OCCUPATIONAL STRESS: AN ANALYSIS Dean Elmuti Eastern Illinois University Charleston, Illinois Yunus Kathawala Eastern Illinois University Charleston, Illinois Sudhir Chawla Angelo State University San Angelo, Texas Introduction Stress is a part and parcel of modem day life, but the stress in the workplace has become a major contributor to many diseases in the modem industrial society. Some of the problems resulting from undue stress are headaches, heart attacks, hypertension, pep- tic ulcers, weakening of the immune system, anxiety, lack of job involvement, job dis- satisfaction, and occupational burnout [13]. This relationship between stress and diseases has had a major impact on American business. It is estimated that it costs about $150 billion annually for health insurance and disability claims, lost productivity, and other expenses [12]. The term stress is difficult to derme as no universal defmition has been accepted among scientists and researchers [3]. In general, stress involves the interaction of environmental stimuli, the physiological and psychological reactions to those stimuli, and coping responses of the person. Stress can be broken down into two basic components: external and inter- nal stressors. External stressors can be separated into physical factors, such as low light- ing, poor ventilation, and taxing physical tasks; and psychological factors such as job or role demands. Internal stressors are primarily a person's attitudes and expectations through which stress is placed on the self [4]. Stress can originate from variou-; spheres of a person's life, for example, home, oommu- oity, job, and soci~conomic conditions of the society at large. Stressors from home and oommunity conditions are age old, but stressors from the workplace have increased dramati- cally due to the rapid technological and economic advancement. These are expressed in the form of many issues such as job satisfaction, job security, participation in decision making, role in organization, career development, male-female equality, etc. Due to changing state oompensation laws businesses have become interested in their employees' health because they may be held liable for a worker's ill health resulting from occupational stress [7]. Another change in the business world is that in recent years women have become a significant part of the labor force with over half of them employed outside the home at any given time. Two-thirds of married women in the 25-44 year-old age group work out- side the home. Even though it is illegal to discriminate against women in the workplace, 2 Journal ofBusiness Strategies Vol. 8, No.1 there are many obstacles to equality of the sexes. The most notable inequality is the earnings gap between males and females [1]. There is also an enormous growth in younger women entering jobs which were for- merly dominated by males. As a result female managers are subjected to a greater num- ber work-related pressures compared to their male counterparts [6]. In their study Davidson and Cooper found that women in junior and middle management experienced the highest overall occupational stress levels; followed by male supervisors; senior women manag- ers; male junior managers; female supervisors, and male middle managers [6]. However, senior male managers reported the lowest occupational stress levels. Other studies contradict these results. Tung found that female administrators experi- enced significantly less occupational stress than male administrators in an educational setting [15]. Senatra also reported no statistically significant differences between men and women in job-related stress in public accounting firms [11]. Since most of the past studies on occupational stress have used exclusively male samples, Beehr and Newman suggest that sex should be used as a moderator [3]. Among types of managers the group considered to be more vulnerable to occupational stress are mid-level management. Stress related illnesses are also more common in this group than among top-level executives [9]. People often make jokes about one, two, and three-ulcer jobs because it is assumed that when executives reach a top-level in their or- ganization, they are more prone to illnesses than their subordinates. However, the Bell Telephone system studies contradicted those assumptions. Top-level executives as a group were found to be less susceptible to heart attacks than other persons in the company. The heart disease rate of both the workers and the supervisors more than doubled that of the executive group [2]. On the basis of this literature review, it seems that marc research is warranted to in- vestigate male and female differences with respect to occupational stress. Furthermore, it would be interesting to study the effects of managerial rank with sex differences, as stress levels are found to be different among different managerial rank. The purpose of this study then is to investigate the relationship of sex and managerial rank with occupational stress. The independent variables in this study are sex and managerial rank. Managerial rank will be varied at three levels, (1) first-line supervisors, (2) mid-level managers, and (3) top-level executives. Occupational stress will be measured by the 14 scales of Occu- pational Stress Inventory (OSI) resulting in the three factors or dependent variables. Hypotheses The three main hypotheses arc presented below. 1. Therc will be a significant diffcrcnce betwecn groups of malc and female managers in regard to their mean scores on all three occupational strcss factors. 2. There will be significant differences among the three groups of managers (first-line supervisors, mid-level managers, and top-level executives) in regard to their mean scorcs on all three occupational stress factors. Spring 1991 Elmuti, Kathawala & Chawla: Occupational Stress 3 3. There will be a significant difference between sex and managerial rank in regard to the mean scores on all three occupational stress factors. Method The relationship of sex and managerial rank with occupational stress was examined using a questionnaire administered to ninety-six male and female managers, employed by a large multinational corporation in a mid-western state. The goal was to obtain six groups consisting of 16 male first-line supervisors, 16 female first-line supervisors, 16 male mid- level mangers, 16 female mid-level managers, 16 male top-level executives, and 16 fe- male top-level executives. The personnel director of the corporation was contacted to obtain permission to con- duct the study and obtain names of managers. Sixteen managers were randomly selected from each group specified by sex and managerial rank. Once all managers were select- ed each individual was contacted by phone to obtain their agreement to participate in the study. In addition, an informed consent form was used to obtain permission of all sub- jects for their volunteer participation in this study and to allow their results to be used in a group form. The subjects were assured of complete confidentiality in this informed consent form. When fourteen managers refused to participate in the study for different reasons, an additional fourteen names were randomly selected from those persons' respective sex by rank category in order to meet our goal of obtaining sixteen managers for each group. Then the packet containing questionnaires were sent to the subjects along with a pre- addressed, pre-paid postage envelope. Approximately one week later, the entire sample was sent a follow up. This procedure produced a total of ninety-six usable questionnaires. Measures Occupational Stress. Occupational stress was measured with Occupational Stress Inventory (OSl). This in- ventory includes 140 items, and is designed to measure different kinds of stresses people experience in their work. Responses are on 5-point scales anchored by most of the time and rarely or never. The OSI has been shown to have acceptable levels of reliability and validity across a variety of settings. Nonnative data for this 14-scale inventory are avail- able for 909 adult subjects in 130 different occupations employed primarily in technical professions and managerial positions. Alpha coefficients for the scales range from .71 to .94. Since OSI was a published and validated inventory, a total of 100 copies were pur- chased from Psychological Assessment Resources in Florida. (A list of items in the in- ventory is presented in Appendix A.) Analysis of Occupational Stress Inventory (OSI). The OSI was developed for two primary reasons: 1) to develop generic measures of occupational stressors that would apply across different occupational levels and environ- 4 Journal ofBusiness Strategies Vol. 8, No.1 ments; and 2) to provide measures for an integrated theoretical model linking sources of stress in the work environment, the psychological strains experienced by individuals as a result of work stressors, and the copying resources available to combat the effects of stressors and alleviate strains [10]. For the purpose of this study, the occupational stress was measured by three major fac- tors resulting in three dependent variables. Factor 1 was measured by a set of six scales which were collectively called the Occupational Roles Questionnaire (ORO). The ORO scales were: Role Overload (RO), Role Insufficiency (RI), Role Ambiguity (RA), Role Boundary (RB), Responsibility (R), and Physical Environment (PE). The second factor in our study was the Personal Strain Questionnaire (PSQ) which was a measure of the domain of psychological strain and composed of four scales: Vocational Strain (VS), Psychological Strain (pSy), Interpersonal Strain (IS), and Physical Strain (PHS). Coping resources was the third factor of the OSI. This factor was assessed by the four scales which made up the Personal Resources Questionnaire (PRQ). These four scales were: Recreation (RE), Self-Care (sq, Social Support (SS), and RationaVCognitive Coping (RC). Coefficients of internal consistency and median item correlations were also calculated by applying the Cronbach's alpha formula to provide further justification for utilizing Occupational Stress Inventory as a multidimensional construct. These were presented in Table 1 which indicated that a coefficient alpha for each index was above .70. Table 1 Pearson Correlation Coefficient for Indices of Occupational Stress Inventory with Coefficient Alpha on Diagonal Factor MatriX- Factorl Factor Occupational Roles - Factor 1 (.79) Personal Strain - Factor 2 .44 (.80) Personal Resources - Factor 3 .38 .39 Median Item Correlations Within .44 .61 Between .20 .18 (.77) .31 .19 • Coefficient alpha are indicated in the diagonal of the Factor Matrix. The question of how large the coefficient alpha should be for an item to be reliable is not consistent among researchers; however, some consider alpha at .70 to be a good cri- terion for adequate scale reliability [5]. In addition, to the scales descnbed above, basic demographic questionnaire including sex, managerial rank, age, and length of employment with the company were included in the packet Spring 1991 Elmuti, Kathawala & Chawla: Occupational Stress Results 5 Demographic characteristics of respondents are shown in Table 2. The average subject was 41 years old and had 10 years of administrative experience. The median hours worked per week per subject were 52. A majority (60%) of respondents were married with children, and the median percentage of total life stress attributed to work was 70%. Approximately 50% (or 48) of the respondents was female, and 50% (or 48) was male. Table 2 Demographic Characteristics of Respondents Percent % Commfot l. Sex Male 50 As planned Female 50 2. Age 20 - 29 22 Average age 30 - 39 40 41 years old 40 - 49 20 50 - 59 12 60 and over 06 3. Managerial Rank First-line Supervisor 34 One individual occupied Mid-line Manager 33 two positions at the Top-level Executive 33 time of survey 4. Service With The Company Average Service 1 - 5 years 30 10 years 6 - 9 years 23 10 - 14 years 18 15 - 19 years 10 20 - 24 years 08 25 and over 11 5. Median Hours Worked Per Week 40 - 45 15 Median Hours Worked 46 - 51 32 Per Week 52 - 57 40 52 58 and over 13 6 Journal ofBusiness Strategies Comparative Analysis of Stress Profiles Vol. B,No.1 To investigate the relationship of sex and managerial rank: with occupational stress factors, a 2 x 3 factorial design was used, (see Figure 1), A 2 x 3 factorial analysis of variance was also used to analyze the data, with an alpha level of .05 for all tests. Figure 1 Research Design Sex Male Female First-Line Supervisor Mid-level Managers Top-level Executives 2* 5 M Score =(23.62) M Score =(22.00) n= 16 n= 16 4 1 M Score = (22.18) M Score = (24.31) n= 16 n= 16 6 3 M Score =(21.06) M Score =(23.00) n= 16 n= 16 Average mean score Average mean score for for Male-Managers Female-Managers (22.29) (23.10) Avg. mean score for first- line supervisors (22.81) Mid-level managers (23.24) Top-level executives (22.03) *Numbers in parenthesis indicate the levels of stress experienced by the groups of managers Hypothesis one predicts that there will be a significant difference between groups of male managers and female managers in regard to their mean scores on the occupational stress inventory factors. To determine the levels of stress experienced by male managers on all three dimen- sions or factors of Occupational Stress Inventory were compared against those of their female counterparts (fable 3). Spring 1991 Elmuti, Kathawala & Chawla: Occupational Stress 7 Table 3 Means, Standard Deviations for the ORQ'l, PSQb, and PRQc Scales for Managerial Rank and Sex QlID rsQ. PRQ ORQ. PSQ &PRQ Group n M SD M SD M SD M SD Male First Line Super. 1 16 9.20 4.75 7.15 2.94 7.37 2.24 23.62 9.57 Mid Level Mgr. 2 16 9.37 3.25 5.62 2.30 7.18 2.98 22.18 8.15 Top Level Exec. 3 16 6.87 2.72 4.75 2.05 10.43 3.90 21.06 8.05 Total Male (1,2,3) 48 9.08 3.20 5.05 2.25 10.14 3.86 22.29 8.17 Female First Line Superv. 4 16 9.12 3.21 5.52 2.26 5.35 2.99 22.00 8.10 Mid Level Mgr. 5 16 11.62 4.84 7.37 2.99 5.31 2.20 24.3110.15 Top Level Exec. 6 16 10.00 3.80 6.75 2.73 6.25 2.47 23.00 9.45 Total Female (4,5,6) 48 10.25 3.90 6.54 2.59 6.31 2.49 23.10 9.50 • ORQ = Occupational Roles Questionnaire. The ORQ scales are: Role, Insufficiency, Ambiguity, Boundary, Responsibility and Physical Environment. " PSQ '" Personal Strain Questionnaire. The PSQ scales are: Vocational Strain, Psychological Strain, Interpersonal Strain and Physical Strain. /) PRQ ;:: Personal Resources Questionnaire. The PRQ scales are: Recreation, Self Care, Social Support and Rational/Cognitive Coping. Table 4 presents the results of one-way analyses of variance between the male and female subgroups on all three factors while controlling for effects such as managerial rank, number of years in present position, and age. Table 4 indicates that the levels of stress experienced by male mangers on all three factors were lower than that of their female counterparts. Mean differences for male and female managers on all three factors are found to be significant with more than 95 percent confidence. This confirmed hy- pothesis one which led to its acceptance. Hypothesis two predicts that there will be significant differences among the three groups of mar.agers (first-line supervisors, mid-level managers, and top-level executives) in re- gard to their mean scores on the occupational stress dimensions. Table 3 presents per- ceived Occupational Stress Inventory indices mean scores and standard deviations for male - fIrst supervisors, male - mid-level managers, male - top-level executives, female - first- line supervisors, female - mid-level managers, and female - top-level executives. Varia- tions can been seen for mean scores on all indices. The question is bow significant sta- tistically are these differences. 8 Journal ofBusiness Strategies Vol. 8, No. 1 Table 4 ANOVA: Male Versus Female Subgroups on aU Three Factors of Occupational Stress Inventory Factor 1 Factor 2 Factor 3 Occupational Personal Personal Catezories Role roRl Strain