Mediating effects of emotional exhaustion on the relationship between job demand-control model and mental health.

16  Download (0)

全文

(1)

Mediating Effects of Emotional Exhaustion on the

Relationship Between Job Demand–Control Model and

Mental Health

Yu-Hwa Huang1, Pey-lan Du2*, Chin-Hui Chen1, Chin-Ann Yang3 & Ing-Chung Huang4

1Institute of Human Resource Management, National Sun Yat-Sen University, Kaohsiung, Taiwan 2Department of Sport & Leisure, National Kinmen Institute of Technology, Kinmen, Taiwan 3Department of Marketing Management, Shu-Te University, Kaohsiung, Taiwan

4National Kaohsiung University, Kaohsiung, Taiwan

Abstract

This study attempted to investigate the role of emotional exhaustion as a mediator on the relationship between job demands–control (JDC) model and mental health. Three-wave data from 297 employees were collected. The results showed that job demands were positively related to emotional exhaustion, and increasing job demands will increase the level of emotional exhaustion. Job control was negatively associated with emotional exhaustion; therefore, increasing job control will decrease the level of emotional exhaustion. Emotional exhaustion was negatively related to mental health. Emotional exhaustion fully mediated the relationship between job demands and mental health, and partially mediated the positive relationship between job control and mental health. In addition, job control was positively associated with mental health directly. The remarkable fi nding of the present study was that emotional exhaustion served as the key mediator between the JDC model and mental health. Theoretical and managerial implications and limitations were discussed. Copyright © 2010 John Wiley & Sons, Ltd.

Received 26 March 2009; Revised 21 May 2010; Accepted 2 June 2010

Keywords

job demands–control model; emotional exhaustion; mental health

*Correspondence

Pey-lan Du, Department of Sport & Leisure, National Kinmen Institute of Technology, Taiwan.

Email: peylandu@gmail.com

Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/smi.1340

Introduction

The role of job demands and job control may enhance intrinsic motivation, but the initiatives may simultane-ously raise levels of job strain and other negative health-related outcomes among employees which will induce signifi cant costs in terms of sickness, lost work time and low productivity for organizations (Parker & Sprigg, 1999; Theorell & Karasek, 1996). Many studies have focused on the outcomes resulting from job stressors,

but few studies have addressed the processes between job stressors and strain. Whereas a focus on both pro-cesses and outcomes is a more sophisticated research strategy than a focus on outcomes alone (e.g. Hunter, 1990), the current study attempted to explore the process of job stressors and outcomes.

Karasek (1979) proposed a job demands–control (JDC) model as the two important characteristics of the work environment, and our study built on the JDC model to explore both positive and negative effects of

(2)

the work environment’s substantive contribution to employee’s emotional exhaustion and mental health. We attempted to provide a clear viewpoint by examin-ing emotional exhaustion as a mediator on the relation-ship between JDC model and mental health. Because the majority of studies have employed cross-sectional research designs on testing the JDC model, the results tend to infl ate correlations between the predictor and outcome variables. Also, the effects of occasional factors can be reduced by assessing the stressors and strain variables at different points in time (Jimmieson, 2000; Podsakoff, MacKenzie, Lee, & Podsakoff, 2003). There-fore, we decreased the bias by using a three-wave study. Specifi cally, our aim was to explore the process of job demands and job control on employees’ mental health. We posited and verifi ed a set of hypotheses based on the notion that emotional exhaustion may mediate the relationship between JDC model and mental health. In so doing, we developed the theoretical rationale that inclusion of emotional exhaustion helps clarify the rela-tionship between the JDC model and mental health. Finally, we provided implications for theory and mana-gerial practice resulting from our fi ndings.

JDC model

The JDC model as proposed by Karasek (1979) has stimulated the greatest amount of research and is gen-erally acknowledged as the dominant theoretical per-spective in the occupational stress area (Guglielmi & Tatrow, 1998; Parker, Jimmieson, & Amiot, 2010; Taris & Feij, 2004). It is a situation-centred model on which much of the current job stress research is based.

In its basic form, the JDC model postulates that the primary sources of job strain lie within two basic char-acteristics of the job itself: (1) ‘psychological job demands’; and (2) ‘job decision latitude’ or ‘job control’. Recent theoretical reviews on the JDC model suggested these work characteristics are still relevant to today’s globalized and post-tayloristic workforce (Hvid, Lund, & Pejtersen, 2008; Johnson, 2008).

Job demands are defi ned as psychological stressors, such as requirements for working fast and hard, having a great deal to do, not having enough time and having confl icting demands (Karasek, 1979). It is important to note that these are psychological demands and not physical ones (Fox, Dwyer, & Ganster, 1993). Thus, a fast and hectic work pace may impose physical require-ments that lead to fatigue, but the stress-related

out-comes the model predicts are related to the psychological effects of the work load—the anxiety associated with the need to maintain the pace of work and the associ-ated consequences of failing to complete the work.

Job control has been described as the worker’s ability to control his own activities and skill usage (Karasek & Theorell, 1990). Fox et al. (1993) pointed out that there are two components included in job control: a worker’s authority to make decisions on the job, or decision authority, and the variety of skills the worker uses on the job, or skill discretion. Previous researchers (e.g. Karasek, 1979) have combined these two components into one measure of job control.

Job strain is defi ned as symptoms of mental strain resulting from work context that result in poor psycho-logical and physiopsycho-logical well-being (Snyder, Krauss, Chen, Finlinson, & Huang, 2008). Empirically, the JDC model has predicted job strain such as depression, sleeping problems, exhaustion, consumption of medi-cation and dissatisfaction (Wallace, 2005), and serious health consequences (Belkic, Landsbergis, Schnall, & Baker, 2004; Kivimäki et al., 2002, 2006; de Lange, Taris, Kompier, Houtman, & Bongers, 2003).

High job demands precipitate job strain, as does low job control (Salanova, Peiró, & Schaufeli, 2002). Regarding the relationship between demands and control on one hand and well-being on the other, an important distinction can be drawn between two hypotheses: the strain hypothesis and the buffer hypoth-esis (Häusser, Mojzisch, Niesel, & Schulz-Hardt, 2010). According to the strain hypothesis of the JDC model, employees working in a high-strain job (high demands– low control) experience the lowest well-being. The buffer hypothesis states that control can moderate the negative effects of high demands on well-being (Van der Doef & Maes, 1999). The two hypotheses are not mutu-ally exclusive (Häusser et al., 2010). Genermutu-ally speaking, the strain hypotheses yielded more consistent support than the buffer hypotheses of the JDC model. Neverthe-less, sometimes it is not clear whether the negative effects could be exclusively attributable to either high demands or lack of control (Alfredsson, Spetz, & Theorell, 1985; Hammar, Alfredsson, & Theorell, 1994). The JDC model assumes that job strains are produced by job stressors: high job demands and low job control. The stressors have their greatest negative impact when job control is low and job demands are high, whereas an increase in job control serves to attenuate the nega-tive effects of job demands on strain.

(3)

Van der Doef and Maes (1999) analysed 63 studies which covered the period from 1979 to 1997, and revealed a greater proportion in supporting the main effect rather than the interactive effect. In addition, Karasek (1979) suggested that the interactive effect is not the primary issue, and that an additive assumption is equally valid. This conclusion is consistent with the fi ndings reported in a review of longitudinal research focusing on the model (de Lange et al., 2003). Specifi -cally, de Lange et al. (2003) found that 63 per cent of studies reported signifi cant main effects of job demand for physical and psychological strain outcomes, whereas 47 per cent reported main effects for job control. Fur-thermore, Häusser et al. (2010) extended the review from 83 studies published between 1998 and 2007, and found the evidence for interactive effects as predicted by the buffer hypotheses of the JDC/JDCS model was very weak overall. Therefore, the main effect of JDC model will be implemented in this research for the examination of the mediating role of emotional exhaus-tion on the relaexhaus-tionships among job demands, job control and mental health.

Emotional exhaustion: the core dimension of burnout

There are many studies examining the antecedents and consequences of burnout, showing that symptoms are associated with high job demands and low levels of job control (cf. de Lange, Taris, Kompier, Houtman, & Bongers, 2002; Taris, Schreurs, & Van Iersel-van Silfhout, 2001; Van der Doef & Maes, 1999). Currently, more and more researchers conceive burnout as a job-induced strain that can emerge in any occupation including management and technology (Leiter & Schaufeli, 1996; Maslach & Schaufeli, 1993; Schaufeli & Bakker, 2004).

Burnout is a stress-related disorder that is currently receiving much interest from behavioural researchers and health specialists (van der Linden, Keijsers, Eling, & van Schaijk, 2005). ‘Burnout’ is a metaphor that is commonly used to describe a state of mental weariness. It refers to a set of symptoms that an individual may develop during prolonged exposure to high levels of work stress, and negatively affects psychological and physical wellness, and perceived performance (Maslach & Schaufeli, 2000; Schaufeli & Enzmann, 1998).

The Maslach burnout inventory contains three dimensions: exhaustion, cynicism and low professional

effi cacy (Maslach, Schaufeli, & Leiter, 2001; Schaufeli, Leiter, Maslach, & Jackson, 1996). The fi rst dimension, emotional exhaustion, refers to the feelings of being emotionally overextended and drained. The second dimension, cynicism, refl ects indifference or a distant attitude towards work in general and the people whom one is working with, losing one’s interest in work and feeling that work has lost its meaning. Finally, low professional effi cacy refl ects a tendency to evaluate oneself negatively. Although Maslach’s (1982, 1993) original three-component conceptualization of burn-out has dominated the burnburn-out research (particularly as it relates to measurement of burnout), a number of researchers have called into question the three-component conceptualization of burnout (e.g. Moore, 2000; Shirom, 2003).

The alternative conceptualizations of burnout differ in a variety of ways; however, all of them include emo-tional exhaustion as a primary component of burnout, suggesting that it is indeed central to the experience of burnout (Halbesleben & Bowler, 2007). Moreover, it is common in the burnout literature to fi nd inconsistent relationships between components of burnout and antecedent or consequent processes (Halbesleben & Bowler, 2007). Empirically, some work has suggested that emotional exhaustion exhibits somewhat stronger relationships than do the other components to impor-tant outcome variables (Lee & Ashforth, 1993, 1996; Wright & Bonett, 1997). Hence, emotional exhaustion appears to be the most consistent in its relationships with outcome variables (Cordes & Dougherty, 1993; Demerouti, Bakker, de Jonge, Janssen, & Schaufeli, 2001; Green, Walkey, & Taylor, 1991; Lee & Ashforth, 1996). Conceptually, scholars argued that emotional exhaustion best captures the ‘core meaning’ of burnout (Cox, Kuk, & Leiter, 1993; Maslach, 1993). In keeping with these statements, we followed the suggestions of Halbesleben and Bowler (2007) and Janssen, Peeters, de Jong, Houkes, and Tummers (2004) that emotional exhaustion is the most appropriate dimension to include as a single burnout dimension. Therefore, in this study, we focused on the emotional exhaustion component of burnout.

Emotional exhaustion and JDC model The fi rst major prediction of this study is that job strain occurs when jobs are high in job demands or low in job control. This prediction rests on the reasoning that high

(4)

job demands produce a state of arousal in a worker that would normally be refl ected in such responses as ele-vated heart rate or adrenaline excretion. When job demands are high or job control is limited, emotional exhaustion develops irrespective of the type of job or occupation (Demerouti et al., 2001). Previous studies in several organizations have confi rmed this hypothesis by showing that badly designed jobs or high job demands exhaust employees’ mental and physical resources, leading to the depletion of energy (i.e. a state of exhaustion) and to health problems (health impairment process) (Bakker, Demerouti, De Boer, & Schaufeli, 2003a; Bakker, Demerouti, & Schaufeli, 2003b; Bakker, Demerouti, & Verbeke, 2004; Demerouti et al., 2001).

There is now a substantial body of evidence showing that higher job demands are associated with greater strain and that greater job control is associated with less strain (Holman & Wall, 2002). High level of job control means that workers have enough opportunity to deal with their tasks. When there is a constraint on the responses of the worker, as would occur when job control is low, arousal cannot be appropriately chan-nelled into a coping response and thus produces an even larger psychological reaction that persists for a longer time (Fox et al., 1993). Therefore, the potential of higher job control may reduce the level of emotional exhaustion. Thus, we proposed:

• Hypothesis 1: Job demands are positively related to emotional exhaustion.

• Hypothesis 2: Job control is negatively related to emotional exhaustion.

Emotional exhaustion and mental health The defi nition of emotional exhaustion lies in the pro-jected mental states when facing work stressors. The notion of ‘mental health’ refers to behaviour, attitudes and feelings that represent an individual’s level of per-sonal effectiveness, success or satisfaction. It has no necessary connection with mental illness in a clinical sense (Banks et al., 1980). The extent of what mental health measures, not only refl ects one’s reaction to his work environment, but also expands to the wider scope as to the individual’s integrated mental states.

Warr (1987) posited the process of mental health into two distinctions: ‘context specifi c’ and ‘context free’. Context specifi c refers explicitly to the mental

health which is job related (i.e. those indices which refl ect affective well-being). Emotional exhaustion can be regarded as ‘context-specifi c’ mental health. In trast, context-free mental health is a more global con-struct which is not tied to a particular setting or context. The extent of what mental health measures can be assumed to be ‘context-free’ mental health. Therefore, we can presume that the consequent effects of emo-tional exhaustion can extend to the individual’s mental state as a whole, and has impact on the individual’s mental health as well.

Emotional exhaustion as a mediator of JDC model and mental health

The JDC model has been used to predict standard occu-pational stress criteria (e.g. emotional exhaustion: Rafferty, Friend, & Landsbergis, 2001), as well as both work-related criteria (e.g. job dissatisfaction: Rodriguez, Bravo, Peiro, & Schaufeli, 2001) and criteria beyond the work context (e.g. cardiovascular disease: Kristensen, 1996). However, studies exploring the process of occupational stress and the relationship between psychological job characteristics, work-related well-being and context-free mental health are limited (e.g. Hakanen, Bakker, & Schaufei, 2006; Kelloway & Barling, 1991; Schaufeli & Bakker, 2004). Moreover, most of these studies employed the cross-sectional design which may pose a potential threat to method-ological rigor or to statistical conclusion validity (Cook & Campbell, 1979).

Based on Warr’s conceptualization, Kelloway and Barling (1991) proposed a model of job-related mental health in which the job characteristics (e.g. autonomy, task variety, etc.) and role stressors give rise to job-related affective well-being (e.g. emotional exhaustion, job satisfaction, etc.). In turn, these context-specifi c reactions predict context-free mental health. In other words, job-related mental health could be the me cha-nism through which the subjective experience of employment affects context-free mental health (Kelloway & Barling, 1991).

In the research of Hakanen et al. (2006) and Schaufeli and Bakker (2004), job demands and job control evoking the energetic process of high demands exhaust employees’ mental and physical resources, and may therefore lead to burnout, and eventually ill health. The effort-driven energetic process in which burnout plays a key role might lead to negative health outcomes

(5)

(Schaufeli & Bakker, 2004); consequently, some schol-ars also argue that job stress leads to the feeling of burnout and then burnout induces negative physical and psychological impacts towards oneself (Kelloway & Barling, 1991; Tang, Au, Schwarzer, & Schmitz, 2001). We therefore predicted:

• Hypothesis 3: Emotional exhaustion mediates the negative relationship between job demands and mental health.

• Hypothesis 4: Emotional exhaustion mediates the positive relationship between job control and mental health.

Methods

Research setting and sample

In order to minimize organizational variation, we focused on a single organization. We tested our hypoth-eses using data collected from employees of Customs Offi ces with four branch bureaus in different areas of Taiwan. In specifi c terms, this organization takes charge of collection of customs duties, smuggling prevention, bonding and duty drawback, trade statistics, building and management for aids to navigation and entrusted taxes and fees collection, as well as enforcement of gov-ernment control.

The data were collected as a three-wave study in order to diminish the effects of common method vari-ance. The fi rst-wave questionnaire included measures of job demands, job control and several demographic variables. Six hundred fi fty-three employees were selected by systematic sampling. All questionnaires were sent through internal organizational mail systems. Among them, 513 questionnaires were returned as valid. Two months later, at time 2, we delivered the second questionnaire including the scale of emotional exhaustion. Four hundred questionnaires were valid for the second-wave data collection. Next, the third-wave data collection was completed 2 months later, at time 3, which measured employees’ mental health. Finally, 297 questionnaires were declared valid and were used in the present study.

Comparison of the demographic characteristics in the sample and the general population revealed no dif-ferences in age, gender, educational level and tenure; therefore, our participants could represent the Customs Offi ces. Moreover, researchers also compared the research variables (job demands and job control)

between the two groups: one whose members attended both surveys of time 1 and time 2, the other whose members only attended the survey of time 1. The results showed no signifi cant difference between these two groups. Therefore, we can assume that those who dropped out at time 2 are not related to our indepen-dent variables.

Most of the respondents were male (68.7 per cent). The mean age of the respondents was 45 years (SD = 7.81). The majority of the participants (53.9 per cent) had completed a university degree. The average annual income was NT$839,000 (approximately US$26,000, SD = NT$174,000). The average tenure in the organiza-tion was 18.7 years (SD = 8.26).

Measures

The questionnaires used in this study were originally developed in English and then translated into the sub-jects’ native language of Mandarin Chinese. We estab-lished the linguistic equivalence of all the measures used in this study through the use of the back transla-tion procedure. The procedure of back translatransla-tion was to discuss each item and the alternatives in a small group of persons fl uent in both languages. Discussion occurred until agreement was reached as to the linguis-tic equivalence of the items in both languages. (cf. Feij, Whitely, Peiró, & Taris, 1995; Taris & Feij, 2004). Exogenous variables

The measures of job demands, job control that we used were developed and validated by Van Veldhoven (1996) in his dissertation research and Van Yperen and Hagedoorn (2003). A four-point response scale from 1 (never) to 4 (very often) followed each item in the scales measuring job demands and job control.

Job demands

The 11 items of the measure of quantitative job demands referred to the degree to which an employee had to work fast and hard, had a great deal to do and had too little time. For example, the participants were asked ‘Do you have to work fast?’, ‘Do you work under time pressure?’ High scores indicate high levels of per-ceived job demands.

Job control

This focused measure of job control consisted of 11 items as well, including items referring to timing

(6)

control and method control. For example, the partici-pants were asked, ‘Can you choose the methods to use in carrying out your work?’, and ‘Do you decide on the order in which you do things?’ High scores indicate high levels of perceived job control.

Endogenous variables

We measured mental health by using the General Health Questionnaire (GHQ-12) (Goldberg, 1972; Goldberg & Williams, 1988). The GHQ is a broad extent measure of psychological well-being comprising a symptoms checklist which is useful in detecting sub-clinical disturbances of individual mental health (Kelloway & Barling, 1991). The 12-item form of the GHQ has been demonstrated to be a sensitive instru-ment within an organizational context (Banks et al., 1980). Items consist of a question asking whether the respondent has recently experienced a particular symptom or item of behaviour rated on a four-point scale (Banks et al. 1980). Some sample items are ‘Have you recently been able to concentrate on whatever you’re doing?, lost much sleep over worry?, been able to enjoy your normal day-to-day activities?’ Higher scores on the GHQ indicate higher level of context-free mental health. This scale was measured at the third wave of the study.

Our mediating variable, emotional exhaustion, was measured by the emotional exhaustion subscale of the MBI-General Survey (MBI-GS; Schaufeli et al., 1996), which is used for non-human service workers and different from the MBI-Human Service Survey (MBI-HSS). The Maslach burnout inventory and its cross-language derivates are the instruments most often used in burnout research. The measurement was based on a seven-point scale ranging from 0 (never) to 6 (every day). High scores indicate high levels of exhaus-tion. The exhaustion subscale includes fi ve items. Exemplary items are: ‘I feel used up at the end of the workday’, and ‘I feel burned out from work’. This scale was measured at the second wave of the study.

Control variables

To isolate the effects of burnout on individual char-acteristic heterogeneity, we controlled for each respon-dents’ gender, age, annual income and tenure suggested by researchers (Bourbonnais, Comeau, & Vézina, 1999; Payne, Wall, Borrill, & Carter, 1999; Xie, 1996). These variables may in principle lead to spurious results if

they were not included in the analyses. To estimate the hypothesized relationships, we followed the approach of two-stage modelling (cf. Arnold, Turner, & Barling, 2007). Each model was estimated twice: once with no statistical controls and once controlling for gender, age, annual income and company seniority. When used, all control variables were estimated as single-indicator latent variables without measurement error, and allowed to covary with the exogenous variables and predict all endogenous variables in the model. The addition of control variables did not change our fi nd-ings, so only the uncontrolled model results are reported in detail.

Results

Goodness-of-fi t of variables

Structural equation modelling with analysis of moment structures (AMOS) was used to examine the applicabil-ity of the hypothesized structural model. As a prelimi-nary step in the analyses, we conducted the confi rmatory factor analysis to examine each scale. In order to main-tain the principle of theoretical consistency and parsi-mony, the items were eliminated according to the modifi cation index and the value of χ2

provided by AMOS to build a valid construct (cf. Byrne, 2001; Jöreskog, 1993).

Several goodness-of-fi t indices were used to assess the model fi t of each variable (cf. Jöreskog, 1993; Jöreskog & Sörbom, 1986): goodness-of-fi t index (GFI), adjusted GFI (AGFI), root mean square error of approximation (RMSEA) and standardized root mean square residual (SRMR). For the indices of GFI and AGFI, values greater than 0.90 indicate an acceptable fi t; values of the RMSEA smaller than 0.06 indicate an acceptable fi t; for the SRMR, values of 0.05 or smaller are acceptable. Because χ2

is sensitive to sample size, meaning that the probability of rejecting a hypothesized model increases with sample size, χ2

/df and two relative goodness-of-fi t indices were included (Hu & Bentler, 1999): the Tucker–Lewis index (TLI) and comparative fi t index (CFI). For these two fi t indices, values of 0.95 or above indicate a model with acceptable fi t. Finally, construct reliabilities were calculated as well (Hair, Anderson, Tatham, & Black, 1998). For the construct reliability, values of 0.5 above are acceptable (Raine-Eudy, 2000). This step is necessary to ensure that the latent variables are adequately representing the observed variable. As can be seen from Table I, the results show

(7)

all the scales have good model fi t and meet the requirements.

Means, standard deviations, internal consistency estimates (Cronbach’s α) and correlations of all study variables are presented in Table II. As can be seen from Table II, Cronbach’s alpha coeffi cients of all scales are considered to be acceptable, if they are larger than 0.70 (Nunnally & Bernstein, 1994). Our results show that all of the absolute values of the correlations are within the moderate level (0.13 to 0.45). The pattern of correla-tions was as expected.

Goodness-of-fi t of hypothesized model and alternatives

In order to test the hypothesized structural model, several competing structure models were employed: the baseline model (M1), the model which assumed zero relationships between variables; the structural model with direct effects of job demands and job control on emotional exhaustion and mental health (M2); and

the M3, which is the structural model of M2 added by the path from emotional exhaustion to mental health.

Model selection criteria are suggested by Kaplan (2009): Akaike Information Criterion (AIC), Bayesian Information Criteria (BIC). The lower index scores indicate a better fi tting model. The indices for com-parative fi t to a baseline model suggested by Hu and Bentler (1999) are normed fi t index (NFI), TLI, CFI and RMSEA.

Table III displays the overall fi t indices of the com-peting models. The model M2 presented better fi t indices than the baseline model (M1) for the lower values of AIC and BIC, and the improved values of NFI, TLI, CFI and RMSEA. Furthermore, the model M3 revealed better fi t than the model M2. Thus, the model comparisons indicated the model M3 best accounts for the data. In addition, comparing each standardized regression weight in model M2 (shown in Figure 1) with the corresponding one in model M3 (shown in Figure 2), the standardized regression weights in model Table I. Fit statistics for variables

Variable χ2 df χ2

/df GFI AGFI RMSEA SRMR TLI. CFI Construct reliability

Demands 3.305 2 1.653 0.99 0.97 0.05 0.02 0.99 0.99 0.80 Control 14.677 9 1.631 0.99 0.97 0.05 0.03 0.99 0.99 0.87 Emotional exhaustion 10.238 5 2.048 0.99 0.96 0.06 0.02 0.99 0.99 0.89 Mental health 6.938 5 1.388 0.99 0.97 0.04 0.02 0.99 0.99 0.80

Table II. Means, standard deviation, internal reliability estimates and correlations for all variables

Variable Cronbach’s α Means SD 1 2 3 4

1. Job demands 0.79 2.40 0.53 —

2. Job control 0.86 2.49 0.57 −0.17** —

3. Emotional exhaustion 0.88 3.01 1.44 0.28*** −0.29*** —

4. Mental health 0.79 2.94 0.46 −0.13* 0.30*** −0.45*** — * p < 0.05; ** p < 0.01; *** p < 0.001.

Table III. Fit statistics for structural equation model comparisons (N = 297)

Model χ2 df χ2

/ df GFI AGFI NFI TLI CFI RMSEA SRMR AIC BIC

M1: Baseline model 370.159 170 2.177 0.88 0.86 0.87 0.91 0.92 0.06 0.16 450.2 597.9 M2: Direct effects of job demands

and job control on emotional exhaustion and mental health

292.446 165 1.772 0.91 0.89 0.90 0.94 0.95 0.05 0.09 382.4 548.7

M3: M2 + emotional exhaustion → mental health (fi nal model)

(8)

Demands .67 D05A .82 .78 D04A e3 .88 .28 D03A e2 .53 .30 D01A e1 .55 Control .69 C19A e8 .69 C18A e7 .58 C17A e6 .24 C16A e5 .83 .83 .76 .49 .12 Mental Health .44 G11C e19 .65 G09C e18 .30 G08C e17 .51 G05C e16 .66 .81 .54 .72 .18 Emotional Exhaustion .59 B01 e11 .65 B02 e12 .73 B03 e13 .78 B04 e14 .77 .81 .86 .88 r1 r2 .30 .28 -.12 -.28 .31 B06 e15 .55 .70 C20A e9 .83 .31 C22A e10 .56 -.16 .32 G12C e20 .56 e4 (n.s.) * p <.05; ** p<.01; *** p<.001

Figure 1 The model 2 (N = 297)

M3 almost decreased. Especially the effect of job control on mental health in M3 largely decreased, but was still signifi cant, showing the partial mediation of emotional exhaustion in the relationship between job control and mental health. These fi ndings suggested that the model M3 was the fi nal model and fi tted well to the data [χ2(df = 164, N = 297) = 242.615, χ2

/df = 1.479, GFI = 0.92, AGFI = 0.90, NFI = 0.91, TLI = 0.97, CFI = 0.97, RMSEA = 0.04, SRMR = 0.05).

Parameter estimates

The fi nal model (see Figure 2) supported our hypoth-eses by and large. Hypothesis 1 predicted that job demands are positively related to emotional exhaustion. In the fi nal model, signifi cant positive relationship was found between job demands and emotional exhaustion

(β = 0.28, p < 0.001). Thus, hypothesis 1 was supported. Hypothesis 2 predicted that job control is negatively related to emotional exhaustion. Signifi cant negative relationship was found between job control and emo-tional exhaustion (β = −0.27, p < 0.001) in the fi nal model, supporting hypothesis 2.

Hypothesis 3 predicted that the relationship between job demands and mental health would be mediated by emotional exhaustion. In addition to job demands being signifi cantly positively related to emotional exhaustion, emotional exhaustion was in turn signifi -cantly negatively related to mental health (β = –0.49,

p < 0.001) in the fi nal model. And the direct path from

job demands to mental health was non-signifi cant (β = 0.03, p > 0.05). Therefore, we can conclude that higher levels of job demands were associated with higher levels of emotional exhaustion, which were in turn associated

(9)

with lower levels of mental health; thus, emotional exhaustion fully mediated the relationship between job demands and mental health. Hypothesis 3 was supported.

Hypothesis 4 predicted that emotional exhaustion will mediate the positive relationship between job control and mental health. Signifi cantly negative rela-tionship was found between job control and emotional exhaustion, and signifi cantly negative relationship was found between emotional exhaustion and mental health as well. In addition, the direct path from job control to mental health was signifi cantly positive (β = 0.16, p < 0.05). Therefore, lower levels of job control were associated with higher levels of emotional exhaus-tion, which will be in turn associated with lower levels of mental health. Moreover, job control was positively related with mental health. Thus, we can conclude that

emotional exhaustion partially mediated the relation-ship between job control and mental health. Hypothesis 4 was partially supported.

Discussion

To summarize, the fi nal model supported: (1) job demands are positively related to emotional exhaustion (hypothesis 1); (2) job control is negatively related to emotional exhaustion (hypothesis 2); (3) the expected mediating process from job demands via emotional exhaustion to mental health (hypothesis 3); and (4) emotional exhaustion mediates the positive relation-ship between job control and mental health (hypothesis 4). Moreover, job control was related to mental health as well. These results were almost in accordance with our hypotheses. Emotional exhaustion is, therefore, an

Demands .66 D05A .81 .79 D04A e3 .89 .27 D03A e2 .52 .30 D01A e1 .55 Control .69 C19A e8 .69 C18A e7 .59 C17A e6 .24 C16A e5 .83 .83 .76 .49 .31 Mental Health .45 G11C e19 .62 G09C e18 .30 G08C e17 .53 G05C e16 .67 .79 .55 .73 .17 Emotional Exhaustion .60 B01 e11 .65 B02 e12 .74 B03 e13 .78 B04 e14 .77 .80 .86 .88 r1 r2 .16 .28 .03 -.27 -.49 .30 B06 e15 .55 .70 C20A e9 .83 .31 C22A e10 .55 -.16 .32 G12C e20 .57 e4 (n.s.) * p <.05; ** p<.01; *** p<.001

(10)

important mediator in the relationship between JDC model and mental health.

Theoretical implications

Our research provides a contribution on taking job demands, job control and emotional exhaustion all into consideration to test the relationship with mental health. We also identifi ed emotional exhaustion as a mediator between job demands and mental health, and between job control and mental health as well. Specifi -cally, this fi nding helps to clarify the process of psycho-logical work characteristics (job demands, job control) and mental health, in which emotional exhaustion plays a key mediator.

The primary goal of the current study was the devel-opment and evaluation of a process model linking JDC model (job demands and job control), job-related mental health (emotional exhaustion) and context-free mental health (mental health). The researchers attempted to look into the process of job stressors leading to mental health. Based on previous research and theorizing (e.g. Bakker, Demerouti, & Euwema, 2005; Kelloway & Barling, 1991; Schaufeli & Bakker, 2004; Warr, 1987), a theoretical model was developed. The researchers proposed the hypotheses that emo-tional exhaustion might be the mediator in the process. The results of the present study confi rmed our expecta-tions. The fi nal model provided a good fi t to the data. These results are taken as support for the hypothesized structural model and, in particular, the mediating role of emotional exhaustion between psychological job characteristics (job demands and job control) and con-text-free mental health.

Moreover, in this study, job demands and job control were found to act through different paths in the rela-tionship with mental health via emotional exhaustion. Emotional exhaustion fully mediated the relationship between job demands and mental health, whereas emo-tional exhaustion partially mediated the relationship between job control and mental health. In addition, job control was related to mental health as well. The plau-sible reason might be that cognitive and emotional workload may evoke chronic stress, overfatigue and fi nally burnout, which may lead to psychosomatic dis-orders and complaints (Hakanen et al., 2006; Rudow, 1999). Thus, job demands are more likely to be related to affective well-being. Whereas, job control is regarded as one of the job resources (Taris et al., 2001), high

levels of job resources will encourage personal invest-ment in work (Hakanen et al., 2006). Thus, job control seems to have more than an emotional connection to overall mental health, and is therefore, not only related to emotional exhaustion, but to the more global con-struct of context-free mental health as well.

Previous research has suggested that job characteris-tics are associated with individual well-being (e.g. Wall & Clegg, 1981; Wall, Clegg, & Jackson, 1978; Wall, Corbett, Martin, Clegg, & Jackson, 1990). Emotional exhaustion is unanimously regarded as a consequence of chronic work-related stress (Maslach et al., 2001), whereas context-free mental health is a more global construct which is not tied to a particular setting or context. Based on Warr’s (1987) conceptualization, ‘context-specifi c’ and ‘context-free’ mental health, the current study investigated the process in which job demands and job control were associated with job-related mental health (emotional exhaustion) and, in turn, related to context-free mental health. This con-clusion will contribute to the mechanism of how psy-chological job characteristics impact on employees’ mental health. Our fi ndings suggest that a focus on both processes and outcomes is a more sophisticated research strategy than a focus on outcomes alone (e.g. Hunter, 1990).

With the exception of the studies of Xie (1996) and Xie and Schaubroeck (2008) among Chinese workers, few of the studies of Karasek’s model have been con-ducted in non-Western societies. Our present study could be the start of the accumulation of understanding the relationship among JDC model, emotional exhaus-tion and mental health from the empirical evidence in Asia. This is especially so given the different cultural conditions Asia faces compared with the United States and other European countries. Our study also provides a re-examination of the JDC model in the work motiva-tion literature. Without more empirical studies in Asia, it would not be possible to compare and contrast the present set of fi ndings with those in other countries except for those found in the United States.

Managerial implications

According to our research results, in the work context, job demands and job control will be related to employ-ees’ emotional exhaustion fi rst and the association with context-free mental health will occur thereafter. Therefore, managers should monitor the symptoms

(11)

of employees’ job-related mental health (emotional exhaustion) and introduce some managerial practices immediately in order to prevent the harmful outcomes of mental health. Although the implemented manage-rial practices may have effects on both emotional exhaustion and context-free mental health, the point is as soon as the symptoms of the employees’ emotional exhaustion are detected, manager can adopt managerial practices at once before the negative, distal effects on employees’ psychological health become manifested; thus, the harmful outcomes can be prevented and employees’ mental health can be protected.

Some suggestions are accordingly proposed for either the organization or the individuals. In terms of the organization, although the gradually increasing pressure of the organization’s job demands becomes a universal and unchangeable fact, organizations still need to consider the affordability of workload when in cases of increasing employees’ job demands, excessive job demands result in seriously consequential negative outcomes. On the other hand, Dollard, Winefi eld, Winefi eld, and de Jonge (2000) have argued that it is not necessary to rely on lowering the job demands in order to alleviate job strain; instead, it can be achieved by higher job control. Hence, organizations should increase employee’s job control such as job redesign, fl exible work schedules and goal setting. Because job demands increase but job control does not accordingly accompany them, this will lead to the high risk of employee health concerns (Cahill, 1996; Schaufeli, Maslach, & Marek, 1993). Job strain is related to a range of health outcomes (e.g. coronary heart disease, mus-culoskeletal injuries) for individuals, as well as increased costs and productivity losses in organizations (e.g. Alavinia, Molenaar, & Burdorf, 2009; Karasek & Theorell, 1990; Polanyi & Tompa, 2004; Theorell & Karasek, 1996). From the viewpoint of organizational intervention, stress management programs such as yoga, meditation, counseling, time management and employee assistance programmes can help to improve employees’ physical and mental health (e.g. Blumenthal et al. 2005; Daubenmier et al. 2007; Kirk & Brown, 2003; Yu, Lin, & Hsu, 2009). Organizations will gain a competitive advantage if they make better use of their human resources (Parker & Sprigg, 1999).

Furthermore, organizations must take active improvement to enhance employee health once manag-ers sense the signs of employees’ emotional exhaustion. In this study, the remarkable fi nding is that emotional

exhaustion serves as the key mediator between the rela-tionship of job demands, job control and mental health. In other words, high job demands and low job control will result in further infl uences on mental health through the mediator of emotional exhaustion. Orga-nizations should lay emphasis on employee health in order to better maintain satisfactory human resources. In addition, managers should have keen awareness to perceive the occurrence of emotional exhaustion in the organizational atmosphere, and rebuild organizational practices, in order to help improve employees’ health.

In terms of the employees, individuals should have self-awareness of emotional exhaustion and acknowl-edge the impact of emotional exhaustion on mental health. Individual-based interventions to reduce the symptoms of emotional exhaustion might also be an avenue to explore and apply. For example, stress man-agement programmes that use a cognitive–behavioural approach are consequently effective in reducing stress reactions, including emotional exhaustion, to improve the psychological health. Even so, such individual-based programmes should be supplemented by organization-based programmes in order to be effective in the long run.

Limitations and future research

Several limitations of our research may offer additional research opportunities. First, all data are based on self-report, meaning that the magnitudes of the effects that we reported may have been biased because of common method variance or the wish to answer consistently (Conway, 2002). However, recently, Spector (2006) has argued that common method variance is not that trou-blesome as one might expect in studies such as the current one. Moreover, although the self-report is applied in this study for the measurement of job stress-ors, emotional exhaustion and mental health, our research design utilizes the three-wave design by col-lecting questionnaires in three different time frames. This could lower the deviation of common method variance. Nevertheless, in order to avoid common method variance, future research should include non-self-reports as well, such as peer ratings from colleagues (Burke & Ng, 2007) or acquaintances (Aziz & Zickar, 2006; Bakker, Demerouti, & Burke, 2009).

Although three-wave design is applied in this study, and the independent variables, mediating variable and dependent variable were collected separately at three

(12)

different points of time, we cannot empirically deter-mine the causal sequencing of effects. Therefore, in the future, a longitudinal panel data for each variable would enhance the ability to verify a causal relationship.

In terms of the issue of generalizability, even though the sample was selected systematically from one single organization, readers should be cautious when general-izing the results to employees of other public or private organizations. Future research should replicate our results in other occupational groups.

Furthermore, previous studies focusing on work stress and well-being are mostly conducted in Western culture countries (e.g. De Jonge et al., 2001; de Lange et al., 2003; Ter Doest & De Jonge, 2006; Wong, Hui, & Law, 1998; Zapf, Dormann, & Frese, 1996). Our study was conducted in a rising Asian country— Taiwan. This study can help us to gain better compre-hension of the relationships between job characteristics, emotional exhaustion and mental health, and reinforce the theoretical foundation. Our fi ndings will contribute to expand understanding of the stressor–strain rela-tionships culturally; nevertheless, more cross-cultural studies undertaken in the future will provide further benefi ts to researchers and practitioners.

As suggested by Hakanena, Schaufelib, and Ahola (2008), job resources may play either an intrinsic moti-vational role because they foster employees’ growth, learning and development, or they may play an extrin-sic motivation role because they are instrumental in achieving work goals. In the former studies, according to self-determination theory (Deci, Vallerand, Pelletier, & Ryan, 1991), any social context that satisfi es the basic human needs of autonomy (job control), competence and relatedness (social support) will enhance well-being. Therefore, future research can consider those variables which will improve employee’s well-being and work performance in order to expand the theoretical model.

Acknowledgments

The authors thank Dr Tahira Probst and two anony-mous reviewers for their helpful comments on earlier drafts of this paper.

REFERENCES

Alavinia, S.M., Molenaar, D., & Burdorf, A. (2009). Pro-ductivity loss in the workforce: Associations with health,

work demands, and individual characteristics. American

Journal of Industrial Medicine, 52, 49–56.

Alfredsson, L., Spetz, C.L., & Theorell, T. (1985). Type of occupation and near-future hospitalization for myocar-dial infarction and some other diagnoses. International

Journal of Epidemiology, 14, 378–388.

Arnold, K.A., Turner, N., & Barling, J. (2007). Transfor-mational leadership and psychological well-being: The mediating role of meaningful work. Journal of

Occupa-tional Health Psychology, 12, 193–203.

Aziz, S., & Zickar, M.J. (2006). A cluster analysis investiga-tion of workaholism as a syndrome. Journal of

Occupa-tional Health Psychology, 11, 52–62.

Bakker, A.B., Demerouti, E., De Boer, E., & Schaufeli, W.B. (2003a). Job demands and job resources as predictors of absence duration and frequency. Journal of Vocational

Behavior, 62, 341–356.

Bakker, A.B., Demerouti, E., & Schaufeli, W.B. (2003b). Dual processes at work in a call centre: An application of the job demands—resources model. European Journal

of Work and Organizational Psychology, 12, 393–417.

Bakker, A.B., Demerouti, E., & Verbeke, W. (2004). Using the job demands–resources model to predict burnout and performance. Human Resource Management, 43, 83–104.

Bakker, A.B., Demerouti, E., & Euwema, M. (2005). Job resources buffer the impact of job demands on burnout.

Journal of Occupational Health Psychology, 10, 170–

180.

Bakker, A.B., Demerouti, E., & Burke, R. (2009). Worka-holism and relationship quality: A spillover–crossover perspective. Journal of Occupational Health Psychology,

14, 23–33.

Banks, M., Clegg, C., Jackson, P., Kemp, N., Stafford, E., & Wall, T. (1980). The use of the general health ques-tionnaire as an indicator of mental health in occupa-tional studies. Journal of Occupaoccupa-tional Psychology, 53, 187–194.

Belkic, K., Landsbergis, P.A., Schnall, P.L., & Baker, D., (2004). Is job strain a major source of cardiovascular disease risk? Scandinavian Journal of Work, Environment

& Health, 30, 85–128.

Blumenthal, J.A., Sherwood, A., Babyak, M.A., Watkins, L.L., Waugh, R., Georgiades, A., Bacon, S.L., Hayano, J., Coleman, R.E., & Hinderliter, A. (2005). Effects of exer-cise and stress management training on markers of car-diovascular risk in patients with ischemic heart disease: A randomized controlled trial. Journal of the American

Medical Association, 293, 1626–1634.

Bourbonnais, R., Comeau, M., & Vézina, M. (1999). Job strain and evolution of mental health among nurses. Journal of Occupational Health Psychology, 4, 95–107.

(13)

Burke, R.J., & Ng, E.S.W. (2007). Workaholic behaviours: Do colleagues agree? International Journal of Stress

Man-agement, 24, 312–320.

Byrne, B.M. (2001). Structural equation modeling with

AMOS: Basic concepts, applications, and programming.

Mahwah, NJ: Lawrence Erlbaum Associates.

Cahill, J. (1996). Psychosocial aspects of interventions in occupational safety and health. American Journal of

Industrial Medicine, 29, 308–313.

Conway, J.M. (2002). Method variance and method bias in industrial and organizational psychology. In S.G. Rogelberg (Ed.), Handbook of research methods in

orga-nizational and industrial psychology (pp. 344–365).

Malden, MA: Blackwell Publishers.

Cook, T.D. & Campbell, D.T. (1979).

Quasi-Experimentation: Design & Analysis Issues for Field Settings. Boston, MA: Houghton Miffl in Company.

Cordes, C.L., & Dougherty, T.W. (1993). A review and integration of research on job burnout. Academy of

Management Review, 18, 621–656.

Cox, T., Kuk, G., & Leiter, M.P. (1993). Burnout, health, work stress and organizational healthiness. In W.B. Schaufeli, C. Maslach, & T. Marek (Eds), Professional

burnout: Recent development in theory and research (pp.

177–197). Washington, DC: Taylor & Francis.

Daubenmier, J.J., Weidner, G., Sumner, M.D., Mendell, N., Merritt-Worden, T., Studley, J., & Ornish, D. (2007). The contribution of changes in diet, exercise, and stress management to changes in coronary risk in women and men in the multisite cardiac lifestyle intervention program. Annals of Behavioral Medicine, 33, 57–68. Deci, E.L., Vallerand, R.J., Pelletier, L.G., & Ryan,

R.M. (1991). Motivation and education: The self-determination perspective. Educational Psychologist, 26, 325–346.

De Jonge J., Dormann, C., Janssen, P.P.M., Dollard, M.F., Landeweerd, J.A., & Nijhuis, F.J.N. (2001). Testing reciprocal relationships between job characteristics and psychological well-being: A cross-lagged structural equation model. Journal of Occupational and

Organiza-tional Psychology, 74, 29–46.

de Lange, A.H., Taris, T.W., Kompier, M.A.J., Houtman, I.L.D., & Bongers, P.M. (2002). Effects of stable and changing demand-control histories on worker health.

Scandanavian Journal of Work, Environment, & Health, 28, 94–108.

de Lange, A., Taris, T.W., Kompier, M.A., Houtman, I.L.D., & Bongers, P.M. (2003). The very best of the millennium: longitudinal research and the demand– control– (support) model. Journal of Occupational

Health Psychology, 8, 282–305.

Demerouti, E., Bakker, A.B., de Jonge, J., Janssen, P.P.M., & Schaufeli, W.B. (2001). Burnout and engagement at

work as a function of demands and control.

Scandina-vian Journal of Work, Environment and Health, 27,

279–286.

Dollard, M.F., Winefi eld, H.R., Winefi eld, A.H., & de Jonge. J. (2000). Sychosocial job strain and productivity in human service workers: A test of the demand– control–support model. Journal of Occupational and

Organizational Psychology, 73, 501–510.

Feij, J.A., Whitely, W.T., Peiró, J.M., & Taris, T.W. (1995). The development of career-enhancing strategies and content innovation: A longitudinal study of new workers. Journal of Vocational Behavior, 46, 231– 256.

Fox, M.L., Dwyer, D.J., & Ganster, D.C. (1993). Effects of stressful job demands and control on physiological and attitudinal outcomes in a hospital setting. Academy of

Management Journal, 36, 289–318.

Goldberg, D. (1972). The detection of mental illness by

questionnaire. London: Oxford University Press.

Goldberg, D., & Williams, P. (1988). GHQ: A user’s

guide to the general health questionnaire. Windsor:

NFER-Nelson.

Green, D.E., Walkey, F.H., & Taylor, A.J.W. (1991). The three-factor structure of the Maslach burnout inven-tory. Journal of Social Behavior and Personality, 6, 453–472.

Guglielmi, R.S., & Tatrow, K. (1998). Occupational stress, burnout, and health in teachers: A methodological and theoretical analysis. Review of Educational Research, 68, 61–99.

Hair, J.F., Anderson, R.E., Tatham, R.L., & Black, W.C. (1998). Multivariate data analysis (5th ed.). UK: Pren-tice Hall International.

Hakanen, J.J., Bakker, A.B., & Schaufeli, W.B. (2006). Burnout and work engagement among teachers. Journal

of School Psychology, 43, 495–513.

Hakanena, J.J., Schaufelib, W.B., & Ahola, K. (2008). The job demands–resources model: A three-year cross-lagged study of burnout, depression, commitment, and work engagement. Work & Stress, 22, 224–280.

Halbesleben, J.R.B., & Bowler, W.M. (2007). Emotional exhaustion and job performance: The mediating role of motivation. Journal of Applied Psychology, 92, 93–106. Hammar, N., Alfredsson, L., & Theorell, T. (1994). Job

characteristics and the incidence of myocardial infarction. International Journal of Epidemiology, 23, 277–284.

Häusser, J.A., Mojzisch, A., Niesel, M., & Schulz-Hardt, S. (2010). Ten years on: A review of recent research on the job demand–control (–support) model and psychologi-cal well-being. Work & Stress, 24, 1–35.

Holman, D.J., & Wall, T.D. (2002). Work characteristics, learning-related outcomes, and strain: A test of

(14)

competing direct effects, mediated, and moderated models. Journal of Occupational Psychology, 7, 283– 301.

Hu, L.T., & Bentler, P.M. (1999). Cutoff criteria for fi t indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation

Modeling, 6, 1–55.

Hunter, J.E. (1990). Causal analysis of experiments and interventions. Paper presented at the annual meeting of the Society for Industrial and Organizational Psychol-ogy, Miami Beach, FL.

Hvid, H., Lund, H., & Pejtersen, J. (2008). Control, fl exi-bility and rhythms. Scandinavian Journal of Work

Envi-ronment and Health Supplement, 6, 83–90.

Janssen, P.P.M., Peeters, M.C.W., de Jong, J., Houkes, I., & Tummers, G.E.R. (2004). Specifi c relationships between job demands, job resources and psychological outcomes and the mediating role of negative work-home interference. Journal of Vocational Behavior, 65, 411–429.

Jimmieson, N. (2000). Employee reactions to behavioural control under conditions of stress: The moderating role of self-effi cacy. Work & Stress, 14, 262–280.

Johnson, J.V. (2008). Globalization, workers’ power and the psychosocial work environment—is the demand– control–support model still useful in a neoliberal era?

Scandinavian Journal of Work Environment and Health Supplement, 6, 15–21.

Jöreskog, K.G. (1993). Testing structural equation models. In K.A. Bollen, & J. Scott Long (Eds), Testing structural

equation models (pp. 294–316). Newbury Park, CA:

Sage.

Jöreskog, K.G., & Sörbom, D. (1986). LISREL VI: Analysis

of linear structural relationships by the method of maximum likelihood. Chicago, IL: National Educational

Resources.

Kaplan, D. (2009). Structural equation modeling:

Founda-tions and extensions (2nd ed.). Los Angeles, CA: SAGE.

Karasek, R.A. (1979). Job demand, job decision latitude, and mental strain: Implications for job redesign.

Admin-istrative Science Quarterly, 24, 285–315.

Karasek, R.A., & Theorell, T. (1990). Healthy work: Stress,

productivity, and the reconstruction of working life. New

York: Basic Books.

Kelloway, E.K., & Barling, J. (1991). Job characteristics, role stress and mental health. Journal of Occupational

Psychology, 64, 291–304.

Kirk, A.K., & Brown, D.F. (2003). Employee assistance programs: A review of the management of stress and wellbeing through workplace counselling and consult-ing. Australian Psychologist, 38: 138–143.

Kivimäki, M., Leino-Arjas, P., Luukkonen, R., Riihimäki, H., Vahtera, J., & Kirjonen, J. (2002). Work stress and

risk of cardiovascular mortality: Prospective cohort study of industrial employees. British Medical Journal,

325, 857.

Kivimäki, M., Virtanen, M., Elovainio, M., Kouvonen, A., Väänänen, A., & Vahtera, J. (2006). Work stress in the etiology of coronary heart disease: Systematic review and meta-analysis of prospective cohort studies.

Scan-dinavian Journal of Work, Environment & Health, 32,

431–442.

Kristensen, T.S. (1996). Job stress and cardiovascular disease: A theoretic critical review. Journal of

Occupa-tional Health Psychology 1, 246–260.

Lee, R.T., & Ashforth, B.E. (1993). A further examination of managerial burnout: Toward an integrated model.

Journal of Organizational Behavior, 14, 3–20.

Lee, R.T., & Ashforth, B.E. (1996). A meta-analytic exami-nation of the correlates of the three dimensions of burnout. Journal of Applied Psychology, 81, 123–133. Leiter, M.P., & Schaufeli, W.B. (1996). Consistency of the

burnout construct across occupations. Anxiety, Stress &

Coping, 9, 229–243.

van der Linden, D., Keijsers, G.P.J., Eling, P., & van Schaijk, R. (2005). Work stress and attentional diffi culties: An initial study on burnout and cognitive failures. Work &

Stress, 19, 23–36.

Maslach, C. (1982). Burnout: The cost of caring. Englewood Cliffs, NJ: Prentice Hall.

Maslach, C. (1993). Burnout: A multidimensional per-spective. In W.B. Schaufeli, C. Maslach, & T. Marek (Eds), Professional burnout: Recent development in theory

and research (pp. 19–32). Washington, DC: Taylor &

Francis.

Maslach C., & Schaufeli W.B. (1993). Historical and con-ceptual development of burnout. In W.B. Schaufeli, C. Maslach, & T. Marek (Eds), Professional burnout:

Recent developments in theory and research (pp. 1–16).

Washington, DC: Taylor & Francis.

Maslach, C., & Schaufeli, W. (2000). Job burnout. Annual

Review of Psychology, 52, 397–422.

Maslach, C., Schaufeli, W.B., & Leiter, M.P. (2001). Job burnout. Annual Review of Psychology, 52, 397–422. Moore, J.E. (2000). Why is this happening? A casual

attribution approach to work emotional exhaustion consequences. Academy of Management Review, 25, 335–349.

Nunnally, J.C., & Bernstein, I.H. (1994). Psychometric

theory (3rd ed.). New York: McGraw-Hill.

Parker, S.K., & Sprigg, C.A. (1999). Minimizing strain and maximizing learning: The role of job demands, job control, and proactive personality. Journal of Applied

Psychology, 84, 925–939.

Parker, S.L., Jimmieson, N.L., & Amiot, C.E. (2010). Self-determination as a moderator of demands and control:

(15)

Implications for employee strain and engagement.

Journal of Vocational Behavior, 76, 52–67.

Payne, R.L., Wall, T.D., Borrill, C., & Carter, A. (1999). Strain as a moderator of the relationship between work characteristics and work attitudes. Journal of

Occupa-tional Health Psychology, 4, 3–14.

Podsakoff, P.M., MacKenzie, S.B., Lee, J-Y., & Podsakoff, N.P. (2003). Common method biases in behavioral research: A critical review of the literature and recom-mended remedies. Journal of Applied Psychology, 88, 879–903.

Polanyi M., & Tompa, E. (2004). Rethinking work-health models for the new global economy: A qualitative analy-sis of emerging dimensions of work. Work, 23, 3–18. Rafferty, Y., Friend, R., & Landsbergis, P.A. (2001). The

association between job skill discretion, decision author-ity and burnout. Work and Stress, 15, 73–85.

Raine-Eudy, R. (2000). Using structural equation model-ing to test for differential reliability and validity: An empirical demonstration. Structural Equation Modeling,

7, 124–141.

Rodriguez, I., Bravo, M.J., Peiro, J.M., & Schaufeli, W. (2001). The demands–control support model, locus of control, and job dissatisfaction: A longitudinal study.

Work and Stress, 15, 97–114.

Rudow, B. (1999). Stress and burnout in the teaching pro-fession: European studies, issues, and research perspec-tives. In A.M. Huberman (Ed.), Understanding and

preventing teacher burnout: A sourcebook of international research and practice (pp. 38–58). New York: Cambridge

University Press.

Salanova, M., Peiró, J.M., & Schaufeli, W.B. (2002). Self-effi cacy specifi city and burnout among information technology workers: An extension of the job demand– control model. European Journal of work and

organiza-tional psychology, 11, 1–25.

Schaufeli, W., & Bakker, A.B. (2004). Job demands, job resources, and their relationship with burnout and engagement. Journal of Organizational Behavior, 25, 293–315.

Schaufeli, W.B., & Enzmann, D. (1998). The burnout

com-panion to study & practice: A critical analysis.

Philadel-phia, PA: Taylor & Francis.

Schaufeli, W.B., Maslach, C., & Marek, T. (1993).

Profes-sional burnout: Recent development in theory and research.

Washington, DC: Taylor & Francis.

Schaufeli, W.B., Leiter, M.P., Maslach, C., & Jackson, S.E. (1996). The Maslach burnout inventory—general survey. In C. Maslach, S.E. Jackson, & M.P. Leiter (Eds),

Maslach burnout inventory—manual (3rd ed., pp.

19–26). Palo Alto, CA: Consulting Psychologists Press. Shirom, A. (2003). Job-related burnout: A review. In J.C.

Quick, & L.E. Tetrick (Eds) Handbook of occupational

health psychology (pp. 245–264). Washington, DC:

American Psychological Association.

Snyder, L.A., Krauss, A.D., Chen, P.Y., Finlinson, S., & Huang, Y.-H. (2008). Occupational safety: Application of the job demand–control–support model. Accident

Analysis and Prevention, 40, 1713–1723.

Spector, P.E. (2006). Method variance in organizational research: Truth or urban legend? Organizational

Research Methods, 2, 221–232.

Tang, C.S.K., Au, W.T., Schwarzer, R., & Schmitz, G. (2001). Mental health outcomes of job stress among Chinese teachers: Role of stress resource factors and burnout. Journal of Organizational Behavior, 22, 887– 901.

Taris, T., & Feij, J.A. (2004). Learning and strain among newcomers: A three-wave study on the effects of job demands and job control. The Journal of Psychology, 138, 543–563.

Taris, T.W., Schreurs, P.J.G., & Van Iersel-van Silfhout, I.J. (2001). Job stress, job strain, and psychological with-drawal among Dutch university staff: Towards a dual process model for the effects of occupational stress.

Work & Stress, 15, 283–296.

Ter Doest, L., & De Jonge, J. (2006). Testing causal models of job characteristics and employee well-being: A repli-cation study using cross-lagged structural equation modeling. Journal of Occupational and Organizational

Psychology, 79, 499–507.

Theorell, T., & Karasek, R.A. (1996). Current issues relat-ing to psychosocial job strain and cardiovascular disease research. Journal of Occupational Health Psychology, 1, 9–26.

Van der Doef, M., & Maes, S. (1999). The job demand– control (–support) model and psychological well-being: A review of 20 years of empirical research. Work and

Stress, 13, 87–114.

Van Veldhoven, M. (1996). Psychosocial job demands and

work stress. Unpublished doctoral dissertation,

Univer-sity of Groningen, The Netherlands.

Van Yperen, N.W., & Hagedoorn, M. (2003). Do high job demands increase intrinsic motivation or fatigue or both? The role of job control and job social support.

Academy of Management Journal, 46, 339–348.

Wall, T.D., & Clegg, C.W. (1981). A longitudinal fi eld study of group redesign. Journal of Occupational

Behav-iour, 2, 31–49.

Wall, T.D., Clegg, C.W., & Jackson, P.R. (1978). An evalu-ation of the job characteristics model. Journal of

Occu-pational Psychology, 51, 183–196.

Wall, T.D., Corbett, M., Martin, R., Clegg, C.W., & Jackson, P.R. (1990). Advanced manufacturing technol-ogy, work design, and performance: A change study.

(16)

Wallace, J.E. (2005), Job stress, depression and work-to-family cinfl ict: A test of the strain and buffer hypo-theses. Relations Industrielles, 60, 510–568.

Warr, P. (1987). Work, unemployment and mental health. New York: Oxford University Press.

Wong, C.S., Hui, C., & Law, K.S. (1998). A longitudinal study of the job perception-job satisfaction relationship: A test of the three alternative specifi cations. Journal

of Occupational and Organizational Psychology, 71,

127–146.

Wright, T.A., & Bonett, D.G. (1997). The contribution of burnout to work performance. Journal of Organizational

Behavior, 18, 491–499.

Xie, J.L. (1996). Karasek’s model in the People’s Republic of China: Effects of job demands, control, and

individ-ual differences. Academy of Management Journal, 39, 1594–1618.

Xie, J.L., & Schaubroeck, J. (2008). Theories of job stress and the role of traditional values: A longitudinal study in China. Journal of Applied Psychology, 93, 831–848. Yu, M.C., Lin, C.C., & Hsu, S.Y. (2009). Stressors and

burnout: The role of employee assistance programs and self-effi cacy. Social Behavior & Personality, 37, 365– 377.

Zapf, D., Dormann, C., & Frese, M. (1996). Longitudinal studies in organizational stress research: A review of literature with reference to methodological issues.

數據

Table II.  Means, standard deviation, internal reliability estimates and correlations for all variables

Table II.

Means, standard deviation, internal reliability estimates and correlations for all variables p.7
Table III.  Fit statistics for structural equation model comparisons (N = 297)

Table III.

Fit statistics for structural equation model comparisons (N = 297) p.7
Table III displays the overall fi t indices of the com- com-peting models. The model M2 presented better fi t  indices than the baseline model (M1) for the lower  values of AIC and BIC, and the improved values of NFI,  TLI, CFI and RMSEA

Table III

displays the overall fi t indices of the com- com-peting models. The model M2 presented better fi t indices than the baseline model (M1) for the lower values of AIC and BIC, and the improved values of NFI, TLI, CFI and RMSEA p.7
Figure 1  The model 2 (N = 297)

Figure 1

The model 2 (N = 297) p.8
Figure 2 The fi nal model (N = 297)

Figure 2

The fi nal model (N = 297) p.9

參考文獻

相關主題 :