Abstract:

Three groups were compared on their perceptions of classroom instructional behaviors. Students from the US, and Eastern culture and a Western culture rated 31 instructional behaviors. US students differed fro Eastern students, and Western students differed from Easter students. The study showed measurable cultural difference existed in the perception of classroom instructional behaviors.

Full Text:

Copyright Jossey-Bass, Incorporated Fall 2001

Samples of university students from the United States, Eastern cultures, and Western cultures rated thirty-one classroom instructional behaviors for appropriateness in their culture. U.S. students had significantly higher appropriateness ratings than Eastern students on clarity of presentation, enthusiasm, classroom interaction, and spatial-behavioral communication behaviors. Western students exceeded Eastern students in all areas except spatial-behavioral communication behaviors; ratings of non-U.S. students were predicted by field of study and age.

In a recent survey of human resource development (HRD) executives from Fortune 500 companies and private companies with $500 million or more in sales, it was found that almost all predicted that their organizations would continue to have heavy involvement in international business. Most respondents expected a substantial increase in training related activities (Thomas, 1995). At the same time, educators and trainers from other countries are increasingly relying on the United States for instructional materials because of limited local availability (Chow, 1995). These factors indicate a trend toward instructional programs being developed, and often presented, by individuals who are not members of the target culture.

Instructional programs and techniques, however, may not be effective cross-culturally because every culture has its own specific expectations about appropriate instructor behaviors and delivery methods. Hofstede (1980, 1986), Laurent (1983), and Thornhill (1993), for example, found very different expectations about instructional behaviors and methods. Learners from societies with a cultural norm of strong uncertainty avoidance (European countries such as Greece, Belgium, France, and Spain) expect an instructor to be an expert and have an authoritative demeanor. In contrast, Marquardt (1993) found that in Saudi Arabia learners desire a friendly relationship with each other and with their instructors, whereas in Mexico learners tended to identify with the instructor and would be loyal to him or her as a person. In contrast, Marquardt (1995) found that in Japan instructors were expected to begin in a humble manner and show their respect for the learners by honoring them in some small way

These culturally different expectations prompted Kohls (1995) to point out that it is very common for organizations simply to transplant existing training programs into a foreign culture without considering learner expectations. He suggested that many organizations "have not noticed that the round pegs are not fitting very well into the square holes. Some, on the other hand, have noticed that the fit hasn't been any too good" (p. 51). Other researchers (Chow, 1995; Harris, 1984; Huang, 1996; Marquardt, 1993, 1995; Morical and Tsai, 1992; Schermerhorn, 1994; Thornhill, 1993) have found that when a training program is written or delivered by someone who is not a member of the population that will receive it, the program must be "acculturalized" or adapted to the target audiences and their cultures. Harris (1984) stated: "Training programs need to be modified to make them suitable for overseas presentation. The central point is that the adaptation has to be made to conform to the client's culture. Trainers must become aware of the unique values, attitudes, and practices that should shape the training content and methods. As in all helping activities, the trainer must start where the client population is, respecting their traditions and practices, with a healthy interest in learning their ways" (p. 50).

Hofstede defines culture as "the collective programming of the mind which distinguishes the members of one group or category of people from another" (1997, p. 5). This definition supports the concept that each culture will have idiosyncratic expectations of training situations. However, Baker, Mohammed, and Boyle (1994) suggest successful cross-cultural training requires consideration not only of the culture but also of the function of the participants, as well as the situation itself. This indicates that when considering cultural differences, training planners must consider what is communicated, how it is communicated, and to whom it will be communicated. Many studies have looked at instructional behaviors (for example, Hall, 1980; Hecht, Andersen, and Ribeau, 1989; LaFrance and Mavo, 1978; Malandro and Barker, 1983; Mehrabian, 1972; Neill, 1991; Wolfgang, 1977). Most authorities agree that the perceived appropriateness of instructional behaviors is culturally based and there are cross-cultural differences in interpreting these behaviors.

The realm of instructional behaviors includes gestures, facial expressions, eye contact and gaze, posture and movement, touching, dress, objects and artifacts, silence, space, time, and paralanguage (Samovar and Porter, 1991). They can express the instructors' attitudes toward the students and the students' attitude toward the instructor. Keith, Tarnatzky, and Pettigrew (1974), for example, analyzed instructional behaviors of student teachers. They discovered that the facial expressiveness of the teacher, through the use of smiling, exerted stronger effects in obtaining students' attention and responsiveness than the verbal statements made by the teacher.

This brief review of the relevant literature reveals the importance of considering the cultural expectations of the target audience when presenting a training program. However, Marquardt (1995), Morical and Tsai (1992), and Swing (1984) point out that it should be possible to make programs culturally relevant for different audiences without compromising the core concepts of the training program. In other words, the core content remains unchanged, only the way in which it is presented is tailored to the target population.

Other research has established that certain classroom behaviors are associated with higher ratings for instructors on evaluation questionnaires. For example, Murray (1983) found that lecturers in a North American university who regularly engaged in certain behaviors received relatively high ratings from students on end-of-course evaluations. Behaviors included those that enhanced the clarity of lectures and those that were associated with instructor enthusiasm.

Although it is clear that individuals from different cultures may have different expectations of instructional materials and techniques, no studies have been done to establish whether there are significant differences between the perceptions of U.S. and non-U.S. groups on the appropriateness of instructionally relevant classroom behaviors. If such differences could be documented, they could provide the basis for adapting classroom instructional delivery for international training. The primary objective of this study was to explore whether classroom instructional behaviors were perceived differently by U.S. students and students from non-U.S. cultural groups. In addition, the effects of background variables were examined. The authors addressed three research questions in the study:

RESEARCH QUESTION 1. What instructional behaviors are perceived by U.S. and non-U.S. cultural groups as appropriate or inappropriate?

RESEARCH QUESTION 2. What are the differences between U.S. students, Eastern-- culture students, and Western-culture students on their ratings of the appropriateness of the trainers instructional behaviors?

RESEARCH QUESTION 3. What is the relationship between the non-U.S. cultural groups' background characteristics on their ratings of the appropriateness of the trainer's instructional behaviors?

International students were chosen as the population for this study because they are native citizens of their countries and experts in their cultures. As "cultural ambassadors" (Du Bois, 1956, p. 44), international students provide "a living curriculum" (Shabahang, 1993, p. 1), and bring with them their home culture's indigenous perceptions about appropriate instructional behaviors. One of the most important practical benefits for choosing international students as the population for this study is that it is common for many different cultures to be represented at one location, such as a university classroom. Thus, the researcher has the opportunity to examine the perceptions of individuals from a variety of backgrounds who are each experiencing, as a group, a similar classroom instructional experience that resembles a U.S. designed and delivered training program.

Because of these research design benefits, it is common for cross-cultural studies to use student samples. Examples of studies with student samples include the Chinese Culture Connection (1987), the collective name adopted by a team of twenty-four researchers organized by Michael H. Bond, which administered the Chinese Value Survey to one hundred students from each of twenty-three countries; and Laurent (1983), who used British and French MBA students to study cultural diversity of Western conceptions of management. Vijver and Leung (1997) pointed out that "college students from different cultures are often compared, and it is assumed that their demographic characteristics are similar across cultures" (p. 30). In other words, it is possible to sample a diversity of characteristics like gender and age when using different cultural groups.

Method

The method is described as follows.

Instrumentation. In developing the instrument for this study a pilot study was conducted. The pilot study employed forty-six subjects, twenty-three U.S. students and twenty-three international students. Participants from both the international and U.S. groups were used to assure that the instrument would be readable and usable by the actual study groups. The total of forty-six subjects in the pilot exceeded the minimum sample size necessary to assure a relatively accurate representation of a variable (Hinkle, Wiersma, and Jurs, 1994).

The questionnaire used for this study consisted of two parts: a section collecting demographic data (Part I) and a survey of student opinions (Part II). Part II contained a list of instructional behaviors derived from past research and theory that were rated for appropriateness.

Instructional behaviors were organized into four scales: Clarity, Enthusiasm, Interaction, and Spatial-Behavioral Communication. Items were placed in the Clarity, Enthusiasm, and Interaction scales primarily based on the empirical findings of Murray (1983) and the findings of the pilot study completed by the authors of the present study This allowed items to be placed in categories based on their relationship to other items. For example, "speaking in a monotone" was in the Interaction scale because research showed that the behavior was related to others in the dimension of interaction. In other words, it was related to a cluster of behaviors that were potentially encouraging or inhibiting to classroom interaction among students and the instructor.

The Clarity scale included the following behaviors: using concrete examples, providing multiple examples, repeating difficult ideas, suggesting practical applications, showing concern for students, and using graphs and diagrams. The Enthusiasm scale included these behaviors: using humor, speaking expressively or emphatically, showing facial expression, moving about while lecturing, exhibiting energy and excitement, smiling or laughing, gesturing with hands and arms, avoiding eye contact with students, and speaking softly The Interaction scale included asking questions of individual students, addressing students by name, providing opportunity for participation, encouraging questions and comments, praising students for good ideas, presenting thought-provoking ideas, talking with students after class, and speaking in a monotone.

The fourth scale, Spatial-Behavioral Communication, included behaviors that previous authors identified as problematic for non-U.S. student populations, such as starting class on time, dressing casually, sitting on the desk, touching students when giving feedback, looking into students' eyes for as long as two to five seconds when listening to students, leaning over students' shoulders to check how they are doing, greeting students warmly, and using various tones, pitches, and volumes. These instructor behaviors generally encompassed nonverbal behaviors and were evidence for the level of formality in the classroom.

A total of thirty-one items in the survey used a Likert scale, with numerical weights assigned as follows: 1 = very inappropriate, 2 = inappropriate, 3 = neutral, 4 = appropriate, and 5 = very appropriate. For purposes of scale construction, some items were reverse weighted when it was known that they were descriptive of ineffective behaviors (for example, speaking in a monotone). Reverse weighting made it possible to combine these items with the other items in the scale.

Before answering the rating scale items, subjects read instructions that explained that they were to rate the appropriateness of classroom instructional behaviors of American instructors-trainers. Instructions on the instrument were as follows.

Purpose: This survey is to determine your opinions regarding the appropriateness of selected classroom instructional behaviors of American instructors-- trainers. Appropriate: "suitable for a particular person, condition, occasion, or place; proper, fitting." Assume that the American instructor-trainer is teaching in your country. Please evaluate the following behaviors as appropriate or not according to the social and behavioral norms of your culture-country. There are no right or wrong answers to these items. Please try your best to answer all the items in the questionnaire.

Internal consistency reliability indices, Cronbach's alpha, were calculated for each scale. Items were deleted if they were not correlated with other items on the scale. This was done to maximize the alpha coefficients. As a consequence, four items of the thirty-one items were dropped from use during the inferential statistical analyses that were performed in the study The reliability coefficients for the four scales were: Clarity (six items), .7909; Enthusiasm (eight items), .7592; Interaction (five items), .6945, and Spatial-Behavioral Communication (eight items), .6725. Item analysis involved processes for scale construction described by Nunnally and Bernstein (1994). The correlation between each item and the remaining items in a scale were calculated. In addition, there was a calculation of how much increase in reliability occurred if the items were deleted. Thus, it was possible to remove items that decreased the reliability of the scale. Sizes of coefficients were consistent with suggestions for the magnitudes of reliability coefficients given by Nunnally and Bernstein.

Design. Causal-comparative and correlation research procedures were selected because they were judged to be the most appropriate methods for the research questions. To address Research Question 1, mean scores and standard deviations were calculated for each item in the questionnaire and for the scales derived from the questionnaire: Clarity, Enthusiasm, Interaction, and Spatial-- Behavioral Communication. In addition to means and standard deviations, other descriptive statistics were calculated, specifically percentages of students marking each item option. The preceding analyses were done for these groups: U.S. students, Eastern-culture group, and Western-culture group.

Multivariate analysis of variance (MANOVA) was used to examine differences among groups on instructional behaviors. MANOVA was chosen because this procedure was developed to study the overall differences among groups on a set of dependent variables (Stevens, 1996). The independent variable was student group (U.S. students 1, Eastern-culture students 2, Western-- culture students 3). The dependent variables were scores on the four scales derived from the questionnaire: Clarity, Enthusiasm, Interaction, and Spatial-- Behavioral Communication.

Multiple regression analysis was selected to measure the relationship between background characteristics and instructional behaviors. Multiple regression is a very flexible statistical technique that can be used to study the degree to which both categorical and continuous variables predict a dependent variable (Stevens, 1996). Four regression models were created, one for each of the four scales from the questionnaire, which served as the dependent variables in the regressions. Five predictors were used: cultural background (West and East), gender, age, length of stay in United States, and field of study The latter variable had four categories: business, science, social studies, and health sciences. For purposes of the regression analysis, the categorical variable field of study was converted into three effect code variables using procedures described in Pedhazur (1982).

Participants. All international students at the University of Louisville, Kentucky (N = 463), were surveyed by mail or measured by administering surveys at classes in English as a Second Language (ESL). A total of 220 responses were returned, for a 48 percent response rate. As already noted, international students were categorized into a Western-culture group (students from Europe and Canada) and an Eastern-culture group (students from China, Japan, and other parts of Asia). This classification is based on the cultural maps derived from cultural studies and theories (Chinese Culture Connection, 1987; Hofstede, 1980, 1984, 1986,1997).

Because one of the main rationales for conducting this study was to understand how to modify U.S.-produced training materials and techniques for use in other cultures, a sample of U.S. students were surveyed in academic areas that were similar to those with international student enrollment so that their responses could be compared to those of the other two groups. The U.S. students (N = 102) were from the schools of education, business, engineering, and medicine. U.S. students were surveyed in classes.

For the U.S., Eastern-culture, and Western-culture groups, the percentages of males were 61 percent, 48 percent, and 52 percent, respectively The majority of subjects in the three groups were at the graduate level: 67 percent, 79 percent, and 81 percent for the U.S., Eastern, and Western groups, respectively. Average ages of respondents for the three groups were similar: 30, 28, and 27.5 years of age, respectively Participants came from a variety of disciplines. The most frequent fields of study of the students were natural sciences and engineering (35 percent), social sciences and education (30 percent), business (21 percent), and health sciences (15 percent).

Several steps were taken to increase the response rate. International students were surveyed by mail. Those who did not respond in four weeks were sent the questionnaire a second time and asked to respond. As another way of reaching international students, data were collected from students attending English as a Second Language (ESL) classes. This allowed students to be reached who were not responsive to a mailed instrument. U.S. students came from the same schools in the university as most international students (again, education, business, engineering, and medicine). Using students in classes boosted the cooperation rate.

Almost all the returned surveys were fully completed. For each of the demographic items, over 90 percent of subjects responded to each item. For the set of rating scale items, 97 percent of the surveys were complete.

Results

The primary purpose of this study was to explore whether classroom instructional behaviors were perceived differently by U.S. students and students from other cultural groups. In addition, the effects of background variables were examined. This study was guided by the three research questions listed earlier.

Results of Research Question 1. Key results related to Research Question 1 are shown in Table 1: means and standard deviations for the four scales that were used in the study For the three comparison groups, most mean scores were above 3.50 and several were above 4.00. For example, all three student groups had average scores for Clarity above 4.00. Because items in the scales were rated from very inappropriate (1) to very appropriate (5), the levels of the mean scores imply that instructor behaviors associated with clarity of presentation, enthusiasm, interaction, and spatial-behavioral communication behavior were viewed as being more appropriate than inappropriate.

Data from individual questionnaire items were similar to the results obtained from mean scores. Most of the thirty-one behaviors addressed by this study were found to be appropriate by all cultural comparison groups. However, a number of behaviors were rated as inappropriate by 15 percent or more of all of the subjects in the sample. The value 15 percent was chosen after inspection of the data; that is, the majority of the items were rated by subjects in the appropriate range. The number 15 percent was a natural dividing line that separated the items. The relatively inappropriate instructional behaviors included three behaviors that revealed lack of expressiveness: avoiding eye contact with students, speaking softly, and speaking in a monotone. The other four behaviors with a relatively large number of inappropriate ratings related to formality of instructor behavior or to student-instructor physical distance: dressing casually (for example, in blue jeans), sitting on the instructor's desk, touching a student when giving feedback, and leaning over students' shoulders to check their work.

Results of Research Question 2. A MANOVA was performed with student group (three levels) as the independent variable and average scores on the four rating scales as the dependent variables. This was statistically significant. Pillars trace statistic = .270, F(8, 622) = 12.12, p < .0005. Pillars trace statistic is a commonly used criterion for testing the null hypothesis in MANOVA (Stevens, 1996). It allows the researcher to state whether there were significant differences among the comparison groups on the combination of all dependent variables. Technically, it is the sum of the eigenvalues obtained from a matrix obtained by multiplying the between-groups sum of squares and cross-- products matrix by the inverse of the total sum of squares and cross-products matrix. The sum of the eigenvalues (called a trace) is converted into an F ratio for the purpose of statistical significance testing.

Separate analyses of variance (ANOVA) for each dependent variable resulted in significant differences among groups for each variable: Clarity, F(2, 313) = 40.32, p < .0005; Enthusiasm, F(2, 313) = 32.21, p < .0005; Interaction, F(2, 313) = 13.73, p < .0005; Spatial-Behavioral Communication, F(2, 313) = 5.24, p < .0005. Each ANOVA was followed by a Tukey multiple comparison procedure to determine significant differences (p = .05 for a set of comparisons) among the three groups.

Table 2 shows mean scores for those groups that were different in the Tukey comparisons. The table reveals that U.S. students had higher mean appropriateness scores for all four scales. For the Enthusiasm scale, U.S. students exceeded both Eastern and Western students. Furthermore, the Western-- culture group exceeded the Eastern-culture group on all scales except Spatial-Behavioral Communication. To put it another way, the Eastern-culture group always had the lowest appropriateness score for the four scales measuring instructional behavior.

The authors followed suggestions about appropriate statistical practices (for example, Stevens, 1996) by calculating effect size measures for each comparison. The eta square statistic was used for this procedure. The eta square statistic is a measure of the magnitude of the relationship between a categorical independent variable and a continuous dependent variable. It is used as a way of determining how strong a relationship there is between the two variables: it provides a procedure for determining whether a significant F ratio is indicative of a weak or strong relationship. Eta square is calculated from the obtained F ratio and its accompanying degrees of freedom. The possible range of this statistic is 0 (weakest possible relationship) to 1.00 (strongest possible relationship). For the four dependent variables, eta square statistics were Clarity, .21 (large); Enthusiasm, .17 (large); Interaction, .08 (moderate); and Spatial-Behavioral Communication, .03 (small to moderate).

Results of Research Question 3. Multiple regression was used to determine the degree of relationship between five predictor variables and four scales, each a dependent variable. The regression analyses did not involve U.S. students because the purpose of the analyses was to determine whether background variables could predict the ratings of international students. The predictor variable cultural background was coded 1 for the Eastern-culture group and 0 for the Western-culture group. Gender was coded 1 for male and 2 for female. Both age and length of stay were continuous variables and were measured in number of years and number of months, respectively The variable field of study had four categories, as noted earlier. For the regression analyses, the categorical variable field of study was converted into three effect code variables using procedures described by Pedhazur (1982).

The same procedure was used for each regression analysis. The variables cultural background, gender, age, and length of stay were entered using the stepwise method of entry of variables (Pedhazur, 1982). Significant predictors (p < .05) were entered into the regression equation one step at a time. If, in a given step, a variable that had been previously entered was no longer significant (at p < . 10) it was removed from the equation. Following the stepwise entry, the three effect code variables were entered into the equation in a single block. This was consistent with the suggestion of Pedhazur (1982), who stated that because all effect codes are needed to represent a categorical variable, all effect codes should be entered at a single step. If the block representing field of study accounted for a significant increment in variance (p < .05) it was retained in the equation.

For the scale Clarity, the variables cultural background and field of study were significant predictors (p < .05). Consistent with the MANOVA, the regression analysis revealed that Eastern-culture students had lower average scores for Clarity than Western-culture students. Regarding the variable field of study, the three effect codes yielded regression coefficients that were tested for significance. It was found that the coefficient associated with the effect code for business was significantly lower than the general mean for all fields of study The total amount of variance accounted for by cultural background and field of study can be summarized by adjusted R squared, the proportion of variance in the dependent variable accounted for by the predictors. For the Clarity scale, adjusted R squared was. 161.

For the scale Enthusiasm, the variables cultural background and age were significant predictors (p < .05). The regression analysis revealed that Eastern-culture students had lower average scores for Enthusiasm than Western-culture students. The variable age had a negative regression coefficient, meaning that the younger the student, the higher his or her score for Enthusiasm. The adjusted R squared for cultural background and age was .082. For the scale Interaction, the variable cultural background was the only significant predictor (p < .05). The regression analysis revealed that Eastern-- culture students had lower average scores for Interaction than Western-culture students; the adjusted R squared was .043. For the Spatial-Behavioral Communication scale, the variable age was the only significant predictor (p < .05). The variable age had a negative regression coefficient, meaning the younger the student, the higher his or her score for Spatial-Behavioral Communication; the adjusted R squared was .037.

Table 3 shows a summary of the regression analyses. The table shows how much variance was accounted for by each statistically significant predictor, and the cumulative total proportion of variance, adjusted R squared. Results of the regression analyses were consistent with the results of the MANOVA, which had revealed that the Eastern-culture group had lower average scores than the Western-culture group for the scales Clarity, Enthusiasm, and Interaction. However, the regression results produced new information. They showed that gender and months in the United States were not significant predictors of any of the scale scores. The variable age did account for significant variance in the scales Enthusiasm and Spatial-Behavioral Communication; age was inversely related to scores on the scales. This meant that younger students rated enthusiastic instructional behaviors and behaviors that encouraged classroom interaction as being more appropriate than older students. The variable field of study had a significant relationship to the variable Clarity. Students in business were relatively lower than the overall mean for business, science, social science, and health science.

In the multiple regression analyses, the amounts of variance in the dependent variables accounted for by the predictor variables were not large. The cumulative adjusted R squared values for the four dependent variables ranged from .037 to .161. If there are variables that strongly predict the ratings of the appropriateness of instructor behavior, future research will be necessary to discover these. Among unmeasured variables that could be examined are those related to the personality and learning style preferences of students.

Discussion

In combination, the results of the descriptive statistics (Research Question 1) and the multivariate analysis of variance (Research Question 2) revealed information that has theoretical and practical utility. A significant finding of this study was that most instructional behaviors are universally considered appropriate. Averages for the scales indicated that most respondents had rated instructional behaviors above a score of 3.00-as appropriate or very appropriate. The classroom instructional behaviors identified as appropriate were the "expressive" and "cognitive or information giving" behaviors such as giving multiple examples. The appropriate spatial-behavioral communication behaviors were those "expressive and/or warm" behaviors such as smiling, laughing, and speaking expressively or emphatically These results are similar to Murray's (1983) study and are consistent with results of research by Keith, Tarnatzky, and Pettigrew (1974). The appropriate behaviors identified also supported the findings of Wolfgang's (1977) study that identified the characteristics of a good teacher, such as being enthusiastic and using instructionally reinforcing spatial-behavioral communication (for example, gestures and smiles).

Despite the overall endorsement of many classroom behaviors, inferential statistics calculated for this study revealed significant effects of cultural background on scale means. The multivariate analysis of variance showed that there were significant differences between U.S. and international students in their perceptions of appropriate instructional behaviors. These differences were found on all four scales.

These results are consistent with statements of authors who have discussed the problems of international training. Students from the non-U.S. groups were likely to have had expectations for instructor behavior that were systematically different from the expectations of U.S. students. For example, if the instructor is seen as an authority figure who is primarily a purveyor of factual information, then behaviors that would be seen as desirable to U.S. students might not seem as desirable to international students. There may be a perception that learning is the student's responsibility and the instructor's role is to present material in a straightforward manner. This perception may be most prevalent in Eastern cultures, less so in the West, and least of all in the United States. Such perceptual differences would be consistent with results of the multivariate analysis of variance.

At first encounter, the descriptive statistics calculated to address Research Question 1 appear to contradict the significance tests used to address Research Question 2. On the one hand, it appeared that the majority of subjects stated that most behaviors were appropriate. On the other hand, differences existed among groups of subjects. However, these two facts are not in opposition. Although mean ratings of international groups were lower than U.S. students, this does not imply that international groups thought that most instructor behaviors in these areas were inappropriate. All groups had average ratings above 3.40, which meant that ratings for all three student groups were toward the appropriate side of the scales. Thus, differences among the three groups should be interpreted as differences in the strength of appropriateness.

Cultural theory (Archer, 1991; Goodman, 1994; Hall, 1980; Hampden-- Turner and Trompenaars, 1993; Hofstede, 1984, 1986, 1997; Samovar and Porter, 1991) states that there are differences in people's perceptions of behaviors because of different cultural and value systems. When people are from different cultures, the possibility of completely different perceptions of the same behavior increases a hundredfold (Archer, 1991). The results of significant group differences between U.S. and international students on all four scales supported cultural theory.

For example, spatial-behavioral communication behaviors were perceived as less appropriate by the Eastern-culture students than the U.S. students. This finding is similar to the results of Powell and Collier's (1990) study, which showed that Asian students did not feel it was appropriate for a teacher to be in close proximity to the students. Other spatial-behavioral communication behaviors include the casual teaching behavior of sitting on the desk during class. People from the United States are noted for their casual style, which is welcomed in many areas in the world. But U.S. instructors and trainers should note that students from other cultures might not like casual instructional behavior, although warm behaviors might be welcomed. This finding is consistent with the results of the previous study showing that Asian students prefer a formal relationship with the teacher. They may prefer formal instructional behaviors as well.

Multiple regression analyses were used to address the third research question, dealing with the effects of background variables on the average scale ratings. It was found that cultural background, defined as Western culture versus Eastern culture, had significant relationships with scores on the verbal scales (Clarity, Enthusiasm, and Interaction), but not with scores on the Spatial-Behavioral Communication scale. Western-culture students rated verbal behaviors as higher in appropriateness. All of these findings replicated what had been found in the Tukey multiple comparisons following the MANOVA.

For the scale Clarity, the variable field of study was a significant predictor. Students in the field of business had lower appropriateness ratings than the overall average of students in all fields (for example, science). This may have occurred because students in business have a practical mindset and expectations that instructors simply present the material without special clarification of the subject matter. For the scales Enthusiasm and Spatial-Behavioral Communication, the predictor age was inversely related to scores on the scale. In other words, younger students were significantly more likely to rate enthusiastic and spatial-behavioral communication behaviors as more appropriate than were older students. These results are likely to have occurred because younger students (in all cultures) are more accepting of informal behaviors and instructor actions that have emotional overtones.

The variables gender and length of stay in the United States did not show significant relationships with students' perceptions of instructional behavior. Thus, the finding that cultural differences were the paramount factor in predicting the instructor behaviors clarity, enthusiasm, and interaction can be generalized to both sexes and individuals of varying lengths of residence in the United States. Note, however, that for subjects in this sample, the median length of residence in the United States was twenty-four months. Most of the respondents (75 percent) were in residence for fifty-four months or less (less than five years).

Implications for HRD

As with all exploratory research, this study has several implications and limitations and it raises far more questions than it answers. Perhaps more than anything else this study supports the cultural theory view of training whose adherents hypothesize that cultural differences affect perceptions of instructor behaviors. Although this research found that most instructional behaviors are acceptable, many were found to be inappropriate to some cultures. This indicates that far more research is needed to determine the exact cultural expectations of every culture so that foreign-produced training materials and techniques can be appropriately modified.

The primary implication of this research is that when instruction is being organized, there should be explicit planning for international audiences, taking into account cultural differences. What works well with U.S. students might not work as well with other groups. Moreover, this study revealed that Eastern-culture students are likely to have greater divergence from U.S. students than Western-culture students. Planning for such differences before the instruction begins is preferable to making on-the-spot changes as a result of negative reactions from students.

Another implication involves the evaluation of instruction. Many educational and training experiences are evaluated by having participants rate the instructor and the instruction. Such evaluative questionnaires are common after training experiences. Given the differences in appropriateness ratings revealed by this study, different forms will be required for each cultural group. In addition, because persons from different cultures differ in the perception of appropriateness of training behaviors, they may well differ in how they evaluate the effectiveness of instruction. As a result, it will be difficult to compare ratings from one culture to another. This suggests that there cannot be a universal evaluation form or a universal standard to compare the effectiveness of training across cultures.

Limitations of This Study and Recommendations for Future Research

One limitation of the study is that it was dependent on a single data collection method: five-step rating scales by respondents who rated the appropriateness of instructional behaviors. Rating scales are sometimes subject to response sets-general behavioral tendencies that affect the ratings on an entire instrument. If such a response set interacted with cultural group (that is, was more evident in one group than another) then the results of this could have been affected or confounded. For example, one culture may have a response set in which extreme responses are avoided in favor of more moderate responses.

Other study limitations include the use of only one university as the population. In addition, multiple disciplines were included. Perhaps future research could involve a larger population and segregate disciplines to determine if location or discipline are important considerations.

Perhaps the most important limitation of this research was the use of university students as the population for the study Although the literature is dear that this is an appropriate protocol for studies such as this, there is little doubt of the possible bias introduced by schooling in the United States.

International training is perhaps the most exciting assignment an HRD professional can get. It is also one of the most difficult because of the number of variables that must be considered. Because an instructor is successful with one audience is not an automatic indicator that he or she will be able to achieve that success with an audience from a different culture. This research found that most instructional behaviors are universally accepted (at least with the populations sampled), but some of the things that instructors do in the classroom may not be interpreted the same way by individuals from different cultural backgrounds. These culturally predisposed differences must be addressed when transplanting training from one location to another. More than anything else, this research indicated the need for a greater understanding of the instructional behaviors appropriate when training cross-culturally.

References

Archer, C. M. (1991). Living with strangers in the USA. Englewood Cliffs, NJ: Prentice-Hall. Baker, G. E., Mohammed, T., dt Boyle, M. A. (1994). A three axis model for conceptualizing the relevant elements in cross-cultural training. Performance Improvement Quarterly, 7 (2), 27-37.

Chinese Culture Connection. (1987). Chinese values and the search for culture-free dimensions of culture. Journal of Cross-Cultural Psychology, 18, 143-164.

Chow, I. H-S. (1995). Management education in Hong Kong: Needs and challenges. International Journal of Educational Management, 9 (5), 10-15.

Du Bois, C. D. (1956). Foreign students and higher education in the United States. Washington, DC: American Council on Education.

Goodman, N. R. (1994). Cross-cultural training for the global executive. In R. W Brislin & T. Yoshida (Eds.), Improving intercultural interactions (pp. 34-54). Thousand Oaks, CA: Sage. Hall, E. T. (1980). The silent language. Westport, CT: Greenwood Press.

Hampden-Turner, C., & Trompenaars, A. (1993). The seven cultures of capitalism. New York: Doubleday

Harris, P R. (Ed.). (1984). Global strategies for human resource development. Washington, DC: American Society for Training and Development.

Hecht, M. L., Andersen, P A., St Ribeau, S. A. (1989). The cultural dimensions of nonverbal communication. In M. K. Asante dT W B. Gudykunst (Eds.), Handbook of international and intercultural communication (pp. 163-185). Thousand Oaks, CA: Sage.

Hinkle, D. E., Wiersma, W, &Jurs, S. G. (1994). Applied statistics for the social sciences (3rd ed.). Boston: Houghton Mifflin.

Hofstede, G. (1980). Motivation, leadership, and organization: Do American theories apply abroad? Organizational Dynamics, 9 (1), 42-63.

Hofstede, G. (1984). Culture's consequences: International differences in work-related values. Thousand Oaks, CA: Sage.

Hofstede, G. (1986). Cultural differences in teaching and learning. International journal of Intercultural Relations, 10, 301-320.

Hofstede, G. (1997). Cultures and organizations: Software of the mind. New York: McGraw-Hill. Huang, Z. (1996). Making training friendly to other cultures. Training and Development, 50 (9), 13-14.

Keith, T. L., Tamatzky, L. G., & Pettigrew, L. E. (1974), An analysis of verbal and nonverbal classroom teaching behaviors. Journal of Experimental Education, 42, 30-38.

Kohls, L. R. (1995). Training know-how for cross-cultural and diversity trainers. Duncanville, TX: Adult Learning Systems.

LaFrance, M., & Mavo, C. (1978). Moving bodies: Nonverbal communication in social relationships. Pacific Grove, CA: Brooks/Cole.

Laurent, A. (1983). The cultural diversity of Western conceptions of management. International Studies of Management and Organization, 8 (1-2), 75-96.

Malandro, L., dt Barker, L. (1983). Nonverbal communication. Reading, MA: Addison-Wesley Marquardt, M. (1993). The impact of the cultural environment on HRD. In A. Reynolds & L. Nadler (Eds.), Globalization: The international HRD consultant and practitioner (pp. 9-132). Amherst, MA: Human Resource Development Press.

Marquardt, M. (1995, May). How to globalize your training. Alexandria, VA: American Society for Training and Development.

Mehrabian, A. (1972). Nonverbal behavior Hawthorne, NY: Aldine de Gruyter.

Morical, K., & Tsai, B. (1992). Adapting training for other cultures. Training & Development, 46 (43. 65-68.

Murray, H. G. (1983). Low-inference classroom teaching behaviors and student ratings of college teaching effectiveness. Journal of Educational Psychology, 75, 138-149.

Neill, S. (1991). Classroom nonverbal communication. London: Routledge.

Nunnally, J. C., di & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). New York: McGraw-Hill. Pedhazur, E. J. (1982). Multiple regression in behavioral research (2nd ed.). Austin, TX: Holt, Rinehart and Winston.

Powell, R. G., & Collier, M. J. (1990). Ethnicity, instructional communication and classroom systems. Communication Quarterly, 38, 334-349.

Samovar, L. A., di Porter, R. E. (Eds.). (1991). Intercultural communication: A reader (6th ed.). Belmont, CA: Wadsworth.

Schermerhorn, J. R. (1994). Intercultural management training: An interview with Asthma Aldullah. Journal of Management Development, 13 (3), 47-64.

Shabahang, J. (1993). Perceptions of American students, faculty, and administrators on the impact of international students on selected community colleges. Unpublished doctoral dissertation, George Washington University, Washington, DC.

Stevens, J. (1996). Applied multivariate statistics for the social sciences (3rd ed.). Hillsdale, NJ: Erlbaum.

Swing, N. (1984). Training in someone else's backyard. Training and Development Journal, 38 (10), 22-23.

Thomas, R. (1995). The rising tide of global training. Training and Development, 49 (2), 9-10. Thornhill, A. R. (1993). Management training across cultures: The challenge for trainers. Journal of European Industrial Trainer, 17 (7), 43-51.

Vijver, E, & Leung, K. (1997). Methods and data analysis for cross-cultural research. Thousand Oaks, CA: Sage.

Wolfgang, A. (1977). The silent language in the multicultural classroom. Theory into Practice, 16, 145-152.

[Author note]

Fengjiao Ji Burba is a senior instructional designer with the E-Learning Development Group, Global Customer Training, Tellabs Operations, in Bolingbrook, Illinois.

Joseph M. Petrosko is a professor of educational foundations at the University of Louisville in Louisville, Kentucky.

Mike A. Boyle is associate professor of human resource education at the University of Louisville in Louisville, Kentucky.

--------------------------------------------------------------------------------

Reproduced with permission of the copyright owner. Further reproduction or distribution is prohibited without permission. BasicAdvancedGuidedPublicationNatural languagePersonal namesLocations & placesCompaniesSubjectsLast searchRecent searchesMarked list & durable links