Reliability and Validity of the Chinese Version of ... - Semantic Scholar

INTERNATIONAL JOURNAL OF BEHAVIORAL MEDICINE, 10(1), 15–30 Copyright © 2003, Lawrence Erlbaum Associates, Inc.

Reliability and Validity of the Chinese Version of the Job Content Questionnaire in Taiwanese Workers Yawen Cheng, Wei-Ming Luh, and Yue-Liang Guo

The reliability and validity of four selected scales—namely, job control, psychological demands, supervisor support, and coworker support—from the Chinese Version of the Job Content Questionnaire (C–JCL) were studied in 551 male and 648 female workers in Taiwan. Cronbach’s alpha coefficients for job control, supervisor support, and coworker support were all above .80. Whereas it was .55 for psychological demands, indicating insufficient internal consistency for this subscale. Participants responded to questionnaire items relatively consistently over a 3-month period. Exploratory factor analyses disclosed 4 empirical factors, which corresponded closely with theoretical constructs of the JCL. As predicted by the model, lowest levels of job satisfaction were found in workers in “iso-strain” category (i.e., high demands combined with low control and low social support). These findings indicated that the C–JCL is reliable and valid for assessing psychosocial work conditions among Taiwanese workers, although further improvement is needed for the psychological demands subscale.

Yawen Cheng, Department of Public Health, National Cheng Kung University, Wei-Ming Luh, Institute of Education, National Cheng Kung University, Yue-Liang Guo, Department of Environmental and Occupational Health, National Cheng Kung University. This study was supported by grants from the National Science Council (NSC 89-2320-B006-088 and NSC 89-2314-B006-155), Taiwan. We thank Professor R. Karasek for permitting the development of the Chinese version of the JCL and providing suggestions regarding the Chinese translation Correspondence concerning this article should be sent to Yawed Cheng, Department of Public Health, College of Medicine, National Cheng Kung University, 1 University Road, Tainan 701, Taiwan. E-mail: [email protected]

16

CHENG, LUH, GUO

Key words: job stress, reliability, validity, occupational health, Taiwan

Research conducted over the past two decades indicated that poor psychosocial work conditions adversely affect health (Karasek & Theorell, 1990; Kawakami & Haratani, 1999; Schnall, Landsbergis, & Baker, 1994). Among many instruments designed for the assessment of psychosocial work environment, the Job Content Questionnaire (JCL) based on Karasek’s demand–control model has been the most popular (Hurrell, Nelson, & Simmons, 1998). This model postulates that the combination of high psychological demands with low control at work causes psychological strain (Karasek, Baker, Marxer, Ahlbom, & Theorell, 1981; Karasek & Theorell, 1990). Most of the epidemiological studies utilizing the JCL have been broadly consistent with Karasek’s hypothesis, showing associations of high levels of job strain with increasing risks of various health problems (Cheng et al., 2000; Fenster, et al., 1995; Karasek, et al., 1988; Schnall et al., 1994; Schnall, Schwartz, Landsbergis, Warren, & Pickering, 1992). The model has also been extended to incorporate work-related social support, as studies have consistently shown greater effects of high strain job on workers with poor social support at work (Johnson & Hall, 1988; Johnson, Hall, & Theorell, 1989; Landsbergis, Schnall, Deitz, Friedman, & Pickering, 1992). Although the importance of job stress as a major occupational hazard has gained wide awareness in many industrialized countries, few studies have been conducted in less developed world (Karasek, et al., 1998; Kawakami & Haratani, 1999). Given the fact that the JCL was developed and tested mostly in Western populations, it remains questionable if it also applies to populations in different cultures. In Taiwan, several recent reports indicated growing problems of work-related stress (Cheng, Go, & Yeh, 2001; Luo et al., 1998; Yang, Ho, Fan, & Yang, 1996; Yang, Ho, Su, & Yang, 1997). However, most of these reports used ad-hoc scales of job stressors, which were either lacking in clear theoretical basis or not tested for validity and reliability. With permission from the JCL center, we translated into Chinese 22 items selected from the JCL, which were core items for assessment of job control, psychological demands, supervisor support, and coworker support. We investigated the psychometric properties of the four scales of the Chinese Version of the JCL (C–JCL) among these workers from four manufacturing factories in Taiwan. The internal consistency, test–retest reliability, construct validity, and criterion-related validity are all considered in the study. Several dialects are spoken in Taiwan, including Mandarin, Taiwanese Ho-Lo, Hakka, and aboriginal languages. We used Mandarin in this study, which is comprehensible to virtually all Taiwanese. Mandarin is also the most

CHINESE JOB CONTENT QUESTIONNAIRE

17

commonly used language in other Chinese-speaking populations, including people in the People’s Republic of China.

METHODS Study Participants Our study sites were offices and plants of four private factories (S, CH, T, and CM) located in southern Taiwan. Factory S manufactures and supplies active pharmaceutical ingredients; more than half of its workforce were professionals holding a master degree or above. Factory CH manufactures light tubes for office use, and Factory T auto lamps and their accessories. Factory CM manufactures and assembles liquid crystal displays, color filters, and other computer appliances; a large proportion of its employees were assembly-line workers working in fabrication rooms. A total of 1,584 workers (S = 165; CH = 127; T = 152; and CM = 1,140) participated in obligatory annual health examinations that took place at the work sites between June and September 2000. Self-administered questionnaires were distributed to employees of Factories S and CM a few days prior to their health exams, and a research assistant checked the returned questionnaires onsite during the exams to assure their completeness. For Factories CH and T, questionnaires were distributed at the time of the exams, and within a week a company staff collected the returned questionnaires. The return rates were 100%, 98%, 83%, and 51% in Factories S, CM, CH, and T, respectively. We further excluded 117 participants who had incomplete information on age, sex, or C–JCL items, 120 workers with work tenure of three months or less, and 28 part-time workers. As a result 1,199 workers (or 76% of 1,584) were available for this study. To assess the test–retest reliability of the C–JCL subscales, a sub-sample of workers was randomly selected from factories S, CM, and T. Participants were contacted by phone three months after their health exams, and, on their agreement, a second questionnaire was mailed to each of them. Of the 72 participants we contacted, 48 participants (67%) agreed to participate and returned by mail their second questionnaires.

Questionnaire The questionnaire included 22 items, which consist a minimum set of questions for assessment of four major JCL scales—job control, psychological demands, supervisor support, and co-worker support. The original 22 items questionnaire was translated into Chinese by the first author, and was translated back into Eng-

18

CHENG, LUH, GUO

lish by two bilinguals who were blind to the original English version. Karasek reviewed our back-translation and some corrections were made accordingly. The job control scale is the sum of two subscales: skill discretion, measured by six items, and decision authority, measured by three items. The psychological demands scale is measured by five items. The work-related social support scale is the sum of two subscales: support from supervisors and support from co-workers, both measured by four items. For each item, the response was recorded on a four-point Likert scale, ranging from 1 (strongly disagree) to 4 (strongly agree). For each scale, a sum of weighted item scores was calculated (see Appendix for calculation formulas). Information on education, marital status, job title, and work history was also obtained. All the possible job titles were listed in the questionnaire that was designed specifically for this survey. Participants were asked to mark their job titles in the questionnaire. Based on job titles, participants were further grouped into the following categories: Grade 1—administrators or managers; Grade 2—professionals (e.g., engineers); Grade 3—skilled, white-collar workers (e.g., industrial hygienists, executives); Grade 4—low-skilled, white-collar workers (e.g., secretaries, bookkeepers); Grade 5—skilled, blue-collar workers (e.g., technicians, mechanists); Grade 6—low-skilled, blue-collar workers (e.g., machine operators, assemblers, cleaners, guards). We also adopted five items from the JCL for assessment of job satisfaction. A sum of weighted item scores was calculated, with a range from 0 (completely unsatisfied) to 100 (completely satisfied). (See Appendix for calculation formula.)

Statistical Analysis Cronbach’s alpha coefficient was calculated for each C–JCL scale to assess internal consistency. Test–retest reliability of each scale was examined with Pearson correlations of the two scores obtained three months apart, and by plotting the differences of the two scores against their means. All of the 22 items were subjected to an exploratory factor analysis, using multiple squared correlation and Varimax rotation method. The effects of job control, job demands, and work-related social support on job satisfaction were assessed using multiple regression models. To aid interpretability, the scales of job control and job demands were divided into tertiles, and work-related social support was split into two groups on the median. By classifying the levels of control and demands into three groups rather than using linear terms in the regression models, we avoided the assumption of linear relationship between independent and outcome variable, and were able to examine the possible gradient effects. All analyses were conducted with the Statistical Analysis System (SAS) program (SAS Institute, 1989).


19

TABLE 1 Demographic Characteristics and Work Conditions of the Study Population (N = 1,199) Men (n = 551)

Women (n = 648)

Variable

Mean

SD

Range

Mean

SD

Range

Age (years) Work tenure (years) Work hours per week (hours)

31.7 2.0 53.0

5.5 2.3 11.1

21.0 ~ 61.0 0.3 ~ 23.0 15.0 ~ 90.0

25.5 1.7 45.9

5.1 2.3 7.9

17.0 ~ 46.0 0.33 ~ 13.0 24.0 ~ 78.0

N Work site Factory S Factory CH Factory T Factory CM Years of education ≤ 9 year (mid-school) 10 ~ 12 (high-school) 12 ~ 16 (college) 16 ~ 18 (master) >18 (doctorate) Data missing Marital status Single Married Widowed/ Separated/ Divorced Data missing Employment grade G1: Managers/administrators G2: Professionals G3: Skilled white-collar G4: Low-skilled white-collar G5: Skilled blue-collar G6: Low-skilled blue-collar Data missing

(%)

N

(%)

109 (19.8) 47 (8.5) 22 (4.0) 373 (67.7)

49 (7.6) 50 (7.8) 33 (5.1) 516 (79.6)

1 (0.2) 34 (6.2) 330 (59.9) 158 (28.7) 27 (4.9) 0 (0.0)

17 (2.6) 319 (49.2) 274 (42.3) 32 (4.9) 6 (0.9) 1 (0.2)

263 (47.7) 283 (51.4) 3 (0.6) 2 (0.4)

477 (73.6) 161 (24.9) 5 (0.8) 5 (0.8)

85 (15.4) 266 (48.3) 56 (10.2) 20 (3.6) 32 (5.8) 67 (12.2) 25 (4.5)

2 (0.3) 31 (4.8) 24 (3.7) 94 (14.5) 0 (0.0) 452 (69.8) 45 (6.9)

RESULTS Descriptive Statistics Table 1 summarizes the demographic characteristics of the study population. The average age of men and women were 32 and 26 years, respectively, and the mean duration of work tenure was about 2 years for both men and women. Men tended to work longer hours than women (53 vs. 46 hours per week). A great

TABLE 2 Mean Values, Standard Deviations, and Ranges of C–JCL Subscales by Sex (N = 1,199) Men (n = 551) JCL Subscale (# of items) Skill discretion (6) Decision authority (3) Job control (9) Demands (5) Supervisor support (4) Coworker support (4) Workplace support (8)

Women (n = 648)

Mean

SD

Range

Mean

SD

Range

32.3 32.6 64.9 33.5 11.2 12.1 23.4

4.7 5.9 9.2 4.4 2.2 1.6 3.1

14 ~ 46 12 ~ 48 30 ~ 94 24 ~ 48 4 ~ 16 4 ~ 16 9 ~ 32

28.8 28.9 57.7 33.6 10.9 12.3 23.2

5.1 6.3 10.3 4.6 2.2 1.7 3.2

14 ~ 48 12 ~ 48 26 ~ 90 21 ~ 48 4 ~ 16 4 ~ 16 8 ~ 32

TABLE 3 Mean Values, Standard Deviations, Correlations, and Cronbach’s α Coefficients of Items of C–JCL (N = 1,199)

Scale Item SD Q1. Learn new things Q2. Work not repetitive Q3. Requires creative Q5. High skill level Q7. Variety Q9. Develop own abilities DA Q4. Allow own decisions Q6. Freedom to make decision Q8. Opinions influential (lot of say) PD Q10. Work fast Q11. Work hard Q13. Excessive work Q14. Insufficient time Q16. Conflicting work SS Q20. Supervisor concerned Q21. Supervisor pays attention Q22. Supervisor helpful Q23. Supervisor good organizer CS Q24. Coworker competent Q25. Coworker interest in me Q26. Coworker friendly Q27. Coworker helpful

Mean

SD

3.23 1.91 2.53 2.56 2.53 2.44 2.54 2.60 2.51 2.95 3.26 2.62 2.44 2.40 2.75 2.76 2.71 2.84 2.99 3.08 3.12 3.02

0.58 0.61 0.75 0.68 0.68 0.72 0.71 0.66 0.67 0.63 0.52 0.70 0.66 0.65 0.69 0.65 0.67 0.61 0.51 0.47 0.48 0.52

Correlation With Total α i f Standardized α of the Items in a Item is Subscale Deleted Subscale 0.38 0.21 0.62 0.47 0.43 0.60 0.57 0.46 0.49 0.32 0.19 0.51 0.49 0.14 0.73 0.72 0.74 0.65 0.63 0.76 0.77 0.66

0.70 0.74 0.62 0.68 0.69 0.63 0.51 0.65 0.62 0.51 0.58 0.38 0.40 0.61 0.82 0.82 0.81 0.85 0.85 0.80 0.79 0.84

0.71

0.69

0.55

0.86

0.86

Note. Job control (9): Standardized Cronbach’s α = 0.80; Workplace social support (8): Standardized Cronbach’s α = 0.86.

20


21

majority of women (70%) were low-skilled, blue-collar workers, whereas only 12% of men were in this category. In contrast, 64% of men were professionals, managers, or administrators, whereas only 5% of women held such positions. Men also scored significantly higher on both skill discretion and decision authority than women (Table 2).

Internal consistency. The mean values, standard deviations, item-subtotal correlations and Cronbach’s alpha coefficients for all the 22 items of C–JCL subscales are presented in Table 3. Little differences were found between men and women, therefore the grouped results were presented. Cronbach’s alpha coefficients were acceptable for skill discretion (0.71), decision authority (0.69), job control (0.80), supervisor support (0.86), and coworker support (0.86). However, it was low (0.55) for the psychological demands subscale. Of this subscale, two items (i.e., Q11—work hard and Q16—conflicting work) had item-subtotal correlations lower than 0.2, indicating poor internal consistency of these two items with the others.

Test–retest reliability. Pearson’s correlation coefficients of the first and second tests were 0.73 (p < .01) for skill discretion, 0.64 (p < .01) for decision authority, 0.62 (p < .01) for psychological demands, 0.62 (p < .01) for coworker support, and 0.36 (p = .02) for supervisor support. For the subscales of skill discretion, decision authority, demands, and coworker support, the differences between the two tests appeared to distribute normally within an acceptable range (i.e., ±2 standard deviations), indicating sufficient test–retest reliability. For the subscale of supervisor social support, however, scores of the first and the second tests were less consistent. We also noted that most subjects reported lower scores on supervisor support in the second test than the first test, especially among subjects with lower average scores (data not shown).

Construct validity. Four components were found separately in male and female samples by principal component analysis with the criterion of Kaiser’s eigenvalue >1. Scree plot analysis also revealed a four-component solution (data not shown). The four factors extracted by the principle axis factoring method in the male sample corresponded very closely to the theoretical constructs (Table 4). After the Varimax rotation, the variance explained for the first three factors was evenly distributed (i.e., 12.7%, 11.6%, and 11.6%). However, the items Q1 “learn new things” and Q16 “conflicting work” did not have factor loadings greater than 0.3 on any of the factors in men.

22

TABLE 4 Factor Analysis of the 22 C–JCL Items Using Multiple Squared Correlation and Varimax Rotation Method (N = 1,199) Men (n = 551) Scale

Item

SD Q1. Learn new things Q2. Work not repetitive Q3. Requires creative Q5. High skill level Q7. Variety Q9. Develop own abilities DA Q4. Allow own decisions Q6. Freedom to make decision Q8. Opinions influential (lot of say) PD Q10. Work fast Q11. Work hard Q13. Excessive work Q14. Insufficient time Q16. Conflicting work SS Q20. Supervisor concerned Q21. Supervisor pays attention Q22. Supervisor helpful Q23. Supervisor good organizer CS Q24. Coworker competent Q25. Coworker interest in me Q26. Coworker friendly Q27. Coworker helpful Variation explained Variation explained (%)

F1

F2

F3

Women (n = 648) F4

F1

F2

F3

F4

0.41 0.33 0.66 0.51 0.41 0.65 0.49 0.41 0.64

–0.36 0.75 0.50 0.54 0.69 0.64 0.37 0.63 0.45 0.34 0.73 0.62

0.76 0.77 0.81 0.76

2.80 12.7

*Items with factor loading < 0.30 are not listed.

–0.35 –0.31 –0.34 0.74 0.67 0.74 0.59

0.34

2.56 11.6

0.69 0.82 0.85 0.66 2.55 11.6

1.46 6.6

2.98 13.6

0.63 0.85 0.88 0.71 2.61 11.9

2.59 11.8

0.63 0.57 0.51 0.39

1.53 7.0


23

TABLE 5 Average Scores of C–JCL Scales and Job Satisfaction Among Workers of Different Employment Grades and Gender (N = 1,199)

Men Job control Skill discretion Decision authority Psychological demands Workplace social support Supervisor support Coworker support Job satisfaction Women Job control Skill discretion Decision authority Psychological demands Workplace social support Supervisor support Coworker support Job satisfaction

Grade 6

Grade 5

Grade 4

Grade 3

Grade 2

Grade 1

(LS/B)

(S/B)

(LS/W)

(S/W)

(Prof)

(Adm)

(n = 67)

(n = 32)

(n = 20)

(n = 56)

(n = 266)

(n = 85)

60.0 29.9 30.1 33.3

61.4 32.2 29.2 31.5

66.8 32.8 34.0 33.3

60.5 31.2 29.5 33.3

65.3 32.5 32.8 33.6

80.0 35.2 36.8 34.4

23.4

22.9

24.0

23.4

23.4

23.4

11.1 12.3 57.1

10.6 12.3 68.8

11.7 12.3 65.8

11.2 12.2 66.2

11.2 12.1 65.3

11.5 11.9 69.7

(n = 452)

(n = 0)

(n = 94)

(n = 24)

(n = 31)

(n = 2)

56.3 28.3 27.9 33.7

— — — —

59.0 28.7 30.3 32.5

63.0 31.2 31.8 33.1

65.6 32.2 33.4 33.8

— — — —

23.2

—

22.9

22.8

23.1

—

10.9 12.4 62.3

— — —

11.0 11.9 61.4

10.6 12.1 66.7

11.0 12.1 61.1

— — —

F test

** ** **

**

** ** **

**p value < 0.01. (F test for difference of group means from overall mean.)

The factor pattern for women was more complex than men. The items for psychological demands appeared to be bipolar, with two items loading negatively on factor III but at the same time loading positively on factor IV. Moreover, the item Q2 “work not repetitive” loaded negatively on factor IV along with four items for psychological demands, whereas the item Q21 “supervisor pays attention” loaded on factor I along with items for job control scale.

C–JCL scores and job satisfaction by employment grade. As shown in Table 5, skill discretion and decision authority scores increased with employ-

24

CHENG, LUH, GUO TABLE 6 Multivariate-Adjusted Regression Coefficients (95% Confidence Intervals) of Job Satisfaction Score by Categories of Job Demands and Control Scales (N = 1,199)

Psychological Demands Job Control High

Intermediate

Low

Low

Intermediate

High

14.74**

10.70**

10.46**

(10.48, 19.00)

(6.70, 14.70)

(6.40, 14.53)

6.54**

6.57**

6.70**

(2.94, 10.15)

(2.79, 10.35)

(2.61, 10.80)

4.04*

2.61

0

(0.23, 7.84)

(–1.15, 6.38)

(Reference)

Note. Adjusted for age, sex, and work-related social support; regression coefficient (95% confidence interval) for work-related social support, high vs. low, was 11.56 (9.58, 13.54); adjusted model Rsquare = 20.35 % *p < 0.05. **p < 0.01.

ment grades. Within the same employment grade, men and women had similar scores on most of the C–JCL subscales, except that low-skilled women had lower job control than men in the same employment categories.

C–JCL scores in association with job satisfaction score. Cronbach’s alpha coefficient for job satisfaction was 0.80, and the mean job satisfaction score was 64.0. When regressed on C–JCL subscales in separate models after controlling for age and sex, higher levels of job control, lower levels of job demands, and higher levels of work-related social support were associated with higher job satisfaction scores. The size of the effects appeared to be greatest in the work-related social support dimension, following by job control and psychological demands (data not shown). Table 6 summarizes the combined effects of job control, and job demands on job satisfaction score after controlling for workplace social support, age, and sex. Workers in the highest tertile of job demands and the lowest tertile of job control (i.e., “high strain” job) were found to have the lowest levels of job satisfaction. DISCUSSION The psychometric properties of the four scales of the C–JCL were examined in this study. The mean scores of work-related social support in our study popula-


25

tion were comparable to the mean scores of nationally representative samples from the United States, Canada, and the Netherlands, and that of Japanese workers from two companies (Karasek et al., 1998; Kawakami et al., 1995). Our study participants scored slightly higher on psychological demands and substantially lower on job control scale, especially for women. The standard deviations of these scales were smaller in our population than most of the Western populations. Such differences were expected, because our study population encompassed very restricted categories of occupations with a great number of female assembly-line workers. Although the average work hours in Taiwan have been declining steadily over the past two decades, Taiwanese workers still work longer hours when compared to workers in Western countries. In 1998, for example, official statistics showed that the average work hours in manufacturing sectors were 198 hours per month (or 48 hours per week; ROC-DGOB, 1998). The means and range of work hours in our study were considered normal in Taiwan. Men reported longer average work hours than women in our study because there were a large proportion of young male professionals who tended to work overtime. Despite of longer work hours, the demands scale did not indicate substantial differences between men and women in our study, nor between Taiwanese populations and Western populations. Conceptually, work hours contribute greatly to job stress, but its importance did not seem to be captured by the five-item version of demands scale. The skill discretion and decision authority scores increased along with employment grade. Compared to men in the same grade, professional women and skilled, white-collar female workers had similar levels of job control, whereas women in low-skilled categories had substantially lower job control score than men. Cronbach’s alpha coefficients for job control, supervisor support, and coworker support were generally acceptable, but it was low for the demands subscale. The internal consistency for the demands subscale is the lowest (Cronbach’s α = 0.55) in our study when compared to other studies from the United States, Canada, the Netherlands, and Japan (Cronbach’s α = 0.59~0.71; Karasek et al., 1998; Kawakami et al., 1995). This indicated that some items in the Chinese version should be rephrased. Particularly, two items (Q11 and Q16) were found to poorly correlate with other items in the demands subscale. We suspected that the item Q11 “hard work” might appear to many subjects like an assessment for work value rather than for psychological demands in the Chinese culture. Item Q16 “conflicting work” contained negative terms in Chinese (I am “not asked” by “not the same” people …), which might have made it difficult to interpret. Nevertheless, lower internal consistency for the psychological demands scale had also been found in several Western and Japanese populations (Karasek et al., 1998; Kawakami et al., 1995), indicating that this five-item version might not be a sufficient measure for various types of psychological demands that workers encountered in the workplace.

26

CHENG, LUH, GUO

Overall, participants responded to questionnaire items relatively consistently over a 3-month period. During 3 months, work conditions might change, which could cause an underestimation of the true test–retest reliability. The true test–retest reliability could have been higher if noise from changes in work conditions over the 3-month period and from applying different methods in data collection could be controlled. Interestingly, we noted that participants tended to report lower scores in supervisor support in the second test, especially those with lower average scores in this subscale. We speculated that participants might have responded more conservatively in the first test with regard to their supervisors, because their first questionnaires were collected in a less confidential manner than in the second test. For men, the four extracted factors corresponded very closely to the theoretical constructs of the JCL. For women, the factor pattern is more complex. Some items belonging to the psychological demands scale (i.e., excessive work, insufficient time, and conflicting work) loaded negatively on the factor for supervisor support, indicating a link between these items and lack of supervisor support. The item Q21 “supervisor pays attention” loaded on factor I along with the items for job control scale, suggesting that supervisor’s attention might also contribute to the perception of job control in women. The item Q2 “work not repetitive” loaded negatively on the factor for psychological demands, suggesting that repetitive work task might contribute to work demands for women in this population. Jobs in the four companies we studied were highly segregated by sex. A majority of male workers (64%) were managers, engineers, or other professionals, whereas most of the female workers (70%) were low-skilled, blue-collar workers. Even in the same occupational categories, men and women might perform different activities. Gender differences in occupational grade and in work content might relate to the discrepancy in the factor pattern shown in this study. Due to an imbalanced distribution of occupations by sex, we were unable to disentangle specific effects of sex from occupational type on the factor pattern. As predicted by the demand-control-support model, lowest levels of job satisfaction were found in workers with jobs that were high in demands but low in control and work-related social support. The size of the effects was greatest in the social support dimension, suggesting that social relationship in the workplace accounted for a great variability of job satisfaction. Job control also accounted for a substantial amount of variability of job satisfaction, whereas the effects of job demands were less notable. Our findings of poor predicting power of the psychological demands scale were in consistent with other studies (Bosma et al., 1997; Marmot et al., 1997; Schnall et al., 1994). Future research will be needed to further improve the measure for this dimension. A major limitation of this study was that the sample was taken from restricted categories of occupations. It would be desirable to examine the validity and reli-


27

ability of the C–JCL scales in a national sample, or at least samples from a broader range of occupations and industries. Furthermore, other aspects of job stressors were not evaluated in this study. For example, a recent study in Taiwanese workers had found that female blue-collar workers ranked the problems of salary and benefits as their major source of job stress (Cheng et al., 2001). In conclusion, the findings of this study indicated that the C–JCL scales are reliable and valid for assessing psychosocial work conditions among Taiwanese working people, although further studies will be needed to improve the psychological demands scale. Societal differences should be considered if our findings are to be generalized to Chinese-speaking populations in other areas.

REFERENCES Bosma, H., Marmot, M. G., Hemingway, H., Nicholson, A. C., Brunner, E., Stansfeld, S. A., et al. (1997). Low job control and risk of coronary heart disease in Whitehall II (prospective cohort) study. British Medical Journal, 314, 558–565. Cheng, Y., Go, Y., & Yeh, W. (2001). A national survey of psychosocial job stressors and their implications for health among working people in Taiwan. International Arch Occupational Environmental Health, 74, 495–504. Cheng, Y., Kawachi, I., Coakley, E., Schwartz, J., & Colditz, G. (2000). Association between psychosocial work characteristics and health functioning in American women: Prospective study. British Medical Journal, 320, 1432–1436. Fenster, L., Schaefer, C., Mathur, A., Hiatt, R., Pieper, C., Hubbard, A., et al. (1995). Psychologic stress in the workplace and spontaneous abortion. American Journal of Epidemiology, 142, 1176–1183. Hurrell, J. J., Jr., Nelson, D. L., & Simmons, B. L. (1998). Measuring job stressors and strains: Where we have been, where we are, and where we need to go. Journal of Occupational Health Psychology, 3, 368–389. Johnson, J., & Hall, E. (1988). Job strain, work place social support, and cardiovascular disease: A cross-sectional study of a random sample of the Swedish working population. American Journal of Public Health, 78, 1336–1342. Johnson, J. V., Hall, E. M., & Theorell, T. (1989). Combined effects of job strain and social isolation on cardiovascular disease morbidity and mortality in a random sample of the Swedish male working population. Scandinavian Journal of Work, Environment & Health, 15(4), 271–279. Karasek, R., Baker, D., Marxer, F., Ahlbom, A., & Theorell, T. (1981). Job decision latitude, job demands, and cardiovascular disease: A prospective study of Swedish men. American Journal of Public Health, 71, 694–705. Karasek, R., Brisson, C., Kawakami, N., Houtman, I., Bongers, P., & Amick, B. (1998). The Job Content Questionnaire (JCL): An instrument for internationally comparative assessments of psychosocial job characteristics. Journal of Occupational Health Psychology, 3, 322–355. Karasek, R., & Theorell, T. (1990). The psychosocial work environment. In Healthy work–stress, productivity, and the reconstruction of working life (pp. 31–82). New York: Basic Books. Karasek, R. A., Theorell, T., Schwartz, J. E., Schnall, P. L., Pieper, C. F., & Michela, J. L. (1988). Job characteristics in relation to the prevalence of myocardial infarction in the US Health Examination Survey (HES) and the Health and Nutrition Examination Survey (HANES). American Journal of Public Health, 78, 910–918.

28

CHENG, LUH, GUO

Kawakami, N., & Haratani, T. (1999). Epidemiology of job stress and health in Japan: Review of current evidence and future direction. Industrial Health, 37(2), 174–186. Kawakami, N., Kobayashi, F., Araki, S., Haratani, T., & Furui, H. (1995). Assessment of job stress dimentions based on the job demands-control model of employees of telecommunication and electric power companies in Japan: Reliability and validity of the Japanese Version of the Job Content Questionnaire. International Journal of Behavioral Medicine, 2, 358–375. Landsbergis, P. A., Schnall, P. L., Deitz, D., Friedman, R., & Pickering, T. (1992). The patterning of psychological attributes and distress by “job strain“ and social support in a sample of working men. Journal of Behavioral Medicine, 15, 379–405. Luo, J. C., Hsu, K. H., Hsieh, L. L., Wong, C. J., & Chang, M. J. (1998). Lung function and general illness symptoms in a semiconductor manufacturing facility. Journal of Occupational & Environmental Medicine, 40, 895–900. Marmot, M. G., Bosma, H., Hemingway, H., Brunner, E., & Stansfeld, S. (1997). Contribution of job control and other risk factors to social variations in coronary heart disease incidence. Lancet, 350(9073), 235–239. Republic of China, Directorate - General of Budget (ROC-DGOB). (1998). Social indicators . Directorate-General of Budget, Accounting and Statistics, Executive Yuan, Republic of China. SAS Institute, Inc. (1989). SAS/STAT user’s guide, Version 6. Cary, NC: Author. Schnall, P. L., Landsbergis, P. A., & Baker, D. (1994). Job strain and cardiovascular disease. Annual Review of Public Health, 15, 381–411. Schnall, P. L., Schwartz, J. E., Landsbergis, P. A., Warren, K., & Pickering, T. G. (1992). Relation between job strain, alcohol, and ambulatory blood pressure. Hypertension, 19, 488–494. Yang, M. J., Ho, C. K., Fan, L. R., & Yang, M. S. (1996). Job strain and drinking behavior (Chinese). Kaohsiung Journal of Medical Science, 12, 670–684. Yang, M. J., Ho, C. K., Su, Y. C., & Yang, M. S. (1997). Job strain, social support, and mental health (Chinese). Kaohsiung Journal of Medical Science, 13, 332–341.


29

APPENDIX 1. Calculation formulas of JCL scale scores Skill discretion = (Q1 + Q3 + Q5 + Q7+Q9 + (5 – Q2) ) * 2 Decision authority = (Q04 + Q08 + (5 – Q06)) * 4 Job control = skill discretion + decision authority Psychological demands = 3 * (Q10 + Q11) + 2 * (15 – (Q13 + Q14 + Q16)) Supervisor support = Q20 + Q21 + Q22 + Q23 Coworker support = Q24 + Q25 + Q26 + Q27 Work-related social support = supervisor support + coworker support 2. Abbreviated forms of 5 items for job satisfaction scale and its calculation formula Q28. Satisfied with current job Q29. Recommend a friend Q30. Choose the same job again Q31. Not consider changing job Q32. Current job is as expected Job satisfaction = (1 – ((Q30 + Q32 – Q29 – Q31)*3 – (Q28*4) + 40)/60) * 100 3. Comparison of the original JCL items and the Chinese back translation Item

Original JCL

Chinese Back Translation

1.

My job requires that I learn new things.

My work requires me to learn new things.

2.

My job involves a lot of repetitive work.

3.

My job requires me to be creative.

There are a lot of repetitive tasks in my work. My work requires me to be creative.

4.

My job allows me to make a lot of decisions on my own.

My work allows me to make my own decision.

5.

My job requires a high level of skill.

My work requires high level of skills.

6.

On my job, I have very little freedom to decide how I do my work.

I don’t get to decide on how I would do my work.

7.

I get to do a variety of different things on my job.

8.

I have a lot of say about what happens on my job.

I can do a variety of different things on my job. On things that happen at work, my opinions are influential.

9.

I have an opportunity to develop my own special ability.

My work provides room for me to develop my own talent. (Continued)

30

CHENG, LUH, GUO APPENDIX (Continued)

Item

Original JCL

Chinese Back Translation

10.

My job requires working very fast.

My work requires me to do things very quickly.

11.

My job requires working very hard.

My work requires me to be very hard-working.

13.

I am not asked to do an excessive amount of work.

My workload is not considered excessive.

14.

I have enough time to get the job done.

I have enough time to accomplish my work.

16.

I am free of conflicting demands that others make.

On my job, I am not asked by different people to do things that are contradictory.

20.

My supervisor is concerned about the welfare of those under him.

My supervisor gives concerns to staff’s welfare.

21.

My supervisor pays attention to what I am saying.

My supervisor listens to my opinions.

22.

My supervisor is helpful in getting the job done.

My supervisor provides assistance to staff.

23.

My supervisor is successful in getting people to work together.

My supervisor organizes well the staff members as to facilitate tasks at work.

24.

People I work with are competent in doing their jobs.

My coworkers attend their duties well.

25.

People I work with take a personal interest in me.

My coworkers show their care for me.

26.

People I work with are friendly.

My coworkers are friendly.