international student stress due in 20 hours

Research in Higher Education, Vol. 33, No. 5, 1992

ACADEMIC STRESS OF INTERNATIONAL STUDENTS ATTENDING U.S. UNIVERSITIES

Teh-yuan Wan, David W. Chapman, and Donald A. Biggs

ù ù o ° , ° . . . . • . . . . • o • . , . . . . . . . . . • . . . . . . . . . . . . . ù . . . . . . . . . . . . . . . . . ° . . ù ° o . ° . . . . . . .

This study investigated factors associated with the academic stress experienced by international students (N = 412) attending graduate school in the United States. The study was grounded in a cognitive framework in which academic stress is under- stood as the consequence of students' appraisal of the stressfulness of role de- mands and their perception of their ability to cope with those demands. These two appraisals, in turn, were posited to be a function of cultural distance, students' social support network, and their role competencies. Results indicated that the primary determinants of the two types of appraisals differed. While self-perceived English- language skills and, to a lesser degree, cultural distance were the predictors of pri- mary appraisal, self-perceived English-language, academic, and problem-solving skills and social support network were the main determinants of secondary ap- praisal. Implications for how universities might werk more effectively with interna- tional students are discussed.

, , , , , , ù ù , o , J , , , , , , o , , o , , , , , , , ù , , , ù ù ù , , , , , o , , ù , , , , . . . . . ù , , , ù , o , o , , , ù , , , . . . . . . o , . . . . . ~ . . . . . . • . . . . . . . o . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ù . . . . . . . . . o ,

The education of international students at the graduate level has become a major activity of U.S. universities (Chapman, Wan, and Xu, 1988; Altbach, Kelly, and Lulat, 1985). At present, international students represent over 20 percent of all graduate enrollment in the United States. For the most part, these students are eagerly courted by American institutions because they offer an important diversity of viewpoint to the student body, help offset a declining American applicant pool, and often come with full funding. At the same time, there is a growing concern about the ability of U.S. universities to adequately address the cross-cultural adjustment problems these students experience once they arrive on campus. International students often find their academic experi- ences to be extremely stressful. Academic demands are heavy, instructor-stu- dent interaction is often fast-paced, and the academic and social support mecha- nisms available to undergraduate foreign students often are not as relevant or as available to graduate students. Classroom experience can involve role expecta-

Teh-yuan Wan, New York State Education Department. David W. Chapman and Donald A. Biggs, State University of New York at Albany. Address correspondence to David W. Chapman, 314 School of Education, State University of New York at Albany, Albany, NY 12222.

607

0 3 6 1 - 0 3 6 5 / 9 2 / 1 0 0 0 - 0 6 0 7 5 0 6 . 5 0 / 0 © 1 9 9 2 H u m a n S c i e n c e s P r e s s , I n c .

608 WAN, CHAPMAN, AND BIGGS

tions that are not clearly defined and that conflict with their previous academic experiences.

The resulting stress can have a strong, negative impact on international stu- dents' experiences in the U.S. educational environment and on their ability to transfer the knowledge and skills they learn in U.S. classrooms to their home countries. However, it cannot be assumed that international students will find similar cross-cultural educational experiences to be equally stressful. Individ- uals differ in their perceptions of the severity of the same academic demands and in their perceptions of their resources for coping with these demands. The present study investigated factors associated with level of academic stress expe- rienced by international students in U.S. universities. To the extent that factors associated with stress can be identified, educators would be better able to coun- sei individuals experiencing such stress, to take institutional actions to help alleviate the sources of the stress, and to prepare incoming students to better handle the pressures they are likely to encounter.

This study employed a cognitive framework based on the work of Lazarus (1974, 1976, 1981) and Lazarus and Cohen (1977), which posits that (1) aca- demic stress is the consequence of students' cognitive appraisal of the stressful- ness of role demands (primary appraisal) and their perception of their ability to cope with those demands (secondary appraisal) and (2) differences in these appraisals of academic stress among students can be influenced by the cultural differences between the home country and the United States, the relative levels of social support available to students, and differences in students' perceived role skills (role competency).

BACKGROUND

Cross-cultural contacts are inherently stressful (Hull, 1978). While stress can be an important component for personal and professional development, too much can negatively affect international students' social adaptation, personal development, and academic attainment (Biggs, 1992; Spaulding and Flack, 1976; Church, 1982; Altbach, Kelly, and Lulat, 1985). The stress experienced by international students tends to center on academic situations. Indeed, re- search indicates that, for international students, problems associated with aca- demic experiences are the most stress-provoking and tend to be more persistent than social or personal problems (HuU, 1978; Klineberg and Hull, 1979; Selby and Woods, 1966; Byrnes, 1966; Higbee, 1969; Sharma, 1973; Hill, 1966). Stress associated with acadernic experiences can negatively effect students' aca- demic performance (Rust, 1960; White, Brown, and Suddick, 1983), mental or physical health (Cole, Allen, and Green, 1980; Eldridge, 1960; Still, 1961; Ward, 1967), level of satisfaction with their cross-cultural experiences, and attitude change toward the host nationals (Veroff, 1963; Hull, 1978; Klineberg

ACADEMIC STRESS 609

and Hull, 1979). Nonetheless, while stress due to problems encountered in academic adjustment has been widely researched, findings have been inconsis- tent and sometimes conflicting. One reason has been the variety of ways in which stress has been defined and operationalized (see, for example, Dohren- wend and Dohrenwend, 1974; Draguns, 1980; Lin and Ensel, 1984; Argyle and Kenton, 1967). Particularly troublesome to the definitional effort has been the recognition that different people responding to the same situation often experi- ence quite different levels of stress.

One view of stress that would account for such differences relies on concepts of adjustment in which stress is viewed as an indication of health or sickness. Alternatively, stress often has been considered a maladaptive response to situa- tional demands. A third explanation of stress is a cognitive mediated model in which individuals male cognitive appraisals of environmental demands (La- zarus and Cohen, 1977; Lazarus, 1981). Within a cognitive framework, stress occurs when the environmental demands are evaluated as exceeding a person's resources for coping with them. It is this cognitive framework in which the present study is grounded.

Lazarus (1976) defines stress as a consequence of a transactional process in which people make two kinds of cognitive appraisals of role demands in a situation. Primary appraisal is the cognitive process of evaluating educational events or experiences in terms of the severity of the demands. Secondary ap- praisal is the cognitive process of evaluating the coping resources and options that are available to deal with the stressful situations (Lazarus, 1981). Differ- ences in levels of stress among students are considered a function of differences in their cognitive appraisals. These two appraisals, in turn, are posited to be a function of (1) the cultural distance between the host and the students' home culture, (2) students' social support networks, and (3) their role competencies.

Cultural distance refers to the extent that a student's home culture differs from the predominant culture of the U.S. (Babiker, Cox, and Miller, 1980). For example, if the cultural distance between Bangladesh and the U.S. is seen as greater than the distance between the U.K. and the U.S., students from Bangladesh would be expected to experience more stress in a U.S. classroom than students from the U.K. Social support refers to the extent to wbich stu- dents have a network of friends in the host culture who offer them encourage- ment, support, and advice. Friends within the social support network operate both to help interpret the new culture to the international student and to rein- force the individual's setf-confidence. The role skills (or competencies) posited to be important factors in academic stress included students' academic skills, problem-solving skills, English-language skilts, and their prior knowledge of the U.S. educational environment. Finally, personal characteristics, including students' age, academic status, field of study, and the length of time they had been in the United States, were entered into the analysis as a set of control

610

I Role Ski l ls

o Social SkUls o Academic Skills o Problem-solving

skills o English Skills o Prior Know{edge

WAN, CHAPMAN, AND BIGGS

Socia l Suppo r t

0 Socia[ Support Network

Cul ture D is tance

o cunure-distance Index

o Differenee in Educ. System

L Charac ter is t i cs ~ ~~~i~i~~ii:~! I o ~oo o Sex I o Aeademic Status o Field ot Study o Length of Study

in U.S. j

FIG. 1. The operational model of academic stress.

variables. Since these characteristics are often related to students' experience of academic stress but were not among the variable sets being investigated, it was important to remove their effects prior to entering the independent variables in the analysis. The conceptual framework for this study is presented as Figure 1 and the particular manner in which each of these variable sets was opera- tionalized is discussed below.

PROCEDURE

During the spring semester of 1989, 689 international graduate students en- rolled in three major universities in upstate New York were selected randomly and invited by mail to complete the Survey of Academic Experiences of Inter- national Students (SAEIS). The SAEIS collected respondents' ratings of the stressfulness of selected academic situations, ratings of their coping resources, and measures of their role skills, social support network, culture distance, and selected demographic information. A total of 412 usable questionnaires were returned for a response rate of 60 percent. Multiple regression was used to test the extent that the four sets of situational and personal variables were related to students' appraisal of the stress associated with academic adjustment (primary appraisal) and, second, their ability to cope with that stress (secondary ap- praisal).

INSTRUMENTATION

The Survey of Academic Experiences of International Students (SAE1S) con- tained measures operationalizing the dependent variables (primary and second- ary appraisal of academic stress) and the four independent variable sets. Situa- tional stress was operationalized using a magnitude estimation scaling (MES)

ACADEMIC STRESS 611

procedure. The independent measures e~nployed in the study were adopted from instruments previously developed b~ other researchers.

Dependent Variables

The two dependent variables used in this study were students' rating of the stressfulness of selected academic situations (primary appraisal) and their rat- ings of their ability to cope with that stress (secondary appraisal). For these ratings to reflect differences in individuals' perception of stress rather than dif- ferences in the stressfulness of the situations themselves, it was necessary to held constant the stressfulness of the academic situations being rated. This was accomplished using a magnitude estimation scaling procedure.

MES is a numeric estimation procedure that has been widely used for con- structing ratio scales of sensation (Lodge, 1981; Stevens, 1975). Using MES, the stress associated with selected situations was scaled (during an instrument development phase) and only situations of comparable stress were retained for the main study. This provided a means of controlling for differences in the stressfulness of the stimulus situations being rated so that differences ratings could be attributed to variation in respondents' perceptions rather than to differ- ences in the situations themselves. Specifically, in developing the measure of academic stress, a pool of 43 vignettes was developed, each describing a stress- ful academic situation. The stressfulness of each item was rated by a sample of international students. Using MES, the stressfulness of the vignettes was scaled (from 1 = low to 10 = high stress) and ten vignettes rated as highly stressful were selected for use as dependent variables in the full study. During the full study, respondents rated the stressfulness of the academic situations represented in each of these ten vignettes on a seven-point scale, ranging from highly stressful to not stressful. Second, they rated their confidence in their ability to cope with the situation represented in each vignette on a seven-point scale (very confident to not at all confident). A respondent's primary appraisal score (alpha = .88) was computed as the mean primary appraisal rating across the ten vignettes while a person's secondary appraisal score (alpha = .89) was the mean secondary appraisal rating across the ten vignettes.

Independent Variables

Two indicators of cultural distance were employed in this study: the culture distance index (CDI) developed by Babiker, Cox, and Miller (1980) and the difference in educational systems scale. The CDI divides international students into three subgroups according to the differences and similarities in religion, language, and home country climate. This measure of cultural distance was found to be significantly correlated (r = .62, p < .0001) with a measure of

612 WAN, CHAPMAN, AND BIGGS

cross-cultural stress (Furnham and Bochner, 1982). The six-item difference in educational systems scale was developed as a supplement to the CDI to reflect the concept of cultural distance within the more specific context of an academic environment. This scale elicited information on the extent of cultural differ- ences on six aspects of academic life, ranging from the educational system in general to the role of students in specific programs. The internal consistency reliability of the DES (alpha = .80) was adequate for purposes of this study.

Two social support subscales (emotional support and instrumentai support) of the life satisfaction inventory (Sorenson, 1988) were used in this study to collect information on perceptions of the availability of people who could pro- vide encouragement, comfort, and practical kinds of help.

Role skills was operationalized using five variable sets: role competence, academic skills, problem-solving skills, English-language skills, and prior knowledge of the host culture. Role competence was measured by combining two subscales (self-management competence and socials skiUs) from the life satisfaction battery (LSB) developed by Sorenson (1988). The scale, consisting of twenty items, measured skills in self-management and skills in getting along with others.

The academic skills scale was drawn from the College Opinion Survey (COS) originally adapted by Biggs, Roth, and Strong (1970) and Biggs and Tinsley (1970) from the Michigan State University's general self-concept of ability to learn scale (GSCA). The scale had been validated with both American high school and college students with satisfactory reliability and validity (Biggs, Roth, and Strong, 1970; Biggs and Tinsley 1970). The measure con- sisted of five items drawn from the COS asking respondents to estimate their academic abilities compared with those of different reference groups of stu- dents, such as fellow students in college in home countries, other international students in the same current graduate programs, and American students in the current graduate programs. Respondents were also asked to report their actual and predicted academic performance (GPA).

Problem-solving skills of international students were measured by using the attitude subscale of the problem-solving inventory (PSI), developed by Heppner (1986), which assesses respondents' attitudes about problem-solving skills. The purpose of the PSI is to assess people's perceptions of their problem-solving behaviors and attitudes. Heppner posited that people with more positive ap- praisal of their problem-solving ability tend to have higher expectations of suc- cess and are more motivated to solve problems. Studies with the PSI have supported this hypothesis (Neal and Heppner, 1982; Heppner, Hibel, Neal, Weinstein, and Rabinowitz, 1982; Phillips, Pazienza, and Ferrin, 1984). While the PSI has three subscales, problem-solving ability, approach-avoidance style, and personal control, only the problem-solving ability subscale was adopted for this study, to serve as a fourth measure of role skills. This subscale had

ACADEMIC STRESS 613

been validated with three different groups of subjects and has an averaged inter- nal consistency reliability of .82 (Heppner, 1986).

Perceived English-language ability was included as a role skill, based on substantial research evidence that perceived language proficiency is a strong and consistent predictor of cross-cultural stress experienced by international students in social interactions, academic adjustment, and personal development (White, Brown, and Suddick, 1983). It was measured as the respondent's total score on the language proficiency scale (LPS) developed by Chapman, Wan, and Xu (1988), which measures English-language ability in six general areas related to academic life: (1) understanding spoken English, (2) reading compre- hension, (3) writing academic papers/research reports, (4) participating in class discussion, (5) taking notes in class, and (6) conversing with faculty and Amer- ican students. Chapman, Wan, and Xu (1988) reported an internal consistency reliability of .89 for this scale.

Prior knowledge of role expectations in a new environment was measured using a four-item subscale adopted from the International Student Adjustment Survey (ISAS) developed by Chapman, Wan, and Xu (1988). The prior knowl- edge subscale elicits information from respondents on (a) their prior knowledge of the U.S. educational system in general, (b) their prior knowledge of the specific academic program in which they were enrolled, (c) their prior knowl- edge of teaching methods and classroom practice, and (d) their prior knowledge of the academic program requirement and expectations. The internal consis- tency reliability of the subscale, when assessed with a group of international graduate students enrolled at a major American university, was .72 (Chapman, Wan, and Xu, 1988).

The SAEIS also collected information on students' demographic characteris- tics: age, gender, academic status, national origin, field of the study, and dura- tion of stay in the United States.

SAMPLE

The sample size of this study was determined through a consideration of the statistical power needed in the analysis and the anticipated response rate. Spe- cifically, minimum sample size was determined using Cohen's (1988) formula and tables and it assumed a medium effect size, a power of .80, and an alpha set at .025. A sample size of 275 subjects out of the total 2,222 international graduate students enrolled in the three universities was estimated as sufficient to reach 80 percent power. The actual sample was 689 students, more than twice the number needed, to adjust for the possibility of nonrespondents.

The 689 students participating in this study represented 31 percent of the 2,222 international graduate students enrolled at the three institutions during the 1988-89 academic year. The profile of respondents, presented in Table 1, indi-

614 WAN, CHAPMAN, AND BIGGS

TABLE 1. Profile of Respondents

Population Respondents Chi Characteristics (N = 2,222) (N = 406) Square

Gender male 75% 75% .72 female 25% 25% (df = 1)

Academic Status master's 43% 40% doctoral 52% 57% 10.23 nondegree 5% 3% (df = 2)

Geographic Region Asia 71% 68 % Middle East 7% 6% 17.58" Africa 4% 3% (df = 4) S. Europe & Latin Arnerica 6% 12% N. Europe 12% 11%

Field of Study Engineering 33% 32% Business & Management 10% 15% Math & Computer Science 11% 14% 16.71' Physical & Life Science 13% 15% (df = 8) Social Science 15% 3% Fine Arts 2% 1% Health Science 2% 2% Humanities 7% 10% Education 7% 9%

*p -< 01.

cated that the majority were male (75%), from Asian countries (68%), and enrolled in natural and applied science programs (61%) for their doctoral de- grees (57%). A chi-square procedure was used to test the significance of differ- ences between the respondent group and the overall population of international students at the three universities on selected dernographic characteristics. The respondents from the southern European countries were overrepresented in the sample (X 2 = 17.58, df = 4, p < .05) while the proportion of students in social science were underrepresented by the respondents (×2 = 16.7, df = 8, p < .05).

DATA ANALYSIS

Multiple regression analyses were computed to test the research questions. The first regression analysis was done to determine the extent that respondents'

ACADEMIC STRESS 615

demographic characteristics, cultural distance, social support network, and role skills were related to the amount of stress they perceived in stressful academic situations. The second regression analysis was computed to determine the ex- tent that these variables were related to respondents' perception of their ability to cope with this stress.

RESULTS

The regression to predict respondents' primary appraisal of stress, reported in Table 2, yielded an overall F ratio of 40.9, significant at the p < .001 level. Cultural distance, social support network, role skiils, and demographic charac- teristics together explained 44 percent of the variance in primary appraisal of stressfnl academic situations. Each of the four variable sets contributed to sta- tistically significant increases in the amount of explained variance. Eight per- cent of the explained variance could be attributed to respondents' demographic characteristics (F = 2.93, p < .01). Within this variable set, the primary predictor was students' academic status (univariate F = 5.30, p < .025). Stu- dents who were working on more advanced degrees tended to perceive the academic situations as less stressful. This finding is consistent with earlier re- search on international students that reported a negative relationship between academic standing and the level of stress students experience (Melby and Wolf, 1961; Hull, 1978; Moleni, 1986). The other demographic characteristics--age, gender, length of study, and major field of study--were not meaningfully re- lated to level of stress.

After controlling for the effect due to demographic characteristics, culture distance (e.g., the differences in education systems scale) accounted for an additional 3 percent of the variance (F = 7.35, p < .001). Students from countries in which the educational system was perceived as more distant from the U.S. perceived their academic experiences as more stressful than did stu- dents from countries having more similar education systems. Contrary to find- ings in earlier research, the culture distance index was not significantly related to stress (Hull, 1978; Furnham and Bochner, 1982; Babiker, Cox, and Miller, 1980). This finding may be due to the limited variability in the range of re- spondents' home countries. The majority of the respondents (and the popula- tion) in this study came from Asian countries.

After controlling for differences in demographic characteristics and culture distance, the extent of respondents' social support network accounted for an additional one percent of the variance in students' primary appraisal of stress (F = 4.03, p < .025). International students who had a stronger social support network in their new academic environment tended to rate the same academic situations as less stressful than did students with less well-developed social networks. However, the strength of association was relatively weak.

TABLE 2. Summary of Regression to Predict Primary Appraisal of Academic Stress

St. F o f R 2 Univariate Independent Variables ~ Mean Dev. R 2 Change Beta F-Ratio

Personal Background .09 2.93** Age 28.29 5.82 .06 .84 Gender 1.27 .53 - . 0 8 3.35 Academic status 2.54 .56 - . 11 5.30 Length of study in U.S. 33.56 20.10 - . 0 3 .59 Academic major

Engineering .20 .60 .05 .67 Math & computer .04 .47 - . 0 5 .86 Business &

management .05 .48 .06 .96 Physical science .06 .49 .02 .11 Social science .05 .34 .01 .03 Humanities .01 .43 - .04 .14 Health science .07 .32 - . 0 5 .32 Fine & applied art .08 .30 .02 .17 Education .09 .28 - - - -

Cultural Distance .12 7 .35*** Country of Origin 2.87 2.63 - . 0 4 .82 Difference in Education

System 4.41 1.24 .16 12.47"** Social Support Network .13 4.03* .03 .52 Role Skills .44 40.90***

Academic skills 4.12 .59 - .08 1.78 Social skills 4.95 .76 - . 0 7 1.20 Problem-solving skills 4.66 .70 - . 0 8 2.64 Prior knowledge of U.S. 3.64 1.00 - . 0 2 .14 Language skills 4.56 .99 - . 5 6 137.17"**

* = p < .025;** = p < .01;*** = p < .001 '~Variables and their scale of measurement

Personal Background Age Gender Academic status Length of study in U.S. Academic major

Cultural Distance Country of origin Difference in

in years 1 = male; 2 = female 1 = nondegree; 2 = master's; 3 = doctorate in years 0 = no; 1 = yes

(l = near; 2 = intermediate; 3 = far from host culture)

Education System 6-item scale; alpha = .80; l = no difference to 7 = great difference Social Support Network 20-item scale; alpha = .93; support available: 1 = never to 7 = always Role Skills all items scaled 1 = poor to 7 = excellent

Academic skills 5-item scale; alpha = .84 Social skills 20-item scale; alpha = .86 Problem-solving skills 11-item scale; alpha = .72 Prior knowledge of U.S. 4-item scale; alpha = .64 Language skills 6-item scale; alpha = .91

ACADEMIC STRESS 617

Role skills were the primary determinant in students' appraisal of stress, accounting for 31 percent of the variance, even after controlling for the effect of the earlier variable sets (F = 40.90, p < .001). Among the five role skills, English-language skill was the only significant contributor to the explained variance (univariate F = 137.17, p < .001). Students who perceived them- selves as having bettet English-language skiUs perceived the same academic situations as less stressful than did students who perceived themselves to have weak language skills. The other role skills (social skills, academic skills, prob- lem-solving skills, and prior knowledge) did not account for a significant addi- tion of variance in primary appraisal of stress, which suggests that neither students' prior knowledge of the United States, self-perceived social skills, self-perceived academic skills, nor their attitudes about problem solving are significantly related to their evaluation of the stressfulness of academic situa- tions. English-language skills appear to override all other concerns, which sug- gests that international students' perceived language skills have the most signif- icant influence on their appraisal of the stressfulness of classroom situations.

In the second analysis, the same four independent variable sets explained 50 percent of the variation in students' appraisal of their ability to cope with aca- demic stress (F = 51.45, p < .001). (Table 3). Of that, most of the variance was explained by students' role skills and the extent of the social support net- work available to them. Only 8 percent of the variation was explained by dem- ographic characteristics (F = 2.78, p < .01), with age, length of study in the U.S., and major field of study of minor importance. Younger students and students who had been studying for a longer period of time in the U.S. tended to have more confidence in their coping skills, a result consistent with those of earlier studies (Church, 1982; Spaulding and Flack, 1976; Altbach, Kelly, and Lulat, 1985). Additionally, students majoring in engineering, mathematics, and computer science tended to rate the same situations as less stressful than stu- dents in other fields.

After controlling for the effects of demographic characteristics and culture distance, social support network accounted for an additional 4 percent of the variance (F = 14.52, p < .001). Students who reported more social support tended to have more confidence in their ability to cope with stressful academic situations.

Role skills was the most important correlate of students' secondary appraisal, accounting for 36 percent of the variance after controlling the effect of the earlier variable sets (F = 151.45, p < .001). Academic skills, problem-solving skills, and perceived English-language skills were the significant predictors of secondary appraisal of academic stress. Students with more positive evaluations of their academic ability, more positive attitudes about their problem-solving skills, and/or more positive perceptions of their English-language abilities tended to give better evaluations of their coping skills than students who be- lieved they had weaker skills in those areas.

TAILLE 3. Summary of Regression to Predict Secondary Appraisal o f Academie Stress

St." F of R 2 Univariate Independent Variables a Mean Dev, R 2 Change Beta F-Ratio

Personal Background .08 2.78** Age 28.29 5.82 - .09 4.90* Gender 1.27 .53 - . 0 8 3.30 Academic status 2.54 .56 .04 .86 Length of study in U.S. 33.56 20.10 - . 1 2 5.56* Academic major

Engineering .24 .60 - . 14 6.31" Math & computer ,04 .47 - . 12 5.35* Business &

management .05 ,48 - .01, .02 Physical science .06 .49 .01 .02 Social science .05 .34 - . 0 5 .55 Humanit ies .01 .43 - .04 .14 Health science .07 .32 .16 3.01 Fine & applied art .08 .30 .07 1.57 Education .09 .28 - - - -

Cultural Distance .09 2.06 Country o f Origin 2.87 2.63 .34 .47 Difference in Education

System 4.41 1.24 - .08 3.11 Social Support Network 4.42 .13 14.52"** .04 .72 Role Skills .49 51 .45"**

Academic skills 4.12 .59 .19 15.58"** Social skills 4.95 .76 .01 .02 Problem-solving skills 4.66 .70 .16 10.42"* Prior knowledge o f U.S . 3.64 1.00 - .03 .57 Language skills 4.56 .99 .47 107.65"**

* = p < .025;** = p < .01;*** = p < .001 ~Vafiables and their scale of measurement

Personal Background Age Gender Academic status Length of study in U.S. Academic major

Cultural Distance Country of origin Difference in

in years 1 = male; 2 = female t = nondegree; 2 = toaster's; 3 = doctorate in years 0 = no; 1 = yes

(1 = near; 2 = intermediate; 3 = far from host culture)

education system 6-item scale; alpha = .80; 1 = no difference to 7 = great difference Social Support Network 20-item scale; alpha = .93; support available: 1 = never to 7 = always Role Skills all items scaled 1 = poor to 7 = excellent

Academic skills 5-item scale; alpha = .84 Social skills 20-item scale; alpha = .86 Problem-solving skills l 1-item scale; alpha = .72 Prior knowledge of U.S. 4-item scale; alpha = .64 Language skills 6-item scale; alpha = .91

ACADEMIC STRESS 619

DISCUSSION

The formulation of stress in terms of a dual (primary and secondary) ap- praisal process helped overcome limitations encountered in previous research, much of which operationalized stress by asking respondents to rate the overall stressfulness of cross-cultural experiences. Results indicated that the primary determinants of the two approaches to evaluating academic stress differed. Some of the factors most strongly related to students' evaluation of a situation as stressful were different from those most strongly related to students' evalua- tion of their ability to cope with that situation. Specifically, perceived English- language skill and, to a lesser extent, social support network were related both to students' primary and secondary appraisals of the stressfulness of academic situations. However, perceived academic and social skills were related only to secondary appraisal.

One implication of this finding is that interventions to assist international students in adjusting to the academic demands of U.S. graduate education would differ, depending on whether the intervention is aimed at reducing the perceived stressfulness of the situation or in bolstering students' ability to cope with the situation. As a practical matter, institutional interventions probably would be better aimed at strengthening students' ability to cope with stress than in trying to reduce the initial stressfulness students are likely to experience, since the determinants of secondary appraisal appear to be more easily manipu- lated through interventions. Results of this study suggest the two most promis- ing interventions are to strengthen students' perceived English-language skills and their social support networks.

The contribution of perceived language skill in reducing stress outweighed the contribution of all other variables in the analyses combined. Students with stronger language skills were less likely to view academic situations as stressful and believed they were more able to cope with the stress they dir experience. Hence, English-language skills have a compound effect; students who believed they had weak English language skills were twice penalized. These findings reinforce the findings of earlier research, that the effectiveness of instructional support and advisement systems to assist international students may be limited if students do not believe they have the language skills necessary to handle the academic workload. The best academic support programs may fall if students do not have the confidence in their language skills to engage productively in their graduate experience (Chapman, Wan, and Xu, 1988).

Fortunately, English-language proficiency of incoming students is one of the factors most easily influenced, either through institutional admissions policy or through direct instructional intervention. Prior language training for students planning to study in the U.S. can be provided by the home government, private groups, or the receiving institution. However, the impact of that intervention on

620 WAN, CHAPMAN, AND BIGGS

reducing subsequent academic stress depends heavily on having the appropriate emphasis in the language training. The English training international students receive often is designed to enable them to pass the standardized Test of Eng- lish as a Foreign Language (TOEFL), widely required for admission to gradu- ate programs in the U.S. (Light, Xu, and Mossop, 1987; Wan and Xu, 1987; Chapman, Wan, and Xu, 1989). This type of language training often falls to adequately help students meet the academic demands of their programs, once admitted. International students can have satisfactory scores on TOEFL and still have an enormous amount of difficulty in their academic pursuits (Light, Xu, and Mossop, 1987; Wan and Xu, 1987; Xu and Wan, 1990). Results of this study suggest that their eventual academic adjustment depends on training that includes attention to functional language skills--such things as note tak- ing, conversing with faculty, and participating in class discussion (Chapman, Wan, and Xu, 1988).

In addition to language skills, students who believed they had a stronger social support network perceived themselves to be more capable of coping with stressful academic situations, which suggests that a promising means by which universities can ease international students' academic adjustment is through as- sistance in establishing these networks. While the importance of this interven- tion is widely understood at the undergraduate level, universities often take less of a role in facilitating networks at the graduate level. Graduate students are more likely to live off-campus, be more focused on their area of specialization, and be more independent than undergraduates. Nonetheless, the present find- ings suggest that active efforts to encourage positive social networks can pay oft. At the same time, even when they are available, international students often fail to take advantage of these types of programs (Dodge, 1990). One reason is that, for some international students, the new educational environrnent is so confusing, ambiguous, and overwhelming that they tend to wrap them- selves up in their academic struggles and appear indifferent to other aspects of academic life on campus. Their academic stress is, then, further compounded by the isolation from the campus community as they overlook the support avail- able to them. To the extent that this is the case, institutions taust be persistent in reaching out to these students.

REFERENCES

Albach, P. G., Kelly, D. H., and Lulat, Y. G-M. (1985). Research on Foreign Students and International Study: An Overview and Bibliography. New York: Praeger.

Argyle, M., and Kenton, A. (1967). The experimental analysis of social performance. In L. Berkowitz (ed.), Advances in Experimental Social Psychology, vol. 3. New York: Academic Press.

Babiker, I. E., Cox, J. L., and Miller, P. M. C. (1980). The measurement of culture distance and its relationship to medical consultation, symptomatology and examina-

ACADEMIC STRESS 621

tion performance of overseas students at Edinburgh University. Social Psychiatry. 15: 109-116.

Biggs, D. A. (1992). Cross-cultural exchange of expertise and training. In D. W. Chap- man and H. J. Walberg (eds.), International Perspectives on Educational Produc- tivity. Greenwich, CT: JAI Press.

Biggs, D. A., Roth, J. D., and Strong, S. R. (1970). Self-made academic predictions and academic performance. Measurement and Evaluation in Guidance 3(2): 81-85.

Biggs, D. A., and Tinsley, D. J. (1970). Student-made academic predictions. Journäl of Educational Research 63(5): 195-197.

Bochner, S. (1982). Cultures in Contact: Stndies in Cross-Cultural Interaction. Oxford: Pergamon.

Byrnes, F. C. (1966). Role shock: An occupational hazard of American technical assist- ants abroad. Annals of the American Academy of Political and Social Science 368: 95-108.

Chapman, D. W., Wan, T. Y., and Xu, M. (1988). Academic adjustment of interna- tional students in American universities. Education and Society 6(2): 96-103.

Church, A. T. (1982). Sojourner adjustment. Psychological Bulletin 91: 540-572. Cohen, J. (1988). Statistical Power Analysis for the Behavioral Sciences, 2nd ed. New

York: American Press. Cole, J., Allen, F., and Green, J. (1980). Survey of health problems of overseas stu-

dents. Social Science & Medicine 14: 627-631. Dodge, S. (1990). Culture shock and alienation remain problems for many foreign stu-

dents on U.S. campuses. Chronicle ofHigher Education 37(25): A33-36. Dohrenwend, B. P., and Dohrenwend, B. S. (1974). Stressful Life Events: TheirNature

and Effects. New York: Wiley. Draguns, J. G. (1980). Psychological disorders of clinical severity. In H. C. Triandis

and J. G. Draguns (eds.), Handbook ofCross-Cultural Psychopathology, vol. 6. Bos- ton: Allyn and Bacon.

Eldridge, J. E. T. (1960). Overseas students at Leicester University: Some problems of adjustment and communication. Race 2: 50-59.

Furnham, A., and Bochner, S. (1982). Social difficulty in a foreign culture: An empiri- cal analysis of culture shock. In S. Bochner (ed.), Cultures in Contact: Studies in Cross-Cultural Interaction. Oxford: Pergamon.

Heppner, P. P. (1986). Manual for the problem solving inventory. A paper presented at the Annual Meeting of the American Psychological Association, Washington, D.C.

Heppner, P. P., Hibel, T., Neal, G. W., Weinstein, C. L., and Rabinowitz, F. E. (1982). Personal problem solving: A descriptive study of individual differences. Jour- nal of Counseling Psychology 29: 580-590.

Higbee, H. (1969). Role shock--A new concept. International Educational and Cul- tural Exchange 4(4): 71-81.

Hilt, J. H. (1966). An analysis of a group of Indonesian, Thai, PakistaN, and Indian student perceptions of their problems while enrolled at Indiana University. Un- published Ed. D. thesis, Indiana University.

Hull, W. F. (1978). Foreign Students in the United Stares of America: Coping Behavior Within the Educational Environment. New York: Praeger.

Klineberg, O., and Hull, W. F. (1979). At a Foreign University: An International Study of Adaptation and Coping. New York: Praeger.

Lazarus, L. (1974). The psychology of coping: Issues of research and assessment. In G. V. Coelho, D. A. Hamburg, and J. E. Adams (eds.), Coping andAdaptation (pp. 249-315). New York: Basic Books.

622 WAN, CHAPMAN, AND BIGGS

Lazarus, R. S. (1976). Patterns ofAdjustment, 3rd ed. New York: McGraw-Hill. Lazarus, R. S. (1981). The stress and coping paradigm. In Eisdorfer, Cohen, Kleinman,

and Maxim (eds.), Theoretical Basesfor Psychopathology. New York: Spectrum. Lazarus, R. S., and Cohen, J. (1977). Environmental stress. In J. Altman and J. Wohl-

weil (eds.), Human Behavior and Environment: Advances in Theory and Research, vol. 2. New York: Plenum Press.

Lazarus, R. S., and Launier, R. (1978). Stress-related transactions between person and environment. In L. Pervin and M. Lewis (eds.), Perspectives in Interfactional Psy- chology. New York: Plenum.

Light, R., Xu, M., and Mossop, J. (1987). English proficiency and academic perform- ance of international students. TESOL Quarterly 21(2): 251-255.

Lin, L., and Ensel, W. M. (1984). Repression-mobility and its social ecology: The role of life events and social support. Journal of Health and Social Behavior 25: 176- 188.

Lodge, M. (1981). Magnitude Scaling. Beverly Hills, CA: Sage Publications. Melby, J. F., and Wolf, E. K. (1961). Looking Glass for Americans: A Study of the

Foreign Students at the University of Pennsylvania. Philadelphia: National Council on Asian Affairs.

Moleni, C. F. (1986). Adjustment problems of foreign students in U.S. colleges and universities. Clearinghouse on Language and Linguistics, ED276296.

Neal, G. W., and Heppner, P. P. (1982). Personality correlate of effective personal problem solving. Paper presented at the meeting of the American Personal and Guid- ance Association, Detroit.

Phillips, S. D., Pazienza, N. J., and Ferrin, H. H. (1984). Decision making styles and problem solving appraisal. Journal of Counseling Psychology 31: 497-502.

Rust, R. (1960). Epidemiology of mental health in college. Journal of Psychology 49: 235-248.

Selby, H., and Woods, C. (1966). Foreign students at a high pressure university. Soci- ology of Education 39: 138-154.

Sharma, S. (1973). A study to identify and analyze adjustment problems experienced by foreign non-European graduate students enrolled in selected universities in the state of North Carolina. California Journal of Educational Research 24:135-146.

Sorenson, J. (1988). Life Satisfaction Battery. Unpublished manuscript, University of California at Los Angeles.

Spaulding, S., and Flack, M. J. (eds.) (1976). The World's Students in the United States. New York: Praeger.

Stevens, S. (1975). Psychophysics. New York: Wiley. Still, R. (1961). Mental health in overseas students. Proceedings of the British Health

Association. Veroff, J. (1963). African students in the United States. Journal of Social Issues 19:

48-60. Wart, T. Y., and Xu, M. (1987). English language proficiency and Chinese graduate

students in U.S. graduate schools. Proceedings of the Beijing International Sympo- sium on Teaching English as a Second Language. Beijing: National Tourism Press.

Ward, L. (1967). Some observations of the underlying dynamics of conflict in a foreign student. Journal of the American College Health Association 10: 430-440.

White, A. J., Brown, S. E., and Suddick, D. (1983). Academic factors affecting the scholastic performance of international students. College Student Journal 17: 268- 272.

ACADEMIC STRESS 623

Xu, M., and Wan, T. Y. (1990). Differential language needs by international graduate students from different academic fields. Paper presented at EERA 1990 Annual Con- ference, Clearwater, Florida.

Received March 28, 1991.