollars in goods, services, and resources were exchanged between the United States and South Korea (United States Census Bureau, 2017a) and the United States and Japan (United States Census Bureau, 2017b).Also, trade and economic relations between Canada and Japan "are bolstered by common values and mutual positive perceptions" (Government of Canada, 2015a, p. 1) and Canada and South Korea continue twoway trade reaching nearly $11.9 billion in 2015 (Government of Canada, 2015b).Moreover, 78,489 South Korean students and 24,452 Japanese students are studying in the United States (U. S. Immigration and Customs Enforcement, 2016) and another 19,760 South Korean and 7,105 Japanese students are studying in Canada(ICEF Monitor, 2016), while thousands of American and Canadian students attend universities in South Korea (Asia Matters for America, 2011) and Japan (Japan Student Service Organization, 2015).