Stress episode in aviation

job performance

CCM 17,1 62 Cross Cultural Management: An International Journal Vol. 17 No. 1, 2010 pp. 62-78 # Emerald Group Publishing Limited 1352-7606 DOI 10.1108/13527601011016916 Stress episode in aviation: the case of China Louise Tourigny Management Department, College of Business & Economics, University of Wisconsin-Whitewater, Whitewater, Wisconsin, USA Vishwanath V. Baba DeGroote School of Business, Hamilton, Canada, and Xiaoyun Wang I.H. Asper School of Business, University of Manitoba, Winnipeg, Canada Abstract Purpose – The purpose of this paper is to investigate the impact of role stressors on job stress among airline employees in mainland China. More specifically, the aggravating effects of shift work and the mitigating effects of decision latitude are explored to facilitate strategies of intervention aimed at reducing job stress. Design/methodology/approach – Data are collected using a field survey in Mandarin from 485 airline employees, including pilots, flight attendants, and service employees in five major cities in mainland China. Findings – The findings demonstrate that role overload and role conflict have significant positive effects on job stress. Furthermore, both shift work and its interference with non-work activities significantly elevated the impact of role overload on job stress. Findings also reveal that decision latitude mitigated the detrimental effect of role overload on job stress for employees working on fixed shift, but not for employees working on rotating shift. Research limitations/implications – This is a cross-sectional study using perceptual measures. Practical implications – The findings suggest that aviation managers in China need to focus not only on decision latitude but also on job and organizational design to mitigate the impact of job demands on stress. While decision latitude works to ease demands among those who work on fixed shifts, it does not work in the same way for those working on rotating shifts. Originality/value – This paper corroborates the cross-cultural applicability of stress theory by demonstrating the detrimental role of rotating shift on stress while at the same time calling attention to some cultural shaping of the findings. Keywords Shift work, Job design, Stress, Airlines, Employee behaviour, China Paper type Research paper 1. Introduction Feelings of stress have been universally experienced. According to the most recent Handbook of Work Stress, ‘‘job stress is an international problem’’ (Liu and Spector, 2005, p. 507). Annual cost to organizations and societies due to stress and its consequences run in the hundreds of billions of dollars worldwide. There has been an ongoing interest in the phenomenon for decades now and it has been studied by scholars, journalists, policy makers, and practicing professionals in every field. It is even a part of routine national population surveys (e.g. Canada and USA). Economists, sociologists, psychologists, epidemiologists, physicians, and social workers have all contributed both empirically and theoretically to the voluminous literature on stress. And not to put too fine a point on it, we know a lot already about the antecedents and consequences of stress in a general way. The current issue and full text archive of this journal is available at www.emeraldinsight.com/1352-7606.htm Financial support for this study from SSHRC (410-2003-1014) is gratefully acknowledged. Stress episode in aviation 63 However, what is not well known is how such universal phenomena appear when looked through cultural lenses. Much of the empirical literature that has shaped our thinking and theorizing in organizational behavior and management has come from North America and Europe (Tsui, 2006). Empirical research on stress and well-being is no exception (e.g. Bakker and Demerouti, 2007; Burke et al., 2006; Elovainio and Kivima ¨ ki, 1996; Johnson and Hall, 1988), although there is some recent work on stress that is located in other parts of the world (Liu and Spector, 2005; Spector et al., 2004). While stress theory, focusing on the individual experiences of stress, is quite portable across national and cultural frontiers, its empirical manifestations have been subject to variation based on culture, industry, occupation, work, gender, and age, to mention a few (Liu and Spector, 2005; Shouksmith and Taylor, 1997). Despite recent cross-cultural attention on stress, its empirical focus on China has been rather sparse (Siu et al., 2007) and practically non- existent when it comes to the aviation industry (Baba et al., 2009). We chose the airline industry because it is well-known that airline employees suffer from chronic fatigue associated with heavy pressure and shift work (Petrilli et al., 2006; Taneja, 2007). In addition, there are factors associated with flying which exacerbate the stress experience of aviation workers (Day et al., 2009; Hickman and Mehrer, 2001). Furthermore, we decided to locate this study in mainland China where there is an increasing demand for air travel and where the airline industry is currently undergoing a major expansion. It has been documented that airlines in China are planning to offer additional services, more flights, and respond to the needs of customers who expect high quality service in a competitive environment (Zhang, 1998). Such demands put additional pressures on employees to work longer hours and to take on additional responsibilities. As a consequence, this can increase the workload and time pressure. In addition, such changes can provoke more stress due to increases in potential role conflict that results from the necessity to work on different shifts with different groups. Rapid changes can also be associated with role ambiguity when employees take on additional responsibilities without receiving enough guidance on how to accomplish the work. Organizational growth represents challenges of recruitment and retention and on the flipside, turnover to the Chinese airlines. These challenges largely contribute to role overload. Therefore, we need to develop a better understanding of the factors that increase job stress and to see how job design can be modified to mitigate the stress and its repercussions. 2. Literature review and hypotheses Karasek’s job strain model (1979) proposes that psychological strain results from the joint effects of the demands related to the job context and the resources an individual has to meet those demands. More specifically, Karasek (1979) identifies job characteristics as the major determinants of psychological strain, which seems to derive from high job demands and low control over the resources necessary to meet those demands. When individuals have to meet demands on their job without adequate resources and without adequate latitude to make appropriate decisions, it makes it difficult to respond effectively. Such situations are known to adversely affect the psychological well-being of employees (Karasek, 1979). The Karasek model predicts that heavy workloads and time pressure represent key job demands that can increase job strain. However, the model does not explore whether role stressors actually contribute to higher job demands by taxing personal resources available to deal with the workloads and time pressure effectively. The literature reveals that clear goals and knowledge of work processes can reduce occupational CCM 17,1 64 stress, and the consequent job strain (Elovainio and Kivima ¨ ki, 1996). Furthermore, role clarity has been found to moderate the relationship between job demands and job strain (Bliese and Castro, 2000). However, the extent to which role ambiguity, role conflict, and role overload contribute to job stress under working conditions that are either exigent or facilitative has not been studied. Hence, we investigate the effects of role stressors, including role ambiguity, role conflict, and role overload, on job stress under different working conditions. Although the focus of this study is on the effects of role stressors on job stress, we explore the joint moderating effects of shift work and job control as exigent and facilitative factors respectively on the relationship between role stressors and job stress. We follow the interactive approach in testing the demand-control (DC) model rather than the additive approach, in light of the current stream of research on job demands and job control (van Vegchel et al., 2005; Totterdell et al., 2006). We want to explore the buffering effect of job control by examining its interactions with specific job demands pertaining to a particular work context. For this purpose, we focus on shift work as a contributing factor in the work context that can intensify the effects of existing job demands. We also focus on decision latitude as another contributing factor that can mitigate or buffer the detrimental effects of excessive job demands by providing greater decision autonomy and control at work. 2.1 Shift work In this study we concentrate on the negative effects of shift work. More specifically, we investigate how fixed or rotating shift have differential effects on job stress and on how employees who report high role stressors actually manage their responses based on whether they work on fixed or rotating shift. Shift work has been the dominant mode of work in communication, health care, transportation and manufacturing and has a pervasive impact on work outcomes (Blau and Lunz, 1999; Burke, 2003; Dunham, 1977; Shouksmith and Taylor, 1997). The literature on occupational mental health amply documents the detrimental effects of shift work (Totterdell, 2005). Poor adaptation to work schedules has been associated with mental and physical health problems (Jamal and Baba, 1992). Shift work has been found to cause sleep disturbances (Barak et al., 1995), which, in turn, can adversely affect moods and decision making that involves risks (Harrison and Horne, 2000; Leith and Baumeister, 1996). Indeed, sleep deprived individuals tend to bypass rational calculation and take greater risks (Harrison and Horne, 2000), and to display lower response speed on psychomotor vigilance task (Petrilli et al., 2006). Furthermore, it is known that early morning shifts are associated with psychosomatic illness, fatigue, and low alertness (Folkard et al., 1990). These conditions may contribute to the severity of operational errors associated with human factors in airlines (Petrie and Dawson, 1997; Schroeder et al., 2006). Airline employees generally work on eight-15 h shifts. However, flight delays and frequent changes in workloads may require them to exceed the length of shift on occasion. We also know that rotating shifts contribute to higher levels of job stress in comparison to fixed shifts (Jamal and Baba, 1992). Consequently, shift work can interfere with decision-making processes critical to safety in airlines. In general, shift work can seriously affect the well-being of employees (Baba et al., 1998). For that reason, reducing the detrimental impact of shift work on airline employees is important. The negative effect of role stressors such as role ambiguity, role conflict, and role overload on stress has been well documented in the literature (Baba et al., 1998). Our Stress episode in aviation 65 interest here is to probe the moderating effect of shift work on that relationship. Rotating shift has always been found to be disruptive to both physical and mental health (Baba and Jamal, 1991; Jamal and Baba, 1992; Totterdell, 2005). Typically, because of the disruptive nature of rotating shifts, employees experience less steady supervision and guidance. This can lead to heightened feelings of role ambiguity and role conflict. In addition, people in night shifts may feel overloaded when they have to step in and take care of many things on their own because of the diminution in managerial guidance as managers tend to put in time during the day and not in the evening or the night. Indeed, job performance tends to be worse on the night shift (Monk and Folkard, 1985). If the employee is already experiencing certain amount of role stressors, the disruptive addition of shift rotations is quite likely to intensify the impact of those role stressors further. The pattern of relationship between role stressors and stress will be more muted in the case of those who are on fixed shifts. Therefore, we hypothesize: H1. The relationship between role stressors and stress will be stronger for those who work on rotating shifts compared to those who work on fixed shifts. Shift work in general interferes with other non-work activities often leading to dissatisfaction with the schedule and the job. However, Barton (1994) proposes that some employees have less aversion to shift work and may actually choose to work on shifts. Therefore, it is likely that those employees will experience less interference or disruption in their lives due to shift work. These employees may actually feel less stress because they themselves have made the choice. Indeed, research among Chinese nurses reveals that when nurses report a match between their actual and preferred shift schedule they experience significantly less mental health problems (Baba and Wang, 2003). We suspect that it is the disruption to normal life that contributes to the impact of role stressors on stress. So, we suggest that the disruption to normal life associated with shift work is the actual aggravating factor. The presence of such disruptions will further exacerbate the relationship between role stressors and stress. On the other hand, employees who choose to work on shifts are likely to organize their work and family lives better by taking into account their work schedule of choice. Consequently, these employees may be less prone to feel the disruption while working on shifts. When employees do not experience such interference associated with shift work, they are less likely to feel stressed out. Therefore, we hypothesize: H2. The relationship between role stressors and stress will be stronger for those who report disruptions to their non-work life due to shift work compared to those who do not report significant disruptions. 2.2 Job control Job control refers to the extent to which an individual has discretion in making decisions pertaining to the work. Decision latitude is an important psychosocial factor in the workplace that can prevent job strain because it increases work predictability. Employees who have responsibilities on the job and yet who lack the opportunity to participate in important decisions affecting their jobs can experience higher job strain (Day et al., 2009). In other words, responsibilities should be matched with an appropriate level of decision latitude. The Karasek model (1979) predicts that employees are more likely to experience higher levels of motivation, job satisfaction, CCM 17,1 66 and mental health when jobs are high in both demands and decision latitude. Xie (1996) found evidence in support to Karasek’s model predictions among health care employees in the People’s Republic of China. In this study, we propose decision latitude as a buffer between role stressors and job stress. Therefore, we hypothesize: H3. The relationship between role stressors and stress will be stronger for those who report lower levels of latitude in decision making compared to those who report greater latitude in decision making. Furthermore, Xie (1996) found that positive outcomes were associated with high job demands and high decision latitude for employees with a perceived ability-job fit. This notion of fit applies to the job context as well. In other words, the employee- environment fit can determine whether high job demands and high decision latitude contribute to better mental health. Rotating shifts may indeed interfere with the sense of control one has over the context of work compared to fixed shifts. Baba and Jamal (1991) in their study of shift work have argued that less variability in the work context might release more degrees of freedom to experiment with the content of work. In other words, holding the context constant may indeed allow for the decision latitude to be put to better use, thus giving the incumbent a greater sense of control. In this study, we analyze whether such relationship is sustainable in the context of shift work. We propose that employees who work on rotating shifts are less likely to benefit from higher levels of decision latitude due to the unpredictability of their work context. Therefore, we hypothesize: H4. In the context of rotating shifts decision latitude will not have a buffering effect on the relationships between role stressors and job stress whereas in the context of fixed shift decision latitude will have a buffering effect on these relationships. The Baba and Jamal study (1991) argued for a compensatory relationship between the context of work and the content of work. In other words, people benefited from enriching work only when the context did not simultaneously compete for their attention. Consequently, employees who experience less disruption in their non-work life due to shift work are more likely to benefit from the enriching prospects of decision latitude. Under such conditions, decision latitude is likely to buffer the detrimental effect of role stressors. So, we suggest that when employees are not disturbed by their schedules, they are less likely to experience negative outcomes, as the schedule does not become one more thing to add to their stress episode. Under such circumstances, an increase in job demands that is matched with higher decision latitude may indeed be beneficial (Karasek, 1979). Therefore, we hypothesize: H5. Decision latitude will have a buffering effect on the relationships between role stressors and job stress only when the disruption in non-work activities due to shift work is low. 3. Methodology 3.1 Sample Data were collected in the People’s Republic of China from 485 employees working for a major national airline in Beijing, Shanghai, Urumqi, Nanjing, and Chengdu with a survey instrument, translated from English to Mandarin, and back translated to ensure accuracy of meaning. The sample consists of pilots, engineers, flight attendants, Stress episode in aviation 67 technicians, plane maintenance assistants, and other service and maintenance employees. There are 315 males and 157 females for a total of 472 respondents with information on gender. Respondents were between 19 and 60 years of age. The mean age is 32.75 with a standard deviation of 8.37. The demographics are presented in Table I. T-tests reveal that males tend to work on shifts more than females. Furthermore, males reported higher levels of role conflict, role overload, and job stress whereas females indicated higher shift work disruption. Males and females did not report any significant difference in role ambiguity and decision latitude. T-tests also reveal that participants who worked on shifts indicated much higher levels of role conflict, role overload, and job stress, and much lower levels of decision latitude. Finally, T-tests show that participants who work on rotating shifts report much higher levels of stress and much lower levels of decision latitude. Respondents who worked on rotating shifts also reported more disruption to their non-work life, although the relationship was not significant. 3.2 Measures 3.2.1 Role stressors. Role ambiguity, role conflict, and role overload were measured using the Beehr et al.’s (1976) instrument with a scale ranging from 1 (strongly agree) to 5 (strongly disagree). Items were reverse scored so that a high score indicates high role Table I. Demographics Age 457: valid responses Education: 472 valid responses Mean: 32.75 High school: 8.5% Standard deviation: 8.37 Professional training: 4% Mode: 30 College: 39% Median: 31 University: 46% Minimum: 19 Master: 1.5% Maximum: 60 Occupation: 283 valid responses Gender Pilots: 36.4% Female: 32.4% Flight attendants: 15.5% Male: 64.9% Transportation: 2.5% Missing values: 2.7% Engineering and maintenance: 15.9% Children: 356 valid responses Other service and administrative: 29.7% 0: 14.7% Tenure: 458 valid responses 1: 81.3% Mean: 8.53 years 2>: 4% Median: 6.25 years Dependents: 356 valid responses Mode: 2 years 0: 3.1% Standard deviation: 7.88 years 1: 20.5% Maximum number of years: 42.25 2: 23.0% Job status 3: 21.1% Permanent: 90.1% 4: 15.4% Temporary: 6.8% 5: 11.8% Missing values: 3.1% 6>: 5.1% Shift work Marital status: 476 valid responses Yes: 189 respondents Never married: 30.9% No: 259 respondents Married: 67.9% Missing values: 37 Divorced: 1.1% Fixed shifts: 59 respondents Separated: 0.2% Rotating shifts: 114 respondents Missing values: 16 CCM 17,1 68 stressors. Role ambiguity was measured with four items from the Beehr et al. (1976). The mean and standard deviation were 2.19 and 0.53, respectively. A sample item is: ‘‘I feel certain about how much authority I have in my job’’ (reverse scored). Role conflict was measured with five items. The mean and standard deviation were 2.96 and 0.65, respectively. A sample item is: ‘‘I often have to bend a rule or a policy in order to carry out an assignment.’’ Role overload was measured with three items pertaining to the performance standards and responsibilities associated with the job. The mean and standard deviation were 3.09 and 0.68, respectively. A sample item is: ‘‘It often seems I have too much work for one person to do’’. 3.2.2 Decision latitude. Decision latitude was measured with three items from the Beehr et al. (1976) instrument with a five-point scale ranging from 1 (strongly agree) to 5 (strongly disagree). Items were reverse scored so that a high score indicates high decision latitude. The mean and standard deviation were 3.02 and 1.09. A sample item is: ‘‘In my job, I have enough say over what happens’’. 3.2.3 Fixed or rotating shift work. Respondents were asked to indicate whether they did shift work. If their response was positive, they were then asked whether they worked on fixed or rotating shifts. A total of 114 respondents indicated working on rotating shifts whereas 59 worked on fixed shifts. Thus, more than 65 percent of the participants who worked on shifts did rotating shifts. The remaining 259 participants did not work on shifts, but rather worked day time only. A total of 37 participants did not indicate whether they worked on shift. Therefore, we had a total of 189 participants who worked on shifts. 3.2.4 Shift work disruption. Shift work disruption was measured with three items from Folkard (1998). These items assess the extent to which the shift work system interferes with other non-work related obligations and activities. We used a five-point scale ranging from 1 (none at all) to 5 (intolerable amount). The mean and standard deviation were 2.71 and 0.97, respectively. A sample item is: ‘‘How much does your shift work system interfere with your domestic life?’’ 3.2.5 Stress. Stress was measured with 13 items from Parker and Decotiis (1983) using a 5-point scale ranging from 1 (strongly agree) to 5 (strongly disagree). Items were reverse scored so that high scores indicate high levels of stress. The mean and standard deviation were 2.89 and 0.65, respectively. A sample item is: ‘‘I have often felt fidgety or nervous as a result of my job.’’ 3.2.6 Control variables. We included both age and gender as control variables. Dummy variables were used to assign a numerical value of 1 and 2 to female and male, respectively. We employed descriptive statistics, correlation, and hierarchical moderated multiple regression to analyze the data. 4. Results Table II presents the correlations for the study variables along with the reliability coefficients. Results indicate that age is positively related to gender and role overload. Gender is positively related to role conflict, role overload, and job stress. Role ambiguity is positively correlated with job stress, and negatively correlated with decision latitude. Role conflict is positively correlated with role overload and job stress, and negatively correlated with decision latitude. Role overload is positively correlated with shift work disruption, and job stress. Rotating shift is negatively correlated with decision latitude, and positively correlated with job stress. Shift work disruption is negatively correlated with decision latitude and positively correlated with job stress. Decision latitude is negatively correlated with job stress. Stress episode in aviation 69 Table II. Descriptive statistics and correlations Variable M SD123456789 1. Age 32.75 8.37 2. Gender a 1.67 0.47 0.13** 3. Role ambiguity 2.19 0.53 À0.08 À0.08 (0.54) 4. Role conflict 2.96 0.65 À0.02 0.16** 0.05 (0.72) 5. Role overload 3.09 0.68 0.12* 0.16** 0.03 0.42** (0.59) 6. Fixed or rotating shift b,c 1.66 0.48 À0.05 0.09 0.01 À0.02 0.13 7. Shift work disruption d 2.71 0.97 À0.11 À0.13 0.09 0.12 0.22** 0.14 (0.90) 8. Decision latitude 3.02 1.09 0.07 0.04 À0.20** À0.23** À0.07 À0.25** À0.19** (0.83) 9. Stress 2.89 0.65 0.03 0.10* 0.13** 0.40** 0.60** 0.21** 0.38** À0.16** (0.88) Notes: n ¼ 425-483. Values in parentheses correspond to Cronbach alpha. **Correlation is significant at the 0.01 level (two-tailed). *Correlation is significant at the 0.05 level (two-tailed); a Female is attributed the dummy variable 1 and male 2; b Fixed shift is attributed the dummy variable 1 and rotating shift 2; c n ¼ 159-173; d n ¼ 159-237 CCM 17,1 70 Results from the multiple regression analysis testing the effects of role stressors on job stress indicate a positive and significant effect for role ambiguity, role conflict, and role overload. These findings confirm that the positive relationships between role stressors and job stress reported in the empirical literature holds good in China’s aviation sector. Table III presents the results for the hierarchical moderated regression analysis where job stress is the criterion variable. The first model tested the interactive effect of role stressors and fixed or rotating shift. Only significant results are reported here. It is worth noting that among role stressors, role overload appears to be the most potent in the aviation sector with the largest mean value. Results show that both role overload and shift work had a significant positive effect on job stress. Furthermore, there is a significant interactive effect between role overload and shift work on job stress as presented in Figure 1. There is a steep positive relationship between role overload and job stress for respondents on rotating shifts whereas such a relationship is much more muted for respondents on fixed shifts. We did not find any significant interactive effect for role Table III. Hierarchical multiple regression analyses ÁR 2  Model 1: Moderating effect of fixed or rotating shift Step 1 0.00 Age À0.05 Gender À0.02 Step 2 0.45** Role overload 0.65** Fixed or rotating shift 0.13* Step 3 0.06** Role overload  fixed or rotating shift 1.49** Model 2: Moderating effect of shift work disruption Step 1 0.00 Age À0.00 Gender À0.04 Step 2 0.44** Role overload 0.56** Shift work disruption 0.26** Step 3 0.02* Role overload  shift work disruption 0.70* Model 3: Moderating effect of decision latitude Step 1 0.00 Age À0.05 Gender À0.02 Step 2 0.46** Role overload 0.64** Fixed or rotating shift 0.10 Decision latitude À0.11 Step 3 0.06** Role overload  fixed or rotating shift 1.39** Role overload  decision latitude À0.21 Decision latitude  fixed or rotating shift À0.02 Step 4 0.01* Role overload  fixed or rotating shift  decision latitude 2.72* Notes: *p-value < 0.05, **p-value < 0.01 Stress episode in aviation 71 ambiguity and role conflict. Thus, the results offer partial support for H1. Nevertheless, the evidence corroborates the exacerbating effect of rotating shifts on the relationship between role overload and job stress, which is in alignment with the predictions of the Karasek model. Model 2 tested the interactive effect of role stressors and shift disruption. Only significant results are reported here. Results reveal that both role overload and shift disruption have a direct positive effect on job stress. Furthermore, there is a significant interactive effect between role overload and shift disruption in predicting job stress as presented in Figure 2. When respondents report high shift disruption, there is a steep positive relationship between role overload and job stress whereas such relationship is less pronounced when respondents are not as affected by shift disruption. Therefore, H2 is partially supported. Model 3 tested the joint interactive effect of role stressors, shift work, and decision latitude. Only significant results are reported here. Results indicate that role overload is a significant predictor of job stress. As predicted, there was a two-way interaction between role overload and shift work in predicting job stress. However, we did not find a significant two-way interaction between role overload and decision latitude in predicting job stress. Therefore, H3 is not supported. Figure 2. Moderating effect of shift work disruption Figure 1. Moderating effect of fixed or rotating shift . collected using a field survey in Mandarin from 485 airline employees, including pilots, flight attendants, and service employees in five major cities in mainland. similarities in the socialization of airline employees are more salient in interpreting experiences of stress in China than cultural differences between China and