Nationality and information behavior: Comparing Koreans and Japanese

Authors


Abstract

This panel discusses “cultural factors in information behavior.” Our special concern is understanding the relation between nationality and information behavior. While many studies have been conducted to investigate the relations between national/cultural factors and information behavior, there is little research regarding the “meaning of nationality in information behavior.” We assume that information behavior would be considerably influenced by nationality, therefore, information systems in a nation, to be successful, should be based on its national traits. Thus the purposes of this panel are to raise a research regarding the “meaning of nationality in information behavior,” and to test the positiveness of the proposed relation theoretically. To the end, at this preliminary stage we are going to explore the relation through comparing Koreans and Japanese. In details, we are going to (1) identify the differences in personal traits between Koreans and Japanese, (2) explore the relation between nationality and information behavior, (3) investigate the relation between personal traits and information behavior, and finally (4) figure out the relative significance in information behavior between nationality and personality.

Introduction (By Jae-Whoan Lee)

In this panel, we discuss “cultural factors in information behavior.” Our special concern is understanding the relation between nationality and information behavior. While many studies have been conducted to investigate the relations between national/cultural factors and information behavior, there is little research regarding the “meaning of nationality in information behavior.” We assume that information behavior would be considerably influenced by nationality, therefore, information systems in a nation, to be successful, should be based on its national traits. For example, Korean information systems should reflect Korean traits in its system design, and so should Japanese information systems.

Thus the purposes of this panel are to raise a research regarding the “meaning of nationality in information behavior,” and to test the positiveness of the proposed relation theoretically. To the end, at this preliminary stage we are going to explore the relation through comparing Koreans and Japanese. In details, we are going to (1) identify the differences in personal traits between Koreans and Japanese, (2) explore the relation between nationality and information behavior, (3) investigate the relation between personal traits and information behavior, and finally (4) figure out the relative significance in information behavior between nationality and personality.

Literature review

A literature survey was conducted two research fields: group psychology and information behavior. Special concerns were on Cultural psychology (Shweder, 1991; Kitayama Markus, 1994; Cole, 1996; Harre, 1998; Miller, 1999, Choi, 2000) and Personality in information searching (Kuhlthau, 1991, 1993 Nahl Tenopir, 1996, Wang et al, 2000, Wilson et al., 2000, Nahl, 2004, 2005, Kim, 2006). A key concept attracted our attention in the process of survey was ‘affective’ in contrast to ‘cognitive.’ Affectiveness was not only a critical factor distinguishing nationality, but it was also a crucial variable influencing information behavior.

Information scientists, such as Kuhlthau, Wilson, Chatman has given their attention to information seekers' affective characteristics, differently from many researchers focusing on cognitive sides only. Kuhlthau focused on that people went through different stages of actions and experienced various emotions such as anxiety and frustration during their search process (Kuhlthau, 1991). Wilson also proposed a new formulation of information searching in formally recognizing that affective goal stated imparts directionally to problem-solving steps (Wilson et al, 2000). Chatman also showed that affective states such as “alienation or disinterest” had a strong influence on information behavior in everyday contexts (Chatman, 2000).

Recently Nahl showed that ‘affective coping skills’ consisting of self-efficacy and optimism had a positive impact on the performance in information tasks and that the higher affective skills could compensate for lower cognitive skills (Nahl, 2004, 2005). Since the “affective load” model by Nahl, several researchers have been joining to the “affective circle.” Among them some scholars have attempted to show how users' emotion control interact and influence the Web search behavior and performance through Web searching experiments (Kim Allen, 2002; Kim, 2006). Thus, the concerns of information behavior researchers have been expanded to the investigation of affective variables such as anxiety or stress, and accordingly, to the inquiries about the effects of personality on information behavior.

However it is very difficult to locate researches focusing on the relation between common personality and information behavior from a group perspective, while many researches focus on the relation between personal abilities and information behavior from an individual perspective. Even Nahl, who insisted that a sociocultural approach was absolutely crucial in information behavior research, has not attempted to investigate the relation between group traits and information behavior from a national perspective. However it has become meaningless to apply a general model for developing information systems in a nation, whose people has quite indigenous information behavior as well as unique national traits. And that explains why this study is focusing on the relation between national traits and information behavior.

Research Design

(1) Participants in this study consist of two national groups: Koreans and Japanese. In terms of their history, culture, and populations, Korean and Japan are much more homogeneous than most other nations. All the participants were drawn from undergraduate students enrolled in top-ranked LIS programs of two national universities: Pusan National University (PNU) of Korea and Tsukuba National University (TNU) of Japan. These two groups of LIS students were selected to minimize the influences by control variables. For reference, these two universities are very similar in their institutional identity and academic excellence. Among 341 students, 158 students were from PNU and, 183 students from TNU. Regarding their backgrounds, 32% of them were male while the others were female, and 53.6% were the lower grades while the others were the upper grades.

(2) Variables and Indicators: nationality, personality, and information behavior are three key concepts in this study. First, nationality consists of various personal traits which are common among the people of a society, and is commonly represented by the “major type of personal traits found in a targeted society.” Thus, in order to identify unique personal traits representing nationality, we must figure out “major type of personal traits found in each of two targeted groups”. Second, to figure out personality, this study employs a method to identify personal traits related to problem-solving or decision-making styles. In the concrete, we consulted and scrutinized models by Heppner and Harren, and decided to utilize Harren's decision-making styles (Fischer Corcoran, 1987). However, instead of attempting to calculate the component ratios of personality styles on the basis of Harren's model, this study focused on finding out the degrees of affectiveness and dependency found in the targeted national groups. Finally, information behavior was was measured by the following 15 indicators: □ strength of information needs, □ contents of information needs, □ activeness of information seeking, □ barriers against information seeking, □ preference in information sources, □ reliability in information sources, □ relative preference between official and unofficial sources, and □ relative preference between mass and private sources, □ value of library as information source, □ use of library catalog, □ utility of library catalog, □ use of librarian, □ use of advanced search functions while they search Web sites, □ use of multiple search terms, while they search Web sites □ preference of re-search while they search Web sites.

(3) Data Collection and Analysis: surveys were conducted for data collection. The questionnaire consisted of two parts. The first part was to measure the degrees of affectiveness and dependency. 20 questions were utilized from Harren's model, however the original 10 questions for affective style were modified for better application to Asian people, while the 10 questions for dependent style used without modification. YES/NO answer was given, and 1 point was allocated for each YES to score the value of each indicator. And the second part consisted of 20 questions to measure the 15 indicators on information behavior. Answers to these questions were rated on a 4-point scale or ranked by preference. The data collected were coded and analyzed using SPSS 14.0. A simple frequency-distribution test was conducted to figure out descriptive statistics on national traits and information behavioral characteristics of each national group. In addition, chi-square tests were conducted to examine the relation between nationality (in terms of citizenship) and information behavior, and the relation between personal traits and information behavior. Also conducted was two-way ANOVA test to investigate the relative significance of nationality and personality on information behavior.

Differences in Personal traits between Koreans and Japanese (By Eun-Ju Lee, You-ra Youn, Ichiro Ohba, and Jae-Whoan Lee)

In this chapter, we discuss major differences in personal traits between 158 Korean students and 183 Japanese students. As mentioned earlier, the personal traits specially investigated in this research were: □ affective vs. cognitive, and □ dependent vs. independent.

(1) Comparison between self-images and test-results: the results showed that there were distinctive differences between the results by self-image test and those by Harren's model test. First of all, in Self-image test the Korean students had a tendency to regard themselves to be more affective than Japanese students (63.3% vs.55.7%) but less dependent (51.3% vs.62.3%). However in Harren's model test, the results were the exact reverse. In affectiveness, both national groups turned out to be cognitive rather than affective, while they judged themselves to be affective rather than cognitive. Between groups, Japanese were much less affective than their Korean counterpart. In dependency, both national groups turned out to be independent rather than dependent, while they judged themselves to be dependent rather than independent. Between groups, Japanese were much less dependent than Korean counterpart (Table 1.1).

(2) Gender differences in personal traits: the results showed, in the total sample group, male students were more affective and more dependent than female students. In the Korean group, male students were more affective and more dependent than female students. In the Japanese group, male students were more affective than female students, but there was no distinctive sexual difference in dependency. In the concrete, the Japanese female group turned out to be the least dependent as well as the least affective among 4 subgroups. And the Korean male group proved to be the most dependent as well as the most affective (Table 1.2).

(3) Grade differences in personal traits: the results showed that in the total sample group, lower grade students were more affective and more dependent than upper grade students. While both national groups showed the same tendency, they were different at the ratio. In the Korean group there was a considerable gap between upper and lower groups. However, the Japanese group did not show a meaningful gap between upper and lower groups. In the concrete, the lower Korean group proved itself to be the most dependent as well as the most affective, while the upper Japanese group turned out to be the least dependent as well as the least affective among 4 subgroups (Table 1.3)

 

Table  . Appendix 1: Tables on Differences in Persornal Traits
original image

Nationality and Information Behavior (Jee-Hyun Rho and Jae-Whoan Lee)

Chi-square tests were carried out to find out statistical significances in the relations between nationality and information behavior. The test results showed that 12 out of 15 indicators representing information behavior turned out to have statistically meaningful relations with nationality at p □0.05. They were: strength of information needs(partially), activeness in information seeking(partially), barriers against information seeking, preference in information sources, reliability in information sources, relative preference between official and unofficial sources, relative preference between mass and private sources, value of library as information source, use of library catalog, utility of library catalog, use of advanced search functions, and use of multiple search terms. Then, let us examine closely the test results, particularly focusing on those 12 indicators representing information behavior, which proved themselves to be associated with nationality.

(1) Three situations were set up to investigate the difference in strength of information needs between Japanese and Koreans: namely, for everyday life, for decision making, and for problem solving. The test results showed that Japanese had stronger information needs than Koreans in all the three situations, and the values of χ2 and p showed that there was a statistically significant relation between nationality and strength of information needs (Table 3-1).

(2) Regarding contents of information needs, both Koreans and Japanese had a common characteristic that they preferred ‘information based on expert knowledge (EK)’ for everyday life, whereas they preferred ‘information based on practical experience (PE)’ for decision making and problem solving. However, The values of χ2 and p showed that there was no statistically significant relation between nationality and this indicator (Table 3-2).

(3) Regarding activeness of information seeking, Koreans proved themselves to be more active than Japanese in all the three situations. However for only everyday life among the three situations, nationality turned out to be partially associated with this indicator (Table 3-3).

(4) Both national groups answered the most critical barrier against information seeking was ‘lack of self-effort.’ And followed were ‘lack of search skills,’ ‘poor information environment,’ and ‘high search cost’ in the exact same order. However both groups were different in the relative weight of barriers. Koreans placed more weight on ‘lack of self-effort’ and ‘poor information environment’, while Japanese did more weight on ‘lack of search skills’ and ‘high search cost.’ The values of χ2 and p showed that there was a critical difference in barriers against information seeking between the two national groups (Table 3-4).

(5) In both national groups, internet Web site was chosen as the most preferred information source. And the other sources were ranked as the following order: family/intimate, book/journal, mass media, and expert/institute. However, the relative weight of sources were different from each other. In particular, internet site was more preferred among Koreans, while family/intimate and book/journal were more preferred among Japanese. The values of χ2 and p showed that there was a statistically significant relation between nationality and this indicator (Table 3-5).

(6) The results on reliability in information sources were in contrast to those on preference in information sources. In details, the most reliable source among Koreans was expert/institute, and it was followed by book/journal, internet site, family/intimate, and mass media. However, the most reliable source among Japanese was book/journal, and it was followed by expert/institute, family/intimate, internet site, and mass media. The values of χ2 and p showed that there was a statistically significant difference in reliability in information sources between Koreans and Japanese (Table 3–6).

(7) Regarding relative preference between official and unofficial sources, both national groups preferred official sources to unofficial sources. In the national difference, Japanese had higher preference for unofficial sources (40.4%) than Koreans (29.7%). Regarding relative preference between mass and private sources, Koreans preferred private sources, while Japanese preferred mass media. The values of χ2 and p showed that there were statistically significant differences in relative preference in information sources between Koreans and Japanese (Table 3-7).

(8) Regarding value of library as information source, the results showed that Koreans were more positive than Japanese, while both national groups had a positive attitude on value of library as information source. The values of χ2 and p showed that there was a statistically significant difference in value of library as information source between Koreans and Japanese (Table 3-8).

(9) library search behavior: First, while both Koreans and Japanese had negative attitudes on use of library catalog, Japanese were more negative than Koreans. The values of χ2 and p showed that there was a statistically significant difference in this indicator between Koreans and Japanese. Second, while both Koreans and Japanese had positive attitudes on utility of library catalog, Koreans turned out to be more positive than Japanese. The values of χ2 and p showed that there was a statistically significant difference in the recognition of utility of library catalog between Koreans and Japanese. Third, regarding use of librarian as a search tool, negative attitudes were commonly found in the majority of each national group. However, the values of χ2 and p also showed that the two groups had no significant difference in use of librarian as a search tool (Table 3-9).

(10) web search behavior: First, Koreans showed themselves to be relatively more active in use of advanced search functions than Japanese, and the values of χ2 and p showed that there was a statistically significant difference in this indicator between the two national groups. Second, Japanese proved themselves to be relatively more active in use of multiple search terms than Koreans, and the values of χ2and p showed that there was a statistically significant difference in this indicator between the two national groups. Third, 55.1% of Koreans preferred re-search while only 43.7% of Japanese preferred re-search. However the values of χ2and p were not enough to support a meaningful relation in preference of re-search between the two national groups (Table 3-10).

Personality and Information behavior (Yong-Wan Cho and Jae-Whoan Lee)

Chi-square tests were carried out to find out if there are statistically significant relations between personality (affectiveness dependency) and information behavior (15 indicators). The test results showed: first, 4 out of 15 indicators representing information behavior had statistically meaningful relations with affectiveness at p □0.05 (Table 3-0). Second, 7 out of 15 indicators had statistically meaningful relations with dependency at p □0.05 (Table 4-0). Thus, dependency proved itself to be more related with information behavior than affectiveness. Then, let us examine closely the relations between personality and information behavior, putting more focus on those indicators which were turned out to have statistically significant relations with affectiveness and/or dependency.

(1) dependency had a statistically significant relation with strength of information needs in all the three situations (everyday life, decision making, and problem solving), while affectiveness did not. The results of frequency test also showed that the more dependent, the stronger information needs in all the three situations. (Table 4-1).

(2) for only problem-solving among the three situations, both indicators(affectiveness and dependency) had statistically significant relations with contents of information needs. The frequency data showed that the less groups (LAg and LDg) had stronger needs on ‘expert knowledge,’ while the more groups (MAg and MDg) had stronger needs on ‘practical experience’ (Table 4-2).

(3) dependency has a statistically significant relation with preference in information sources, while affectiveness did not. The frequency data showed that there was a clear difference in preferring information sources between MDg and LDg. For example, MDg had more preference on family and intimate, while LDg did more preference on book/journal, and to less extent, on internet site (Table 4-5).

(4) dependency proved itself to have a statistically significant relation with reliability in information sources, while affectiveness did not. In the concrete, MDg was more reliable on Mass Media, while LDg placed more reliability on book/journal. Although it was not proved, there was a positive but small difference between MAg and LAg. For instance, MAg had more reliability on family/intimate, while LAg did more reliability on book/journal (Table 4-6).

(5) regarding relative preference between official and unofficial sources, both affectiveness and dependency proved themselves to have statistically significant relations with it. Between groups, the less groups (LAg and LDg) had more preference on official source, while the more groups (MAg and MDg) did more preference on unofficial source. Regarding relative preference between mass and private sources, only affectiveness proved itself to have a statistically significant relation. Between groups, LAg had more preference on Mass media, while MAg did on private source (Table 4-7).

(6) among the three indicators representing Web search behavior (use of advanced search functions, use of multiple search terms, preference of re-search), only use of multiple search terms turned out to have a statistically significant relation with both affectiveness and dependency (Table 4-10).

Relative Significance of Nationality and Personality (By Jae-Whoan Lee and Ichiro Ohba)

The final discussion concerns the relative significance of nationality and personality on information behavior. As discussed in previous sections, the results of chi-square tests showed that 12 out of 15 indicators representing information behavior had statistically meaningful relations with nationality at p □0.05. Also turned out was 4 out of 15 indicators representing information behavior had statistically significant relations with affectiveness at p □0.05, and 7 out of 15 indicators had statistically significant relations with dependency at p □0.05. Thus the results of chi-square test showed that nationality had more significant relations with information behavior than personality.

In this chapter, we re-examined their relative significance by using two-way ANOVA test. For the ANOVA test, the independent variables consisted of three: nationality, affectiveness, and dependency. And the dependent variable information behavior was represented by 21 indicators, mainly due to that the early mentioned three situations (i.e., everyday life, decision making, problem solving) were treated separately. The ANOVA test was carried out to assess only the relative significance of nationality, affectiveness, and dependency on information behavior. Therefore we referred to the F and P-values of three main effects and disregarded the values of interaction effects.

The results of ANOVA test showed that the F and P-values of (at least one) main effects were significant in 13 out of 21 results. Among them, it was 6 times when nationality proved itself to be the most significant main effect among three variables, and it was 6 times when dependency proved itself to be the most significant main effect. Finally, it was only 1 time when affectiveness proved itself to be the most significant main effect. When we considered all the 21 ANOVA results, nationality won 8 times, dependency won 7 times, and affectiveness won 6 times. Thus, the results of the two-way ANOVA test also showed that nationality was more significant than personality in relation with information behavior, and that between the two personal traits, dependency was more significant than affectiveness (Table 5-0).

Table  . Appendix 4: Table on Results of Two-way ANOVA for relative significance
original image

Conclusion (by Jae-Whoan Lee)

In this research, we attempted to figure out key factors influencing information behavior, with emphasis on national traits. We had two assumptions before investigation: first, there would be a significant relation between nationality and information behavior, and second, affectiveness and dependency would be key factors in national traits, therefore, both indicators could explain the differences in information behavior between Koreans and Japanese.

The research results showed that our assumptions were partially right. Nationality, in terms of citizenship, proved itself to be a key factor influencing information behavior. Dependency, one of the two variables developed to represent national traits, also proved itself to be a key factor influencing information behavior, However, affectiveness turned out to be not so influential as we assumed. Why would that be? What is wrong with affectiveness? The research findings on affectiveness raise some questions on the procedure and method that we employed for this research.

Above all, we have a question on whether the measurement tool for affectiveness was adequate. As discussed earlier, Harren's model was developed to distinguish personality into three types, and the test results based on Harren's model were quite different from those of self-image test. Thus, such result shows us that a more adequate tool to measure affectiveness needs to be developed, if we want to keep our eyes on affectiveness as one of the main variables related to information behavior. Also needed is a careful consideration on other factors of national traits, which might be more appropriate for figuring out the differences in information behavior among various national groups.

The second question is about whether using survey method only was proper enough to identify major differences in information behavior among national groups. As is generally known, information behavior researches employ a variety of methods for data collection. Besides the survey method, interview, observation, and experiment have been frequently used. Consequently, employing only the survey method seems to be not enough to investigate complex characteristics in information behavior. Therefore, methods for data collection also need to be made up for both weakness and limitations of this research.

The final question is whether the samples were fit for the objectives of this research. Needless to say, college students are too limited to represent the general public of a nation as well as common traits in nationality. Thus it would be inappropriate to apply the results of this research to the entire populations of the targeted two nations. Therefore a variety of national samples need to be investigated in the following research. Such research is sure to complement the validity of this study as well as to generalize the findings of this study.

Ancillary