The purpose of this section is to describe the methods and procedures used to test the research hypotheses and answer the research questions. This section is divided into four areas: (1) subjects, (b) procedures, (c) instrumentation, and (d) statistical analyses.

Subjects

　The subjects required in this study were randomly selected from international students studying at Kent State University in the Spring semester 1986. April 1986 data from the Kent State Office of International Student Affairs showed that in Fall 1985 there were 611 foreign students from sixtythree different countries (see Appendix A). Among these, 309 students were enrolled in undergraduate courses, 299 were enrolled in graduate courses, and 3 were enrolled in postgraduate nondegree programs. These foreign students were mainly from the Far East, Asia, Middle East, Africa, Europe, and North America. A few others were from South and Central America, Australia, and the Caribbean. These students, therefore, represented different parts of the world. For the purpose of this study, each international student was assigned a number and 200 potential subjects were selected randomly (without replacement) for this study. Of these, 149 international students agreed to participate in the study (see Table 1).

　Table 1 reports the proportion of students from seven major areas of the world. Fifteen subjects were from Africa, accounting for 10.1% of the total number of subjects, 23 were from Asia, accounting for 15.4%, 18 were from Europe, accounting for 12.1%, 78 were from the Far East, accounting for 52.5%, eight were from the Middle East, accounting for 5.4%, one was from North America, accounting for 0.7%, and four were from South/Central America and the Caribbean, accounting for 2.7%.

　Among these, 94 subjects were male and 55 were female. The average age of the subjects was 27.5 (ages ranged from 18 to 42). Six subjects had English as their native language, all others were from nonEnglish speaking countries. In addition, the average time subjects had stayed in the United States was 2 years and 9 months (this ranged from 3 months to 9 years). 

　

Table 1. Distribution of International Students

Procedures

　The researcher first telephoned the subjects and explained the purpose of the study. The researcher then hand delivered the questionnaires to the subjects upon agreement to participate in the study. These were collected upon completion. At the beginning of the questionnaire, subjects were asked to complete the Kent State Human Subjects consent form (see Appendix B). At the end of the questionnaire, each international student subject was asked to suggest two American friends, roommates, instructors, or other persons who know him or her well. This group, called "raters" in this study, were sent instrumentation booklets and were asked to rate the international student's level of intercultural communication competence. Booklets were returned to the researcher by mail. When both raters returned booklets to the researcher, the first booklet received was used in the study. Americans returned questionnaire booklets for 129 of the 149 international subjects.

Instrumentation

　As noted earlier, both groups (subjects and raters) were given instrument booklets. The American group was asked to rate the overall intercultural communication competence and communication message skills of the subjects. Subjects were asked to rate themselves on 15 different aspects of intercultural communication which comprised the four previouslydescribed dimensions of intercultural communicationPersonal Attributes, Communication Skills, Psychological Adaptation, and Cultural Awareness. All instruments used in this study are discussed separately below.

　Measurement of Intercultural Communication Competence

　The Intercultural Behavioral Assessment Indices (IBAI) developed by Ruben (1976) (see Appendix C) were utilized in this study to measure the general intercultural communication competence. The American raters were asked to rate the international students on seven items which measure seven different dimensions of intercultural communication competence: display of respect, interaction posture, orientation to knowledge,, empathy, role behavior, interaction management, and tolerance of ambiguity. Each index is comprised of a continuum containing either four or five levels of competence and descriptions of these levels. Raters were asked to indicate the position best describing the subject on the continuum. The scores obtained from the IBAI constituted the degree of intercultural communication competence of the international student subjects within an American environment. For analyses requiring a totaled score, the seven individual component scores were first converted to zscores and then summed. 

　Ruben (1976) found that the IBAI could be easily administered by untrained observers with efficiency and reliability. The IBAI had previously been tested on a group of trainees bound for overseas assignments and had been found reliable and useful. Three untrained staff members of a sevenday intercultural adaptation training program (who had quite different vocational and educational backgrounds and had little prior experience working together before the training program) were asked to observe the participants (consisting of adults and young adults) during meals, coffee or cocktails, and during formal training sessions, and to complete, without collaboration, the IBAI at the conclusion of the training program. The results indicated that interrater reliability on most indices was quite encouraging, but the author did not report the reliability coefficients (Ruben, 1976). Interrater correlations on the dimensions of display of respect, empathy, selforiented role behavior, and interaction management were reported to be significant at the .001 level. Interaction posture, orientation to knowledge, and tolerance of ambiguity were significant at the .05 level. As to other indices, task role behavior and relational role behavior yielded interrater correlations in the appropriate range, but the report indicated that these fell short of the .05 significance level. 

　A Qfactor analysis by Ruben (1976) revealed that there were three types of subjects. Type I individuals were considered to be successful in interpersonal interaction in the host culture. They had high rating scores on orientation to knowledge, tolerance of ambiguity, interaction management, and display of respect, but low rating scores on selforiented role behavior. Type II individuals were considered to be essentially a lowprofile group that exhibited a mixed behavioral pattern. This group had high rating scores on display of respect, tolerance of ambiguity, and empathy, but low rating scores on task roles behavior, selforiented roles behavior, and interaction management. Type III individuals were expected to encounter difficulty in intercultural interaction in the host culture. They had high rating scores on selforiented roles behavior, but low rating scores on orientation to knowledge, interaction management, relational group role behavior, empathy, tolerance for ambiguity, and interaction profile. 

　Because both Type I and Type II individuals did not show high rating scores on selforiented role behavior and taskoriented role behavior, these role behavior scales were not used in the present study. The dimensions used in this investigation included: display of respect, interaction profile, orientation to knowledge, empathy, relational group role behavior, interaction management, and tolerance for ambiguity. A general consideration about the scores obtained from the IBAI is that high degree of intercultural communication competence is indicated by high scores on each index.

　Measurement of Personal Attributes

　As explained in the previous literature review, Personal Attributes in the study of intercultural communication competence contains four major personal traits: selfdisclosure, selfawareness, selfconcept, and social relaxation. Although each of these traits accounts for part of Personal Attributes and can be measured separately, only selfdisclosure, selfawareness, and social relaxation were measured in this study. This decision was made to reduce the number of questionnaires subjects were asked to complete, and because reliable and valid selfdisclosure and selfawareness scales were available in abbreviated format. 

　The General Disclosiveness Scale (GDS) (see Appendix D) developed by Wheeless (1978) was used in this study to measure the subjects' general tendency to disclose to others, i.e., a trait of disclosiveness. The GDS is composed of 31 items which can be divided into five dimensions: amount of disclosure, consciously intended disclosure, honesty/accuracy of disclosure, positiveness/negativeness of disclosure, and depth/intimacy of disclosure. Wheeless reported the following reliability coefficients for these dimensions: intent, .65; amount, .82; positiveness, .90; depth, .78; and honesty, .84.

　Wheeless (1978) has found that the GDS could also predict some of the variables which are related to interpersonal solidarity. In addition, Wheeless and Grotz's (1975) earlier study leading to the development of the GDS showed that a selfreport, selfdisclosure measure about a specific target person could be used to measure intent and amount of disclosure; these were related to level of trust in the target person. 

　Subjects in the present study were asked to indicate the degree to which each statement reflects how they communicate with Americans by marking a 7point Likerttype continuum (ranging form "strongly agree" to "strongly disagree"). Average subscale scores were used in all data analyses. 

　The SelfConsciousness Scale (SCS) (see Appendix E), developed by Fenigstein, Scheier, and Buss (1975), was used in this study to measure subjects' selfawareness and social relaxation. The 23item SCS consists of three factors. The private selfconsciousness factor is concerned with attention to one's inner thoughts and feelings. The publicconsciousness factor is concerned with general awareness of the self as a social object, one that has an effect on others. And the socialanxiety factor is concerned with discomfort people experience in the presence of other. The social anxiety factor, then, was used as a measure of social relaxation. Public selfconsciousness was found to correlate moderately with private selfconsciousness and social anxiety, but the correlation between private selfconsciousness and social anxiety was negligible. The Fenigstein et al. study also showed that the factors were internally reliable (coefficient alphas): public selfconsciousness, .84; private selfconsciousness, .79; social anxiety, .73; and .80 for the total set of items. 

　Further study by Fenigstein (1974) has found that women who had high public selfconsciousness were more sensitive to rejection by a peer group. In addition, people who were high in private selfconsciousness were found to be more responsive to their transient affective state.

　Items on the SCS were originally scored on a scale ranging from 0 (extremely uncharacteristic) to 4 (extremely characteristic). In the present study, subjects were asked to indicate their degree of selfawareness by rating themselves on a Likerttype scale ranging from 1 (not at all like me) to 5 (exactly like me). Mean subscale ratings were used in data analysis. 

　Measurement of Psychological Adaptation

　Psychological Adaptation is typically associated with personal ability to handle situations such as frustration, stress, alienation, and ambiguity within a host culture. That is, the dimension of Psychological Adaptation indicates how individuals become acclimated to the socalled "social difficulties" caused by the host culture. In order to measure the ability to adapt psychologically or hand "social difficulties," Furnham and Bochner's (1982) Social Situations Questionnaire (SSQ) was used in the present study (see Appendix F). The questionnaire measures mainly the individual's ability to deal with ambiguous and stressful situations in a host culture.

　The SSQ consists of 40 statements about commonly occurring social situations identified as ambiguous and stressful for people living another culture. The 40 statements were adapted by Furnham and Bochner from a scale developed by Trower, Bryant, and Argyle (1978) and from interviews with over 50 foreign language students. All the statements have been identified as particularly difficult. 

　Research by Furnham and Bochner (1982) has shown that the SSQ is reliable and valid. One hundred and fifty students selected from approximately 400 foreign language students from 51 countries were divided into three experimental subgroups (according to the geographic region from which they originated): northern Europe, southern Europe, and the East. Another 50 British students served as the control group. The results showed that these three geographic subgroups differed significantly on the SSQ. The results also revealed that social difficulty was a positive function of cultural distance. In other words, the bigger the difference between the host culture and the sojourner's culture, the greater the social difficulty is. This study provided evidence about the construct validity of the SSQ as a measure of social difficulty. This study also demonstrated face validity in that foreign students experienced greater social difficulty than the host culture students. 

　In the present study, subjects were instructed to indicate how much difficulty they experienced in each of the SSQ situations in the United States. A 5point scale ranging from "extreme difficulty" to "no difficulty" was used in this questionnaire. As defined, "having difficulty" indicates that the subject feels anxious, uncomfortable, frightened, embarrassed or uneasy in the host culture (Furnham & Bochner 1982). For data analysis involving the ability to deal with stressful and ambiguous situations, the mean rating for the 40 items was utilized. 

　Measurement of Communication Skills

　As mentioned previously, Communication Skills are comprised of message skills, interaction management, social skills, and flexibility. Cegala's (1981) Interaction Involvement Scale (IIS) was used to measure interaction management (see Appendix G), and Wheeless and Duran's (1982; Duran, 1983) Communicative Adaptability Scale (CAS) was used to measure flexibility (see Appendix H). In addition, an adaptation of Rubin's (1985) Communication Competency OtherReport Scale was completed by the American rater group about their foreign friends/acquaintances to measure message skill (see Appendix I). 

　The Interaction Involvement Scale (IIS) (Cegala, 1981) consists of 18 items that are considered to be fundamental elements in the interpersonal communication process. Basically, these items were designed to describe social behaviors related to personal abilities of attentiveness and perceptiveness in interpersonal interaction. Cegala's research showed that interaction involvement contains three major factors: responsiveness, perceptiveness, and attentiveness (Cegala, 1981, 1984; Cegala, Savage, Brunner, & Conrad, 1982). Internal reliabilities for the three factors were .88, .86, and .87 (Cegala, 1981) and .89, .8, and .88 (Cegala et al., 1982) respectively. Moreover, these factors were found to be related to Wiemann's (1977) five dimensions of interpersonal communication competence (i.e., empathy, behavioral flexibility, interaction management, affiliation/support, and social relaxation), and were related to additional concepts such as extroversion, neuroticism, selfconsciousness, and communication apprehension (Cegala, et al., 1982). 

　In this study, subjects were asked to complete the scale using a 7point Likerttype continuum ranging from "very much like me" to "not at all like me." Also, subjects were instructed to complete the scale based on their experiences of interacting with Americans in the United States.

　The Communicative Adaptability Scale (CAC), developed by Wheeless and Duran (1982), consists of 20 5point Likerttype items that are used to measure two prominent dimensions of communicative competence: adaptability and rewarding impressions. Wheeless and Duran (1982) have reported reliability coefficients of .85 and .87 respectively for the two dimensions. Their study also showed that gender orientation was related to the adaptability and rewarding impressions dimensions of communicative competence. The findings confirmed three assumptions: (a) masculinity characteristics (i.e., instrumentality, control, and dominance) were highly correlated with adaptability, (b) femininity characteristics (i.e., expressiveness and sensitivity) were highly correlated with rewarding impressions dimension, and (c) androgynous individuals scored high on both dimensions.

　Subsequently, Duran (1983) reassessed the validity of the CAS and its dimensions. The results indicated that the CAS and its dimensions. The results indicated that the CAS might include four additional dimensions. For the adult samples, the CAS factored into social confirmation, articulation, social experience/composure, wit, and appropriate disclosure dimensions (the reliability coefficients ranged from .7 to .89). For the student samples, the CAS factored into social confirmation, social composure, articulation, social experience, wit, and appropriate disclosure dimensions (the reliability coefficients ranged from .71 to .85). The results also showed strong relationship and significant differences between these dimensions of communication adaptability and the two constructs of communication apprehension and selfesteem.

　In this study subjects were asked to complete the CAS using 7point Likerttype scales ranging from "strongly agree" to "strongly disagree." Because there were too few subjects in this study to perform a factor analysis, the original dimensions of adaptability and rewarding impressions were used and scores were obtained for each by averaging scores on the items in each index. 

　The Communication Competency OtherReport Scale (Rubin, 1985) was created as a possible alternative to the Communication Competency Assessment Instrument (Rubin, 1982b). These two scales were found to be virtually unrelated due to the different response modes (selfreport and rater assessed). However, a recent revision of the CCSR, the Communication competency OtherReport (CCOR) (alpha = .91) has been found to correlate significantly with organizational communication competence (L. Wilson, personally communication, May 1986). Thus, a revised version of this scale (see Appendix I) was given to the American group who was asked to complete the scale about their foreign acquaintances. Scores on this 19item index were summed and the total served as a measure of message sending and receiving skills.

　Measurement of Cultural Awareness

　Because measurement of Cultural Awareness focuses on social values, social customs, social norms, and social systems of the United States, the Test of American Culture (TOAC) (see Appendix J) created by Kenji Kitao (1981) was used in this study to measure Cultural Awareness. 

　The TOAC concerns 49 different areas of American culture including history, politics, superstition, religion, media, economy, family, customs, American character, food, manners, nonverbal behaviors, and so on. These areas obviously cover many aspects of Cultural Awareness about American culture.

　The TOAC originally contained of 100 questions. For each of 50 items, two questions were created. Questions 150 and 51100 are two parallel tests. All the questions are multiple choices with four options. In questions 150, subjects are asked to select the word or phrase most closely related to the given keywords or phrases. In questions 51100, subjects, given a sentence with a blank, are asked to choose the best word or phrase from a list of four.

　Kitao's (1981) study demonstrated that the TOAC was valid and reliable. Kitao used the TOAC with three different groups: Americans, Japanese students studying in the United States, and Japanese students studying in Japan. As predicted, the American group obtained the highest scores, the Japanese students studying in the United States obtained the second highest, and the Japanese students studying in Japan obtained the lowest.

　In order to test the reliability and to reduce the number of questions to be used in the present study, a pretest was performed. Since the two sets of TOAC questions are parallel, only questions 51100 were used. Fiftyone American students enrolled in basic speech courses at Kent State University were tested. The results showed that 25 out of 50 questions were answered correctly by over ninety percent of the students. Therefore, subjects in the present study were asked to complete these 25 questions, some of which were slightly reworded to counteract vagueness and obsolescence in the original test. Cultural awareness, then, was represented by the number of correct answers provided on the TOAC by the international subjects. 

　Thus, American group received a cover letter (see Appendix K) and completed the Intercultural Behavioral Assessment Indices and the Communication Competency OtherReport Scale. The international subjects completed a booklet containing the following:

　(1) A Human Subjects consent form and cover letter. 

　(2) A survey questionnaire form used to solicit subjects' demographic data (see Appendix L)

　(3) The General Disclosiveness Scale, used to measure one of the Personal Attributes, selfdisclosure. 

　(4) The SelfConsciousness Scale, used to measure the Personal Attributes of selfawareness and social relaxation.

　(5) The Social Situations Questionnaire, used to measure the dimension of Psychological Adaptation.

　(6) Two instruments used to measure the dimension of Communication Skills: The Interaction Involvement Scale and the Communicative Adaptability Scale.

　(7) The Test of American Culture, used to measure the dimension of Cultural Awareness of the United States.

　Each of the instruments was in the form indicated in its respective appendix. Naturally, the scoring instructions were not included in the booklets. 

Statistical Analyses

　Several statistical methods were employed in the analysis of the data generated by the research. First, Cronbach (1851) coefficient alphas were computed for the various scales to determine reliability. Second, to test Hypothesis 1, Pearson productmoment correlations were used to assess relationships among the 16 measures of Personal Attributes, Psychological Adaptation, Communication Skills and Cultural Awareness. Third, a canonical correlation analysis was performed on the data in order to test Hypothesis 2. Fourth, in order to explore the first research question, multiple regression was used in a stepwise procedure to determine the ability of the 16 measures of Personal Attributes, Psychological Adaptation, Communication Skills, and Cultural Awareness to predict the components of the IBAI. Lastly, to answer the second research question, a factor analysis was performed to investigate the soundness of the four proposed dimensions of intercultural communication competence.

Summary

　This chapter presents the methodology of the study. One hundred and forty nine international students from different parts of the world enrolled at Kent State University in Spring 1986 served as subjects in this study. 

　Eight instruments are discussed. Subjects completed six instruments: the General Disclosiveness Scale, the SelfConsciousness Scale, the Social Situations Questionnaire, the Interaction Involvement Scale, the Communicative Adaptability Scale, and the Test of American Culture. These instruments were completed by the international student subjects to measure dimensions of Personal Attributes, Communication Skills, Psychological Adaptation, and Cultural Awareness. The Intercultural Behavioral Assessment Indices and the Communication Competency OtherReport Scale were completed by 129 Americans who were identified as well acquainted with the subjects. 

　Finally, Pearson productmoment correlations, regression analysis, and canonical analysis were used in this study to interpret the data collected from the subjects and those who knew them. Moreover, factor analysis was used to test the original model of the dimensions of intercultural communication competence.