Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Burnout and engagement at work as a function of demands and control

Burnout and engagement at work as a function of demands and control Downloaded from www.sjweh.fi on October 30, 2021 Original article Scand J Work Environ Health 2001;27(4):279-286 doi:10.5271/sjweh.615 Burnout and engagement at work as a function of demands and control by Demerouti E, Bakker AB, de Jonge J, Janssen PPM, Schaufeli WB Affiliation: University of Nijmegen, Department of Work & Organizational Psychology, Postbox Office 9104, 6500 HE Nijmegen, Netherlands. E.Demerouti@psych.kun.nl The following articles refer to this text: 2006;32(5):339-348; 2008;34(5):345-355 Key terms: absorption; burnout; burnout at work; dedication; demandcontrol model; engagement; engagement at work; vigor This article in PubMed: www.ncbi.nlm.nih.gov/pubmed/11560342 This work is licensed under a Creative Commons Attribution 4.0 International License. Print ISSN: 0355-3140 Electronic ISSN: 1795-990X Demerouti et al Scand J Work Environ Health 2001;27(4):279—286 1, 2 3 3 4 by Evangelia Demerouti, PhD, Arnold B Bakker, PhD, Jan de Jonge, PhD, Peter PM Janssen, PhD, Wilmar B Schaufeli, PhD Demerouti E, Bakker AB, de Jonge J, Janssen PPM, Schaufeli WB. Burnout and engagement at work as a function of demands and control. Scand J Work Environ Health 2001;27(4):279—286. Objectives The present study was designed to test the demand-control model using indicators of both health impairment and active learning or motivation. Methods A total of 381 insurance company employees participated in the study. Discriminant analysis was used to examine the relationship between job demands and job control on one hand and health impairment and active learning on the other. Results The amount of demands and control could be predicted on the basis of employees’ perceived health impairment (exhaustion and health complaints) and active learning (engagement and commitment). Each of the four combinations of demand and control differentially affected the perception of strain or active learning. Job demands were the most clearly related to health impairment, whereas job control was the most clearly associated with active learning. Conclusions These findings partly contradict the demand–control model, especially with respect to the validity of the interaction between demand and control. Job demands and job control seem to initiate two essentially independent processes, and this occurrence is consistent with the recently proposed job demands–resources model. Keywords absorption, burnout, dedication, demand–control model, engagement, vigor The demand–control model (1, 2) is a theoretical mod- The demand-control model is based on two central el that attempts to provide insight into the relationships assumptions, as reflected by diagonals A and B in fig- between psychosocial work characteristics on one hand ure 1. The first assumption (diagonal A) is that psycho- and health and motivation on the other. Partly because logical strain (such as chronic fatigue, anxiety, and car- of its elegant simplicity, the model has stimulated sev- diovascular complaints) is particularly caused by the eral scientific studies (3, 4, 5). It looks for the determi- combination of high psychological demands and low de- nants of workers’ health and motivation in the interac- cision latitude (quadrant 1), while the opposite, lack of tion of two psychosocial job characteristics, namely, the strain, can be found in a job with low psychological de- “psychological demands” corresponding to psycholog- mands and much decision latitude (quadrant 3). The ical stressors present in the work environment (eg, high second important assumption (diagonal B) is that time pressure, difficult and mentally taxing work) and work motivation, as well as learning and development “decision latitude” (or “job control”) comprising the opportunities, occur if job demands are high (but not worker’s authority to make decisions on the job (“deci- overwhelming) and decision latitude is high (quadrant sion authority”) and the breadth of skills used by the 2). The opposite type of work situation occurs in a job worker on the job [“skill discretion” (6)] (figure 1). in which neither job demands nor decision latitude is Carl von Ossietzky Universität Oldenburg, Department of Work & Organizational Psychology, Oldenburg, Germany. University of Nijmegen, Department of Work & Organizational Psychology, Nijmegen, The Netherlands. Utrecht University, Department of Social and Organizational Psychology, Utrecht, The Netherlands. University of Maastricht, Department of Health Studies, Maastricht, The Netherlands. Reprint requests to: Dr Evangelia Demerouti, University of Nijmegen, Department of Work & Organizational Psychology, Postbox Office 9104, 6500 HE Nijmegen, The Netherlands. [E-mail: E.Demerouti@psych.kun.nl] Scand J Work Environ Health 2001, vol 27, no 4 279 Demand – control model Psychological demands B diagonal ently of the level of job demands. Job resources repre- sent a broader category of “positive” features of the Low High work environment than job control in the demand–con- trol model. Applied to the demand–control model, this statement would mean that the low strain quadrant is Low strain Active High 3 not only related to diminished health impairment, but also to motivation and active learning. The aim of our study was thus to test the demand– control model with special attention to the active-pas- sive hypothesis and to integrate the strain and active learning hypotheses. Specifically, we attempted to pre- Passive High strain dict employees’ levels of job demands and job control Low 4 1 using stress reactions (burnout, health complaints), but also indicators of active learning (engagement, commit- ment). In addition, we used an alternative statistical technique, which allows for a different perspective of the predictions of the demand–control model and an A diagonal examination of the postulated interaction. Figure 1. The demand–control model [source: Karasek (2)]. Burnout and engagement The concept of burnout that was initially relevant only to working with people has recently been expanded to- very pronounced (quadrant 4). This “passive” work sit- wards all other professions and occupational groups (18, uation is characterized by a decrease in work activities 19). Based on the original instrument to measure hu- and “negative learning” — a gradual loss of acquired man service-related burnout [Maslach Burnout Inven- skills. tory (20)] a new scale [Maslach Burnout Inventory— Recent reviews of the model contend that the de- General Survey (21)] has been developed for use out- mand–control model is appropriate for further investi- side the human services. The dimensions of the latter gation since it can predict health and, to a less extent, (ie, exhaustion, cynicism, professional efficacy) paral- motivational and productivity outcomes (3—5, 7—9,). lel those of the original inventory (ie, emotional exhaus- However, the “active-passive” dimension of the model tion, depersonalization, personal accomplishment) in has been underutilized in organizational research (10). the sense that they are more generic and do not refer to Only a few studies provide some evidence for the ac- the people one is working with. Exhaustion is measured tive-passive dimension of the model (11—14). Studies by items that refer to fatigue, but do not make direct measuring both strain and motivation or active learning reference to people as the source of such feelings. The find stronger support for the model than do those re- items assessing cynicism reflect indifference or a dis- stricted to strain consequences (15). tant attitude towards work in general, not necessarily One possible reason for the lack of attention to the towards people. Finally, professional efficacy encom- active-passive hypothesis may be that most of the inter- passes both social and nonsocial aspects of occupational est in the model originated in the field of epidemiolo- accomplishments. Psychometric research with the gy, with cardiovascular diseases as the main form of Maslach Burnout Inventory—General Survey demon- strain (7, 8). Another reason may be that the B diagonal strated that the three-factor structure is invariant across [“motivation for learning new behavior patterns” (6)] is occupational groups (19). hard to operationalize as an independent psychological A recent development in burnout research is the shift construct, since it may be conceptually confounded by towards its opposite, engagement (22). In the view of the job characteristic skill discretion (3). Maslach & Leiter (23), workers have a sense of ener- According to Karasek (1), strain and learning are getic and effective connection with their work activi- two, more or less, independent phenomena, and conse- ties (energy, involvement), and they see themselves as quently no attempts have been made to integrate the two able to deal with the demands of their job (professional assumptions in the model. When, for example, job con- efficacy). Bakker & Schaufeli (24) partly agree with trol is high and job demands are low, we should expect their description and defined engagement as a positive, low strain, but nothing is explicitly said about the level fulfilling, work-related state of mind that is character- of motivation. Demerouti et al (16, 17) have argued and ized by vigor, dedication, and absorption. Rather than shown that, when job resources are high, we should ex- a momentary and specific state, engagement refers to a pect motivation in the form of engagement, independ- 280 Scand J Work Environ Health 2001, vol 27, no 4 Decision latitude (control) Demerouti et al more persistent and pervasive affective-motivational the respondents were men (66%). The mean age of the state that is not focused on any particular object, event, group was 40 (SD 10.4) years, and the mean organiza- individual, or behavior. Vigor is characterized by high tional tenure was 12.5 (SD 10.5) years. Twelve percent levels of energy and mental resilience while working, of the sample had a managerial position. Most of the the willingness to invest effort in one’s work, and per- participants had a full-time job (64%). The group was sistence even in the face of difficulties. Dedication is recruited at their job sites after an informative meeting characterized by a sense of significance, enthusiasm, with the personnel department and management. All 624 inspiration, pride, and challenge. Absorption is charac- employees received an informative letter about the study terized by being fully concentrated and engrossed in from the management team, together with the question- one’s work, whereby time passes quickly. Bakker & naire and a return envelope. The confidentiality and ano- Schaufeli (24) found evidence for the psychometric nymity of the data were emphasized. The participants were quality (ie, internal consistency, factorial validity) of the kindly requested to fill out the questionnaire at home and engagement construct and the independence of the burn- to place it in a special box in their department. out and engagement dimensions. Measures Present study In this study, psychological job demands refer to the Our study was designed to test the job strain and active extent to which the work pace is high and the availabil- learning hypotheses generated by the demand–control ity of sufficient time to execute the required work. Job model (1, 2). Insurance company employees were cho- demands were measured with five items of the Dutch sen for this purpose. The sample was suitable for test- version of the Job Content Questionnaire (26). The four ing the model for several reasons. First, because of the response categories ranged from 1 “never” to 4 “always”. different types of tasks and specialties, insurance em- [Example item: “Do you have a lot of work to do?”] ployees are a relatively heterogeneous group as far as Job control was conceived as the degree to which their work characteristics are concerned, which implies employees have the freedom to make decisions regard- that there is considerable variance in their job demand ing the work assignment and their work methods. Nine and control. Second, the group chosen had relatively items of the Dutch version of the Job Content Question- long job tenure. Since the demand–control model refers naire (26) were used. The response format was identi- to chronic stressors (2, 25), the effect of such stressors cal to that used for job demands. [Example item: “Do is most obvious for job holders with long experience in you participate in decisions regarding your work?”] their current job. Burnout was measured with the Dutch version (27) A special feature of the study was not only the meas- of the Maslach Burnout Inventory—General Survey urement of psychological strain but also that of indica- (21). The instrument consists of 16 items that measure tors of active learning and motivation. As strain and ac- the three components of burnout (exhaustion, cynicism tive learning are not one-dimensional, we operational- and professional efficacy). The items were scored on a ized them through the use of several indicators. For that 7-point scale (0 = never, 6 = always). [Example items: purpose, burnout and engagement were included. How- “I feel used up at the end of the workday” (exhaustion), ever, burnout is a complex syndrome that consists of an “I have become less enthusiastic about my work” (cyn- orthodox stress reaction (exhaustion), as well as of an icism), and “In my opinion, I am good at my work” (pro- attitudinal component (cynicism and lack of efficacy) fessional efficacy).] that is closely linked to motivation. In other words, when Psychosomatic health complaints refer to the number workers are cynical and do not feel efficacious, they of psychosomatic symptoms that a person experienced may not be willing to learn actively. In addition psy- during the year preceding the study. Health complaints chosomatic complaints on one hand and engagement (in- were measured with 13 items of a well-validated ques- cluding vigor, dedication, and absorption) and commit- tionnaire on health complaints (28). Employees could ment on the other were included as indices of strain and respond using a 4-point scale (1 = seldom or never, 4 = active learning, respectively. very often). [Example items: “Do you have unpleasant pains on the breast?” and “Do you have headaches?”] Engagement was assessed with the Utrecht work engagement scale (24). The scale includes 15 items, Subjects and methods which measure the three components of engagement (vigor, dedication and absorption). All the items were Sample and procedure rated on a 7-point scale ranging from 0 “never” to 6 Altogether, 381 employees from an insurance company “always”. [Example items: “During my work, I feel participated in the study (response rate = 61%). Most of vital and strong” (vigor), “I am enthusiastic about my Scand J Work Environ Health 2001, vol 27, no 4 281 Demand – control model Table 1. Means (M), standard deviations (SD), internal consistencies (α) and Pearson correlations of the variables included in this study (N = 381). Variable M SD α 1 2 34 56 7 8 9 1. Job demands 2.69 0.43 0.85 2. Job control 2.87 0.35 0.75 0.04 3. Exhaustion 1.79 1.08 0.89 0.25** -0.20** 4. Cynicism 1.49 0.94 0.74 0.11* -0.29** 0.60** 5. Personal competence 4.02 0.85 0.78 0.15** 0.44** -0.18** -0.30** 6. Health complaints 1.44 0.37 0.86 0.13** -0.23** 0.56** 0.41** -0.17** 7. Vigor 3.92 0.88 0.86 0.08 0.40** -0.41** -0.44** 0.65** -0.37** 8. Dedication 4.19 0.96 0.90 0.05 0.53** -0.29** -0.50** 0.69** -0.28** 0.76** 9. Absorption 3.56 0.98 0.80 0.07 0.42** -0.16** -0.32** 0.52** -0.18*** 0.73** 0.75** 10. Commitment 3.61 0.60 0.80 -0.05 0.30** -0.36** -0.44** 0.40** -0.24** 0.49** 0.59** 0.45** * P < 0.05, ** P < 0.01. 1.00 job” (dedication), and “Time flies when I am working” (absorption).] Commitment refers to the relationship of employees to the organization in which they work. It is measured 0.50 with the 5 items of the affective commitment scale de- HD & LC HD & LC veloped by Mowday et al (29). [Example item: “I tell HD & HC HD & HC my friends and family that my organization is a pleas- ant organization to work for” (1 = totally disagree, 5 = 0.00 totally agree).] LD & LC LD & LC LD & H C LD & HC Data analysis strategy -0.50 The strategy followed in the present study was to look for different generalized patterns of health impairment and active learning to indicate the differences between -1.00 the four combinations of job demands and job control –1.00 –0.50 0.00 0.50 1.00 (ie, the four quadrants in figure 1). If these four combi- Discriminant function 1 nations are related to health impairment and active learning, it should be possible to separate the four quad- Figure 2. Group centroids for the four combinations of demands and rants on the basis of the response patterns to questions control in a two-dimensional space. LD & LC = low demands and low about health impairment (exhaustion, health complaints) control; LD & HC = low demands and high control; HD & LC = high and active learning (engagement, professional efficacy, demands and low control; HD & HC = high demands and high control. cynicism, commitment) by means of a discriminant anal- ysis. This form of statistical treatment of the problem offers the advantage of making use of the common var- Table 2. Standardized canonical coefficients for the scales used iance of the individual aspects or items of health im- in the discriminant function analyses (N = 381). pairment and active learning and thus ignores singular- Discriminant function ities that might otherwise blur the picture. If, on the oth- er hand, there is no significant difference in the effects Scales I II of a combination of job demand and job control, no dis- Dedication 0.97 -0.19 crimination based on the reported health impairment and Personal competence 0.82 0.35 active learning should be possible. Vigor 0.76 -0.18 The procedure used yielded the following target Absorption 0.71 -0.09 groups for the discriminant analysis: low demands and Commitment 0.52 -0.24 low control (LD & LC), low demands and high control Cynicism -0.40 0.46 (LD & HC), high demands and low control (HD & LC), Exhaustion -0.25 0.77 and high demands and high control (HD & HC). After Health complaints -0.24 0.44 a median split on the measures of job demands and job 282 Scand J Work Environ Health 2001, vol 27, no 4 Discriminant function 2 Demerouti et al interpreted as follows: positive values on the first control, low (high) was considered the employees who discriminant function delineate an overrepresentation of scored under (above) the mean value of 0 after a z-trans- motivation and learning, and negative values on this dis- formation. criminant function indicate an underrepresentation or absence of motivation. Similarly, positive values on the second discriminant function stand for an overrepresen- Results tation of health problems, while negative values indi- The means, standard deviations, internal consistencies, cate the absence or underrepresentation of these prob- and the bivariate correlations of the study variables are lems. Specifically, LD & LC jobs were characterized by presented in Table 1. As can be seen from this table, all an absence of both health impairment and worker moti- the scales had good internal consistency (30), and the vation (because they were located in the quadrant where pattern of correlations was as expected. both discriminant functions are negative), while LD & The stepwise discriminant analysis revealed a sig- HC jobs were characterized by the presence of motiva- nificant separation of the four groups [Wilk’s λ=0.73, tion and an absence of health impairment. Furthermore, χ (9)=108.92, P<0.001]. More specifically, two discri- the distinctive characteristics of HD & LC jobs were minant functions were significant for an optimal dis- health impairment (positive values on the first discri- crimination between the four groups. The eigenvalue of minant function) and an absence of worker motivation the first discriminant function was 0.26 and the canoni- (negative values on the second discriminant function), cal correlation was 0.45. For the second discriminant whereas HD & HC jobs were characterized by both function, these values were 0.09 and 0.29, respectively. health impairment and motivation. As can be seen from the plot of the group centroids Finally, figure 3 shows the proportion of participants in figure 2, the first discriminant function clearly dis- classified correctly, as well as the proportion of false criminates between the presence and absence of job con- classifications. As can be seen, the best prediction was trol, while the second discriminant function separates made for job control when job demands were low, and the two levels of job demands. Overall, 42.1% of the the worst prediction was made for high demands when total sample could be correctly classified; this result in- job control was low. Furthermore, figure 3 shows that dicates that the classification by these discriminant func- the proportion of extremely wrong classifications was tions is obviously superior to a random assignment based relatively low (eg, predicted membership HD & HC on prior group membership probabilities, which would when the jobs were actually characterized by LD & LC). be only 25% of the sample (31). The plot of the group centroids showed that job control had been classified more accurately than job demands on the basis of the Discussion discriminating measures since the distances between the Our study shows that positive and negative work expe- groups with low and high control (independent of the riences are differently predictive of the four combina- level of demands) were larger than the distances be- tions of job demands and job control, as proposed by tween the groups with low and high demands. the demand–control model (1, 2). Our hypothesis that The standardized canonical coefficients for the the amount of demands and control could be predicted scales of health impairment and motivation in the dis- on the basis of reported health impairment and active criminant functions are displayed in table 2. The first learning was confirmed. By means of a discriminant canonical variable — discriminating for the presence or analysis, we were able to separate the four quadrants of absence of control — was characterized by features re- the demand–control model according to the employees’ lated to motivation and active learning, including dedi- pattern of response to questions about different indica- cation, professional efficacy, vigor, absorption, and tors of health impairment and active learning and not commitment. The second canonical variable — separat- according to that of unique indicators, as is the case with ing low job demands from high job demands — was the use of a multivariate analysis of variance. In addi- mainly characterized by health impairment, namely, by tion, for more than two groups, “discriminant analysis exhaustion, and by psychosomatic health complaints. is preferred because it is computationally more efficient Cynicism had almost equally high loadings on both dis- than logistic regression analysis” (32, p 335). In other criminant functions and thus made no contribution to words, our alternative methodology was successful in discrimination among the groups. Indeed, the results detecting the reported combinations of job demands and were identical after the elimination of cynicism. control on the basis of employees’ burnout and engage- Since positive values in a discriminant function ment. Each of the four combinations of demands and mean dominance of the respective canonical (or discri- control was differentially related to the experience of minant) variables and negative values represent an un- strain consequences or motivational outcomes. derrepresentation of these variables, figure 2 can be Scand J Work Environ Health 2001, vol 27, no 4 283 Demand – control model Predicted group membership Actual group membership Low demands & low control Low demands & high control High demands & low control Predicted group membership Low demands & low con- Low demands & high control High demands & low control High demands & high control High demands & high control 0% 20% 40% 60% 80% 100% Figure 3. Results of the classification procedure. The indicators that were responsible for the discrim- ogy) experienced few strain consequences but scored ination between low versus high job control were pro- relatively high on active learning as well. Furthermore, fessional efficacy, vigor, dedication, absorption, and employees in active jobs (high demands and high con- commitment. In other words, the experience of job con- trol) reported not only high motivation but also high lev- trol is probably responsible for the experience of moti- els of strain, a finding also not completely in line with vation and learning. Note, however that cynicism (ie, the prediction of the model, since one would expect ac- distancing from work) — which was seen as an indica- tive learning and average psychological strain in such tor of low motivation (33) — made no contribution to work environments. this discrimination. Thus cynicism is not only strongly Our study produced some remarkable findings. First, (and negatively) related to the conceptual similar con- our findings suggest that active jobs may not only lead structs, for example, dedication, but is also strongly (and to increased worker motivation and learning, but also positively) related to the other core burnout component to high strain. This simultaneous existence of both mo- of exhaustion. It therefore justifies the designation of tivation and health impairment in so-called active jobs burnout as an affective-motivational syndrome (ie, the requires a systematic observation in future studies. It existence of one component increases the possibility of may represent a specific feature of our population, or it the existence of the other component). On the other might mean that a high level of job demands is linked hand, the discrimination of low versus high job demands to health impairment with or without the additional was mainly based on employees’ levels of exhaustion availability of job control. Furthermore, it could also be and health complaints. It seems, in other words, that job that, in this group, job demands were in fact too high. demands were the factor most clearly responsible for the If this were the case, much of the energy mobilized by development of health impairment. job demands could not be translated into effective prob- However, the combination of both discriminant lem solving (ie, job control), the consequence being functions gives a more accurate insight into the relation- much residual strain (2). This possibility is also consist- ship between demands and control on one hand and pos- ent with the ideas of Warr’s Vitamin Model (34), which itive and negative experiences at work on the other hand. assumes that neither too many nor too few demands are The employees who reported low demands and low con- good for employees. trol reported both low active learning and little strain, A second remarkable finding is that low-strain jobs which is in agreement with Karasek’s (1, 2) predictions. may not only lead to reduced health impairment, but also In addition, the job strain hypothesis, referring to a work to the experience of active learning. This finding also situation of high demands and low control was also con- contradicts the prediction of the demand–control model firmed. The employees who worked under such condi- and may be due to method artifacts (motivated employ- tions experienced pronounced strain consequences and ees assessing their jobs as less demanding), or it may low active learning. However, inconsistent with predic- represent another point of view in the interpretation of tions from the demand–control model, employees who job demands (ie, demands are a troublesome aspect of reported low demands and high job control (ie, in so- the job and require additional effort from the employee called low strain jobs according to Karasek’s terminol- in order to be fulfilled). According to our findings, high 284 Scand J Work Environ Health 2001, vol 27, no 4 Demerouti et al job demands are linked to an overrepresentation of strain inflated the relationships of the variables, since a con- (independent of job control), whereas high job control firmatory factor analytic approach, using only a single is related to an overrepresentation of active learning (in- factor method (38), showed that a method factor could dependent of the level of demands). In other words, job not adequately explain the variance of the study varia- control seems to be an insufficient buffer that could tem- bles. In a similar vein, the two-factor confirmatory fac- per the effects of job demands, or job demands seem to tor analytic approach, in which one factor was opera- have an overwhelming character for employees. Con- tionalized by all positively phrased scales (job control, sequently, the results of our study raise the necessity of vigor, dedication, absorption, commitment and profes- additional investigation into the specific mechanisms of sional efficacy) and the second factor was operational- how high job demands must be to become a challenge ized by all negatively phrased scales (job demands, ex- (and further lead to motivation and active behavior) and haustion, cynicism, and health complaints), revealed that from what level they are overwhelming (and conse- response bias due to item wording was again insufficient quently result in health impairment). in explaining the variance of the variables under study. These findings generally support the assumption of An advantage of our study was the inclusion of both different ill-health indicators for the distinct combina- positive and negative outcomes of job (re)design with tions of demands and control. Although the discriminat- respect to the individual, as recommended by Karasek ing solution was not completely accurate, the correct (2). The results of our study show that, to obtain better classifications had the highest frequency, and the ex- insight into the postulated effects of the demand–con- tremely wrong classifications had the lowest frequen- trol model and to find support for the postulated (inter- cy. More importantly, the combinations of demands and action) effects (15), the inclusion of both positive and control that were on the active learning diagonal, and negative outcomes is necessary. clearly represent the interaction effect (LD & LC and In summary, our methodology appeared to be suc- HD & HC), could be efficiently discriminated from each cessful in investigating the importance of combinations other. The same applies for the strain diagonal (LD & between job demands and job control in predicting HC and HD & LC). However, the resulting relationships health impairment and active learning or motivation. between job demands and job control on one hand and Health impairment differentiated between low and high positive and negative outcomes on the other do not com- demanding jobs, whereas active learning differentiated pletely confirm the expectations of the demand–control between jobs allowing low and high job control. Nev- model, but they do confirm the job demands–resources ertheless, the specific hypotheses of Karasek (1) that are model (16, 17). The job demands–resources model states inherent in the demand–control model were only that excessive job demands are primarily related to neg- partly confirmed, since the low-strain jobs were filled ative stress reactions, like exhaustion, whereas job re- by employees without health impairments but with mo- sources are mainly related to motivational outcomes, tivation, whereas the active jobs included employees like engagement and professional efficacy. Recent stud- with motivation but also reporting health complaints. ies with this model have clearly shown that the relation- Future studies should therefore set out the interplay of ships between job demands and motivational indicators, the two core assumptions of the demand–control model as well as between job resources and strain are margin- in combination with longitudinal research designs al (16, 17). and focus more explicitly on the motivation or active A weakness of our study is that both job character- learning hypothesis. istics and outcomes were measured with a cross-section- al, self-report questionnaire. In principle, this method- ology throws the causality hypothesis into a questiona- Acknowledgments ble state. However, several longitudinal studies in this This research is part of the concerted research action on domain have shown that job characteristics, like job de- “Fatigue at Work” granted by The Netherlands Organi- mands and job control have causal predominant relation- zation for Scientific Research (NWO), 580-02-202. ships with health outcomes in such a way that the out- We thank Michiel Baldal and Marco van Putten for comes tend to occur after job perceptions, rather than their help with data collection. vice versa (35). As far as the self-report measures are concerned, the job holder seems to be the most impor- tant source to offer information regarding his or her References unique job position (36). Other, “objective” methods, such as observers’ ratings, appear to be good alterna- 1. Karasek RA. Job demands, job decision latitude and mental tives, but they also suffer from problems, including ob- strain: implications for job redesign. Adm Sci Q 1979;24: server bias and halo and stereotyping effects (37). Fur- 285—308. 2. Karasek RA. Demand/control model: a social, emotional, thermore, common method variance should not have Scand J Work Environ Health 2001, vol 27, no 4 285 Demand – control model Maslach burnout inventory — general survey. In: Maslach C, and psychological approach to stress risk and active behav- iour development. In: Stellman JM, editor. Encyclopaedia of Jackson SE, Leiter MP, editors. Maslach burnout inventory occupational health and safety. Geneva: International Labour [manual]. 3rd ed. Palo Alto (CA): Consulting Psychologists Office, 1998:34.6—34.14. Press, 1996. 3. De Jonge J, Kompier MAJ. A critical examination of the 22. Maslach C, Schaufeli WB, Leiter MP. Job burnout. Annu demands-control-support model from a work psychological Rev Psychol 2001;52:397—422. perspective. Int J Stress Manage 1997;4:235—58. 23. Maslach C, Leiter MP. The truth about burnout: how organi- 4. Van der Doef M, Maes S. The job demand-control(-support) zations cause personal stress and what to do about it. San model and physical health outcomes: a review of the strain Francisco (CA): Jossey-Bass, 1997. and buffer hypotheses. Psychol Health 1998;13:909—36. 24. Bakker AB, Schaufeli WB. De Utrechtse bevlogenheidss- chaal: UBES [The Utrecht engagement scale: UBES]. Utrecht 5. Van der Doef M, Maes S. The job demand-control(-support) model and psychological well-being: a review of 20 years (The Netherlands): Utrecht University, Department of Social empirical research. Work Stress 1999;13:87—114. and Organizational Psychology, 1999. 25. Levi L, Bartley M, Marmot M, Theorell T, Karasek RA, 6. Karasek RA, Theorell T. Healthy work: stress, productivity and the reconstruction of working life. New York (NY): Siegrist J, et al. Stressors at the workplace: theoretical mod- Basic Books, 1990. els. In: Schnall PL, Belkic ´ K, Landsbergis P, Baker D, edi- 7. Kristensen TS. The demand-control-support model: method- tors. The workplace and cardiovascular disease. Philadelphia ological challenges for future research. Stress Med 1995; (PA): Hanley & Belfus, 2000:69—106. Occupational medi- 11:17—26. cine: state of the art reviews, vol 15. 8. Kristensen TS. Job stress and cardiovascular disease: a theoret- 26. Karasek RA, Brisson C, Kawakami N, Houtman I, Bongers P, Amick B. The Job content questionnaire (JCQ): an instru- ical critical review. J Occup Health Psychol 1996;1:246—60. 9. Schnall PL, Landsbergis PA, Baker D. Job strain and cardio- ment for internationally comparative assessments of psycho- vascular disease. Annu Rev Public Health 1994;15:381—411. social job characteristics. J Occup Health Psychol 1998;3:322—55. 10. Theorell T, Karasek RA. Current issues relating to psycho- logical job strain and cardiovascular disease research. J Oc- 27. Schaufeli WB, Van Dierendonck D. Maslach burnout inven- cup Health Psychol 1996;1:9—26. tory: Nederlandse versie [Maslach burnout inventory: Dutch 11. De Jonge J, Dollard MF, Dormann C, Le Blanc PM, Hout- version]. Lisse (The Netherlands): Swets Zeitlinger, 2000. man ILD. The demand-control model: specific demands, spe- 28. Joosten J, Drop MJ. De betrouwbaarheid en vergelijkbaar- cific control, and well-defined groups. Int J Stress Manage heid van drie versies van de VOEG [Reliability and compara- 2000;7, 269—87. bility of three versions of the VOEG]. Gezondheid en Sa- menleving 1987;8:251—65. 12. Karasek RA. Job socialization and job strain: the implica- tions of 2 related psychosocial mechanisms for job design. 29. Mowday RT, Steers RM, Porter LW. The measurement of In: Gardell B, Johansson G, editors. Man and working life. organizational commitment. J Vocat Behav 1979;14:224—47. Chichester (UK): Wiley, 1981:75—94. 30. Cortina JM. What is coefficient alpha? An examination of 13. Landsbergis PA, Schnall PL, Deitz D, Friedman R, Pickering theory and applications. J Appl Psychol 1993;78:98—104. T. The patterning of psychological attributes and distress by 31. Tabachnik BG, Fidell LS. Using multivariate statistics. 2nd ‘job strain’ and social support in a sample of working men. J ed. New York (NY): Harper Collins Publishers, 1989. Behav Med 1992;15:379—405. 32. Sharma S. Applied multivariate techniques. New York (NY): 14. Meijman TF, Ulenbelt P, Lumens ME, Herber RF. Behavio- Wiley, 1996. ral determinants of occupational exposure to chemical agents. 33. Schaufeli WB, Enzmann D. The burnout companion to study and practice: a critical analysis. London: Taylor & Francis, J Occup Health Psychol 1996;1:85—91. 15. De Rijk AE, Le Blanc PM, Schaufeli WB, De Jonge J. Active 1998. coping and need for control as moderators of the job de- 34. Warr PB. Work, unemployment and mental health. Oxford: mand-control model: effects on burnout. J Occup Organ Psy- Oxford University Press, 1987. chol 1998;71:1—18. 35. Buunk BP, de Jonge J, Ybema JF, De Wolff CJ. Psychoso- 16. Demerouti E, Bakker AB, Nachreiner F, Schaufeli WB. A cial aspects of occupational stress. In: Drenth PJD, Thierry model of burnout and life satisfaction among nurses. J Adv H, De Wolff CJ, editors, Handbook of industrial and organi- Nurs 2000;32:454—64. zational psychology: vol 2. 2nd ed. Brighton: Psychology 17. Demerouti E, Bakker AB, Nachreiner F, Schaufeli WB. The Press, 1998:145—82. job demands–resources model of burnout. J Applied Psy- 36. Frese M, Zapf D. On the importance of the objective environ- ment in stress and attribution theory. J Organ Behav chol. In press. 18. Taris TW, Schreurs PJ, Schaufeli WB. Construct validity of 1999;20:761—6. the Maslach burnout inventory — general survey: a two- 37. De Jonge J, Van Breukelen GJP, Landeweerd JA, Nijhuis FJN. Comparing group and individual level assessments of sample examination of its factor structure and correlates. Work Stress 1999;13:223—37. job characteristics in testing the job demand-control model: a 19. Schutte N, Toppinnen S, Kalimo R, Schaufeli WB. The fac- multilevel approach. Hum Relat 1999;52:95—122. torial validity of the Maslach burnout inventory — general 38. McFarlin DB, Sweeney PD. Distributive and procedural jus- survey (MBI-GS) across nations and occupations. J Occup tice as predictors of satisfaction with personal and organiza- Organ Psychol 2000;73:53—66. tional outcomes. Acad Manage J 1992;35:626—37. 20. Maslach C, Jackson SE. Maslach burnout inventory. 2nd ed. Palo Alto (CA): Consulting Psychologists Press, 1986. 21. Schaufeli WB, Leiter MP, Maslach C, Jackson SE. The Received for publication: 7 August 2000 286 Scand J Work Environ Health 2001, vol 27, no 4 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Scandinavian Journal of Work, Environment & Health Unpaywall

Burnout and engagement at work as a function of demands and control

Scandinavian Journal of Work, Environment & HealthAug 1, 2001

Loading next page...
 
/lp/unpaywall/burnout-and-engagement-at-work-as-a-function-of-demands-and-control-u1opaeW4um

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Unpaywall
ISSN
0355-3140
DOI
10.5271/sjweh.615
Publisher site
See Article on Publisher Site

Abstract

Downloaded from www.sjweh.fi on October 30, 2021 Original article Scand J Work Environ Health 2001;27(4):279-286 doi:10.5271/sjweh.615 Burnout and engagement at work as a function of demands and control by Demerouti E, Bakker AB, de Jonge J, Janssen PPM, Schaufeli WB Affiliation: University of Nijmegen, Department of Work & Organizational Psychology, Postbox Office 9104, 6500 HE Nijmegen, Netherlands. E.Demerouti@psych.kun.nl The following articles refer to this text: 2006;32(5):339-348; 2008;34(5):345-355 Key terms: absorption; burnout; burnout at work; dedication; demandcontrol model; engagement; engagement at work; vigor This article in PubMed: www.ncbi.nlm.nih.gov/pubmed/11560342 This work is licensed under a Creative Commons Attribution 4.0 International License. Print ISSN: 0355-3140 Electronic ISSN: 1795-990X Demerouti et al Scand J Work Environ Health 2001;27(4):279—286 1, 2 3 3 4 by Evangelia Demerouti, PhD, Arnold B Bakker, PhD, Jan de Jonge, PhD, Peter PM Janssen, PhD, Wilmar B Schaufeli, PhD Demerouti E, Bakker AB, de Jonge J, Janssen PPM, Schaufeli WB. Burnout and engagement at work as a function of demands and control. Scand J Work Environ Health 2001;27(4):279—286. Objectives The present study was designed to test the demand-control model using indicators of both health impairment and active learning or motivation. Methods A total of 381 insurance company employees participated in the study. Discriminant analysis was used to examine the relationship between job demands and job control on one hand and health impairment and active learning on the other. Results The amount of demands and control could be predicted on the basis of employees’ perceived health impairment (exhaustion and health complaints) and active learning (engagement and commitment). Each of the four combinations of demand and control differentially affected the perception of strain or active learning. Job demands were the most clearly related to health impairment, whereas job control was the most clearly associated with active learning. Conclusions These findings partly contradict the demand–control model, especially with respect to the validity of the interaction between demand and control. Job demands and job control seem to initiate two essentially independent processes, and this occurrence is consistent with the recently proposed job demands–resources model. Keywords absorption, burnout, dedication, demand–control model, engagement, vigor The demand–control model (1, 2) is a theoretical mod- The demand-control model is based on two central el that attempts to provide insight into the relationships assumptions, as reflected by diagonals A and B in fig- between psychosocial work characteristics on one hand ure 1. The first assumption (diagonal A) is that psycho- and health and motivation on the other. Partly because logical strain (such as chronic fatigue, anxiety, and car- of its elegant simplicity, the model has stimulated sev- diovascular complaints) is particularly caused by the eral scientific studies (3, 4, 5). It looks for the determi- combination of high psychological demands and low de- nants of workers’ health and motivation in the interac- cision latitude (quadrant 1), while the opposite, lack of tion of two psychosocial job characteristics, namely, the strain, can be found in a job with low psychological de- “psychological demands” corresponding to psycholog- mands and much decision latitude (quadrant 3). The ical stressors present in the work environment (eg, high second important assumption (diagonal B) is that time pressure, difficult and mentally taxing work) and work motivation, as well as learning and development “decision latitude” (or “job control”) comprising the opportunities, occur if job demands are high (but not worker’s authority to make decisions on the job (“deci- overwhelming) and decision latitude is high (quadrant sion authority”) and the breadth of skills used by the 2). The opposite type of work situation occurs in a job worker on the job [“skill discretion” (6)] (figure 1). in which neither job demands nor decision latitude is Carl von Ossietzky Universität Oldenburg, Department of Work & Organizational Psychology, Oldenburg, Germany. University of Nijmegen, Department of Work & Organizational Psychology, Nijmegen, The Netherlands. Utrecht University, Department of Social and Organizational Psychology, Utrecht, The Netherlands. University of Maastricht, Department of Health Studies, Maastricht, The Netherlands. Reprint requests to: Dr Evangelia Demerouti, University of Nijmegen, Department of Work & Organizational Psychology, Postbox Office 9104, 6500 HE Nijmegen, The Netherlands. [E-mail: E.Demerouti@psych.kun.nl] Scand J Work Environ Health 2001, vol 27, no 4 279 Demand – control model Psychological demands B diagonal ently of the level of job demands. Job resources repre- sent a broader category of “positive” features of the Low High work environment than job control in the demand–con- trol model. Applied to the demand–control model, this statement would mean that the low strain quadrant is Low strain Active High 3 not only related to diminished health impairment, but also to motivation and active learning. The aim of our study was thus to test the demand– control model with special attention to the active-pas- sive hypothesis and to integrate the strain and active learning hypotheses. Specifically, we attempted to pre- Passive High strain dict employees’ levels of job demands and job control Low 4 1 using stress reactions (burnout, health complaints), but also indicators of active learning (engagement, commit- ment). In addition, we used an alternative statistical technique, which allows for a different perspective of the predictions of the demand–control model and an A diagonal examination of the postulated interaction. Figure 1. The demand–control model [source: Karasek (2)]. Burnout and engagement The concept of burnout that was initially relevant only to working with people has recently been expanded to- very pronounced (quadrant 4). This “passive” work sit- wards all other professions and occupational groups (18, uation is characterized by a decrease in work activities 19). Based on the original instrument to measure hu- and “negative learning” — a gradual loss of acquired man service-related burnout [Maslach Burnout Inven- skills. tory (20)] a new scale [Maslach Burnout Inventory— Recent reviews of the model contend that the de- General Survey (21)] has been developed for use out- mand–control model is appropriate for further investi- side the human services. The dimensions of the latter gation since it can predict health and, to a less extent, (ie, exhaustion, cynicism, professional efficacy) paral- motivational and productivity outcomes (3—5, 7—9,). lel those of the original inventory (ie, emotional exhaus- However, the “active-passive” dimension of the model tion, depersonalization, personal accomplishment) in has been underutilized in organizational research (10). the sense that they are more generic and do not refer to Only a few studies provide some evidence for the ac- the people one is working with. Exhaustion is measured tive-passive dimension of the model (11—14). Studies by items that refer to fatigue, but do not make direct measuring both strain and motivation or active learning reference to people as the source of such feelings. The find stronger support for the model than do those re- items assessing cynicism reflect indifference or a dis- stricted to strain consequences (15). tant attitude towards work in general, not necessarily One possible reason for the lack of attention to the towards people. Finally, professional efficacy encom- active-passive hypothesis may be that most of the inter- passes both social and nonsocial aspects of occupational est in the model originated in the field of epidemiolo- accomplishments. Psychometric research with the gy, with cardiovascular diseases as the main form of Maslach Burnout Inventory—General Survey demon- strain (7, 8). Another reason may be that the B diagonal strated that the three-factor structure is invariant across [“motivation for learning new behavior patterns” (6)] is occupational groups (19). hard to operationalize as an independent psychological A recent development in burnout research is the shift construct, since it may be conceptually confounded by towards its opposite, engagement (22). In the view of the job characteristic skill discretion (3). Maslach & Leiter (23), workers have a sense of ener- According to Karasek (1), strain and learning are getic and effective connection with their work activi- two, more or less, independent phenomena, and conse- ties (energy, involvement), and they see themselves as quently no attempts have been made to integrate the two able to deal with the demands of their job (professional assumptions in the model. When, for example, job con- efficacy). Bakker & Schaufeli (24) partly agree with trol is high and job demands are low, we should expect their description and defined engagement as a positive, low strain, but nothing is explicitly said about the level fulfilling, work-related state of mind that is character- of motivation. Demerouti et al (16, 17) have argued and ized by vigor, dedication, and absorption. Rather than shown that, when job resources are high, we should ex- a momentary and specific state, engagement refers to a pect motivation in the form of engagement, independ- 280 Scand J Work Environ Health 2001, vol 27, no 4 Decision latitude (control) Demerouti et al more persistent and pervasive affective-motivational the respondents were men (66%). The mean age of the state that is not focused on any particular object, event, group was 40 (SD 10.4) years, and the mean organiza- individual, or behavior. Vigor is characterized by high tional tenure was 12.5 (SD 10.5) years. Twelve percent levels of energy and mental resilience while working, of the sample had a managerial position. Most of the the willingness to invest effort in one’s work, and per- participants had a full-time job (64%). The group was sistence even in the face of difficulties. Dedication is recruited at their job sites after an informative meeting characterized by a sense of significance, enthusiasm, with the personnel department and management. All 624 inspiration, pride, and challenge. Absorption is charac- employees received an informative letter about the study terized by being fully concentrated and engrossed in from the management team, together with the question- one’s work, whereby time passes quickly. Bakker & naire and a return envelope. The confidentiality and ano- Schaufeli (24) found evidence for the psychometric nymity of the data were emphasized. The participants were quality (ie, internal consistency, factorial validity) of the kindly requested to fill out the questionnaire at home and engagement construct and the independence of the burn- to place it in a special box in their department. out and engagement dimensions. Measures Present study In this study, psychological job demands refer to the Our study was designed to test the job strain and active extent to which the work pace is high and the availabil- learning hypotheses generated by the demand–control ity of sufficient time to execute the required work. Job model (1, 2). Insurance company employees were cho- demands were measured with five items of the Dutch sen for this purpose. The sample was suitable for test- version of the Job Content Questionnaire (26). The four ing the model for several reasons. First, because of the response categories ranged from 1 “never” to 4 “always”. different types of tasks and specialties, insurance em- [Example item: “Do you have a lot of work to do?”] ployees are a relatively heterogeneous group as far as Job control was conceived as the degree to which their work characteristics are concerned, which implies employees have the freedom to make decisions regard- that there is considerable variance in their job demand ing the work assignment and their work methods. Nine and control. Second, the group chosen had relatively items of the Dutch version of the Job Content Question- long job tenure. Since the demand–control model refers naire (26) were used. The response format was identi- to chronic stressors (2, 25), the effect of such stressors cal to that used for job demands. [Example item: “Do is most obvious for job holders with long experience in you participate in decisions regarding your work?”] their current job. Burnout was measured with the Dutch version (27) A special feature of the study was not only the meas- of the Maslach Burnout Inventory—General Survey urement of psychological strain but also that of indica- (21). The instrument consists of 16 items that measure tors of active learning and motivation. As strain and ac- the three components of burnout (exhaustion, cynicism tive learning are not one-dimensional, we operational- and professional efficacy). The items were scored on a ized them through the use of several indicators. For that 7-point scale (0 = never, 6 = always). [Example items: purpose, burnout and engagement were included. How- “I feel used up at the end of the workday” (exhaustion), ever, burnout is a complex syndrome that consists of an “I have become less enthusiastic about my work” (cyn- orthodox stress reaction (exhaustion), as well as of an icism), and “In my opinion, I am good at my work” (pro- attitudinal component (cynicism and lack of efficacy) fessional efficacy).] that is closely linked to motivation. In other words, when Psychosomatic health complaints refer to the number workers are cynical and do not feel efficacious, they of psychosomatic symptoms that a person experienced may not be willing to learn actively. In addition psy- during the year preceding the study. Health complaints chosomatic complaints on one hand and engagement (in- were measured with 13 items of a well-validated ques- cluding vigor, dedication, and absorption) and commit- tionnaire on health complaints (28). Employees could ment on the other were included as indices of strain and respond using a 4-point scale (1 = seldom or never, 4 = active learning, respectively. very often). [Example items: “Do you have unpleasant pains on the breast?” and “Do you have headaches?”] Engagement was assessed with the Utrecht work engagement scale (24). The scale includes 15 items, Subjects and methods which measure the three components of engagement (vigor, dedication and absorption). All the items were Sample and procedure rated on a 7-point scale ranging from 0 “never” to 6 Altogether, 381 employees from an insurance company “always”. [Example items: “During my work, I feel participated in the study (response rate = 61%). Most of vital and strong” (vigor), “I am enthusiastic about my Scand J Work Environ Health 2001, vol 27, no 4 281 Demand – control model Table 1. Means (M), standard deviations (SD), internal consistencies (α) and Pearson correlations of the variables included in this study (N = 381). Variable M SD α 1 2 34 56 7 8 9 1. Job demands 2.69 0.43 0.85 2. Job control 2.87 0.35 0.75 0.04 3. Exhaustion 1.79 1.08 0.89 0.25** -0.20** 4. Cynicism 1.49 0.94 0.74 0.11* -0.29** 0.60** 5. Personal competence 4.02 0.85 0.78 0.15** 0.44** -0.18** -0.30** 6. Health complaints 1.44 0.37 0.86 0.13** -0.23** 0.56** 0.41** -0.17** 7. Vigor 3.92 0.88 0.86 0.08 0.40** -0.41** -0.44** 0.65** -0.37** 8. Dedication 4.19 0.96 0.90 0.05 0.53** -0.29** -0.50** 0.69** -0.28** 0.76** 9. Absorption 3.56 0.98 0.80 0.07 0.42** -0.16** -0.32** 0.52** -0.18*** 0.73** 0.75** 10. Commitment 3.61 0.60 0.80 -0.05 0.30** -0.36** -0.44** 0.40** -0.24** 0.49** 0.59** 0.45** * P < 0.05, ** P < 0.01. 1.00 job” (dedication), and “Time flies when I am working” (absorption).] Commitment refers to the relationship of employees to the organization in which they work. It is measured 0.50 with the 5 items of the affective commitment scale de- HD & LC HD & LC veloped by Mowday et al (29). [Example item: “I tell HD & HC HD & HC my friends and family that my organization is a pleas- ant organization to work for” (1 = totally disagree, 5 = 0.00 totally agree).] LD & LC LD & LC LD & H C LD & HC Data analysis strategy -0.50 The strategy followed in the present study was to look for different generalized patterns of health impairment and active learning to indicate the differences between -1.00 the four combinations of job demands and job control –1.00 –0.50 0.00 0.50 1.00 (ie, the four quadrants in figure 1). If these four combi- Discriminant function 1 nations are related to health impairment and active learning, it should be possible to separate the four quad- Figure 2. Group centroids for the four combinations of demands and rants on the basis of the response patterns to questions control in a two-dimensional space. LD & LC = low demands and low about health impairment (exhaustion, health complaints) control; LD & HC = low demands and high control; HD & LC = high and active learning (engagement, professional efficacy, demands and low control; HD & HC = high demands and high control. cynicism, commitment) by means of a discriminant anal- ysis. This form of statistical treatment of the problem offers the advantage of making use of the common var- Table 2. Standardized canonical coefficients for the scales used iance of the individual aspects or items of health im- in the discriminant function analyses (N = 381). pairment and active learning and thus ignores singular- Discriminant function ities that might otherwise blur the picture. If, on the oth- er hand, there is no significant difference in the effects Scales I II of a combination of job demand and job control, no dis- Dedication 0.97 -0.19 crimination based on the reported health impairment and Personal competence 0.82 0.35 active learning should be possible. Vigor 0.76 -0.18 The procedure used yielded the following target Absorption 0.71 -0.09 groups for the discriminant analysis: low demands and Commitment 0.52 -0.24 low control (LD & LC), low demands and high control Cynicism -0.40 0.46 (LD & HC), high demands and low control (HD & LC), Exhaustion -0.25 0.77 and high demands and high control (HD & HC). After Health complaints -0.24 0.44 a median split on the measures of job demands and job 282 Scand J Work Environ Health 2001, vol 27, no 4 Discriminant function 2 Demerouti et al interpreted as follows: positive values on the first control, low (high) was considered the employees who discriminant function delineate an overrepresentation of scored under (above) the mean value of 0 after a z-trans- motivation and learning, and negative values on this dis- formation. criminant function indicate an underrepresentation or absence of motivation. Similarly, positive values on the second discriminant function stand for an overrepresen- Results tation of health problems, while negative values indi- The means, standard deviations, internal consistencies, cate the absence or underrepresentation of these prob- and the bivariate correlations of the study variables are lems. Specifically, LD & LC jobs were characterized by presented in Table 1. As can be seen from this table, all an absence of both health impairment and worker moti- the scales had good internal consistency (30), and the vation (because they were located in the quadrant where pattern of correlations was as expected. both discriminant functions are negative), while LD & The stepwise discriminant analysis revealed a sig- HC jobs were characterized by the presence of motiva- nificant separation of the four groups [Wilk’s λ=0.73, tion and an absence of health impairment. Furthermore, χ (9)=108.92, P<0.001]. More specifically, two discri- the distinctive characteristics of HD & LC jobs were minant functions were significant for an optimal dis- health impairment (positive values on the first discri- crimination between the four groups. The eigenvalue of minant function) and an absence of worker motivation the first discriminant function was 0.26 and the canoni- (negative values on the second discriminant function), cal correlation was 0.45. For the second discriminant whereas HD & HC jobs were characterized by both function, these values were 0.09 and 0.29, respectively. health impairment and motivation. As can be seen from the plot of the group centroids Finally, figure 3 shows the proportion of participants in figure 2, the first discriminant function clearly dis- classified correctly, as well as the proportion of false criminates between the presence and absence of job con- classifications. As can be seen, the best prediction was trol, while the second discriminant function separates made for job control when job demands were low, and the two levels of job demands. Overall, 42.1% of the the worst prediction was made for high demands when total sample could be correctly classified; this result in- job control was low. Furthermore, figure 3 shows that dicates that the classification by these discriminant func- the proportion of extremely wrong classifications was tions is obviously superior to a random assignment based relatively low (eg, predicted membership HD & HC on prior group membership probabilities, which would when the jobs were actually characterized by LD & LC). be only 25% of the sample (31). The plot of the group centroids showed that job control had been classified more accurately than job demands on the basis of the Discussion discriminating measures since the distances between the Our study shows that positive and negative work expe- groups with low and high control (independent of the riences are differently predictive of the four combina- level of demands) were larger than the distances be- tions of job demands and job control, as proposed by tween the groups with low and high demands. the demand–control model (1, 2). Our hypothesis that The standardized canonical coefficients for the the amount of demands and control could be predicted scales of health impairment and motivation in the dis- on the basis of reported health impairment and active criminant functions are displayed in table 2. The first learning was confirmed. By means of a discriminant canonical variable — discriminating for the presence or analysis, we were able to separate the four quadrants of absence of control — was characterized by features re- the demand–control model according to the employees’ lated to motivation and active learning, including dedi- pattern of response to questions about different indica- cation, professional efficacy, vigor, absorption, and tors of health impairment and active learning and not commitment. The second canonical variable — separat- according to that of unique indicators, as is the case with ing low job demands from high job demands — was the use of a multivariate analysis of variance. In addi- mainly characterized by health impairment, namely, by tion, for more than two groups, “discriminant analysis exhaustion, and by psychosomatic health complaints. is preferred because it is computationally more efficient Cynicism had almost equally high loadings on both dis- than logistic regression analysis” (32, p 335). In other criminant functions and thus made no contribution to words, our alternative methodology was successful in discrimination among the groups. Indeed, the results detecting the reported combinations of job demands and were identical after the elimination of cynicism. control on the basis of employees’ burnout and engage- Since positive values in a discriminant function ment. Each of the four combinations of demands and mean dominance of the respective canonical (or discri- control was differentially related to the experience of minant) variables and negative values represent an un- strain consequences or motivational outcomes. derrepresentation of these variables, figure 2 can be Scand J Work Environ Health 2001, vol 27, no 4 283 Demand – control model Predicted group membership Actual group membership Low demands & low control Low demands & high control High demands & low control Predicted group membership Low demands & low con- Low demands & high control High demands & low control High demands & high control High demands & high control 0% 20% 40% 60% 80% 100% Figure 3. Results of the classification procedure. The indicators that were responsible for the discrim- ogy) experienced few strain consequences but scored ination between low versus high job control were pro- relatively high on active learning as well. Furthermore, fessional efficacy, vigor, dedication, absorption, and employees in active jobs (high demands and high con- commitment. In other words, the experience of job con- trol) reported not only high motivation but also high lev- trol is probably responsible for the experience of moti- els of strain, a finding also not completely in line with vation and learning. Note, however that cynicism (ie, the prediction of the model, since one would expect ac- distancing from work) — which was seen as an indica- tive learning and average psychological strain in such tor of low motivation (33) — made no contribution to work environments. this discrimination. Thus cynicism is not only strongly Our study produced some remarkable findings. First, (and negatively) related to the conceptual similar con- our findings suggest that active jobs may not only lead structs, for example, dedication, but is also strongly (and to increased worker motivation and learning, but also positively) related to the other core burnout component to high strain. This simultaneous existence of both mo- of exhaustion. It therefore justifies the designation of tivation and health impairment in so-called active jobs burnout as an affective-motivational syndrome (ie, the requires a systematic observation in future studies. It existence of one component increases the possibility of may represent a specific feature of our population, or it the existence of the other component). On the other might mean that a high level of job demands is linked hand, the discrimination of low versus high job demands to health impairment with or without the additional was mainly based on employees’ levels of exhaustion availability of job control. Furthermore, it could also be and health complaints. It seems, in other words, that job that, in this group, job demands were in fact too high. demands were the factor most clearly responsible for the If this were the case, much of the energy mobilized by development of health impairment. job demands could not be translated into effective prob- However, the combination of both discriminant lem solving (ie, job control), the consequence being functions gives a more accurate insight into the relation- much residual strain (2). This possibility is also consist- ship between demands and control on one hand and pos- ent with the ideas of Warr’s Vitamin Model (34), which itive and negative experiences at work on the other hand. assumes that neither too many nor too few demands are The employees who reported low demands and low con- good for employees. trol reported both low active learning and little strain, A second remarkable finding is that low-strain jobs which is in agreement with Karasek’s (1, 2) predictions. may not only lead to reduced health impairment, but also In addition, the job strain hypothesis, referring to a work to the experience of active learning. This finding also situation of high demands and low control was also con- contradicts the prediction of the demand–control model firmed. The employees who worked under such condi- and may be due to method artifacts (motivated employ- tions experienced pronounced strain consequences and ees assessing their jobs as less demanding), or it may low active learning. However, inconsistent with predic- represent another point of view in the interpretation of tions from the demand–control model, employees who job demands (ie, demands are a troublesome aspect of reported low demands and high job control (ie, in so- the job and require additional effort from the employee called low strain jobs according to Karasek’s terminol- in order to be fulfilled). According to our findings, high 284 Scand J Work Environ Health 2001, vol 27, no 4 Demerouti et al job demands are linked to an overrepresentation of strain inflated the relationships of the variables, since a con- (independent of job control), whereas high job control firmatory factor analytic approach, using only a single is related to an overrepresentation of active learning (in- factor method (38), showed that a method factor could dependent of the level of demands). In other words, job not adequately explain the variance of the study varia- control seems to be an insufficient buffer that could tem- bles. In a similar vein, the two-factor confirmatory fac- per the effects of job demands, or job demands seem to tor analytic approach, in which one factor was opera- have an overwhelming character for employees. Con- tionalized by all positively phrased scales (job control, sequently, the results of our study raise the necessity of vigor, dedication, absorption, commitment and profes- additional investigation into the specific mechanisms of sional efficacy) and the second factor was operational- how high job demands must be to become a challenge ized by all negatively phrased scales (job demands, ex- (and further lead to motivation and active behavior) and haustion, cynicism, and health complaints), revealed that from what level they are overwhelming (and conse- response bias due to item wording was again insufficient quently result in health impairment). in explaining the variance of the variables under study. These findings generally support the assumption of An advantage of our study was the inclusion of both different ill-health indicators for the distinct combina- positive and negative outcomes of job (re)design with tions of demands and control. Although the discriminat- respect to the individual, as recommended by Karasek ing solution was not completely accurate, the correct (2). The results of our study show that, to obtain better classifications had the highest frequency, and the ex- insight into the postulated effects of the demand–con- tremely wrong classifications had the lowest frequen- trol model and to find support for the postulated (inter- cy. More importantly, the combinations of demands and action) effects (15), the inclusion of both positive and control that were on the active learning diagonal, and negative outcomes is necessary. clearly represent the interaction effect (LD & LC and In summary, our methodology appeared to be suc- HD & HC), could be efficiently discriminated from each cessful in investigating the importance of combinations other. The same applies for the strain diagonal (LD & between job demands and job control in predicting HC and HD & LC). However, the resulting relationships health impairment and active learning or motivation. between job demands and job control on one hand and Health impairment differentiated between low and high positive and negative outcomes on the other do not com- demanding jobs, whereas active learning differentiated pletely confirm the expectations of the demand–control between jobs allowing low and high job control. Nev- model, but they do confirm the job demands–resources ertheless, the specific hypotheses of Karasek (1) that are model (16, 17). The job demands–resources model states inherent in the demand–control model were only that excessive job demands are primarily related to neg- partly confirmed, since the low-strain jobs were filled ative stress reactions, like exhaustion, whereas job re- by employees without health impairments but with mo- sources are mainly related to motivational outcomes, tivation, whereas the active jobs included employees like engagement and professional efficacy. Recent stud- with motivation but also reporting health complaints. ies with this model have clearly shown that the relation- Future studies should therefore set out the interplay of ships between job demands and motivational indicators, the two core assumptions of the demand–control model as well as between job resources and strain are margin- in combination with longitudinal research designs al (16, 17). and focus more explicitly on the motivation or active A weakness of our study is that both job character- learning hypothesis. istics and outcomes were measured with a cross-section- al, self-report questionnaire. In principle, this method- ology throws the causality hypothesis into a questiona- Acknowledgments ble state. However, several longitudinal studies in this This research is part of the concerted research action on domain have shown that job characteristics, like job de- “Fatigue at Work” granted by The Netherlands Organi- mands and job control have causal predominant relation- zation for Scientific Research (NWO), 580-02-202. ships with health outcomes in such a way that the out- We thank Michiel Baldal and Marco van Putten for comes tend to occur after job perceptions, rather than their help with data collection. vice versa (35). As far as the self-report measures are concerned, the job holder seems to be the most impor- tant source to offer information regarding his or her References unique job position (36). Other, “objective” methods, such as observers’ ratings, appear to be good alterna- 1. Karasek RA. Job demands, job decision latitude and mental tives, but they also suffer from problems, including ob- strain: implications for job redesign. Adm Sci Q 1979;24: server bias and halo and stereotyping effects (37). Fur- 285—308. 2. Karasek RA. Demand/control model: a social, emotional, thermore, common method variance should not have Scand J Work Environ Health 2001, vol 27, no 4 285 Demand – control model Maslach burnout inventory — general survey. In: Maslach C, and psychological approach to stress risk and active behav- iour development. In: Stellman JM, editor. Encyclopaedia of Jackson SE, Leiter MP, editors. Maslach burnout inventory occupational health and safety. Geneva: International Labour [manual]. 3rd ed. Palo Alto (CA): Consulting Psychologists Office, 1998:34.6—34.14. Press, 1996. 3. De Jonge J, Kompier MAJ. A critical examination of the 22. Maslach C, Schaufeli WB, Leiter MP. Job burnout. Annu demands-control-support model from a work psychological Rev Psychol 2001;52:397—422. perspective. Int J Stress Manage 1997;4:235—58. 23. Maslach C, Leiter MP. The truth about burnout: how organi- 4. Van der Doef M, Maes S. The job demand-control(-support) zations cause personal stress and what to do about it. San model and physical health outcomes: a review of the strain Francisco (CA): Jossey-Bass, 1997. and buffer hypotheses. Psychol Health 1998;13:909—36. 24. Bakker AB, Schaufeli WB. De Utrechtse bevlogenheidss- chaal: UBES [The Utrecht engagement scale: UBES]. Utrecht 5. Van der Doef M, Maes S. The job demand-control(-support) model and psychological well-being: a review of 20 years (The Netherlands): Utrecht University, Department of Social empirical research. Work Stress 1999;13:87—114. and Organizational Psychology, 1999. 25. Levi L, Bartley M, Marmot M, Theorell T, Karasek RA, 6. Karasek RA, Theorell T. Healthy work: stress, productivity and the reconstruction of working life. New York (NY): Siegrist J, et al. Stressors at the workplace: theoretical mod- Basic Books, 1990. els. In: Schnall PL, Belkic ´ K, Landsbergis P, Baker D, edi- 7. Kristensen TS. The demand-control-support model: method- tors. The workplace and cardiovascular disease. Philadelphia ological challenges for future research. Stress Med 1995; (PA): Hanley & Belfus, 2000:69—106. Occupational medi- 11:17—26. cine: state of the art reviews, vol 15. 8. Kristensen TS. Job stress and cardiovascular disease: a theoret- 26. Karasek RA, Brisson C, Kawakami N, Houtman I, Bongers P, Amick B. The Job content questionnaire (JCQ): an instru- ical critical review. J Occup Health Psychol 1996;1:246—60. 9. Schnall PL, Landsbergis PA, Baker D. Job strain and cardio- ment for internationally comparative assessments of psycho- vascular disease. Annu Rev Public Health 1994;15:381—411. social job characteristics. J Occup Health Psychol 1998;3:322—55. 10. Theorell T, Karasek RA. Current issues relating to psycho- logical job strain and cardiovascular disease research. J Oc- 27. Schaufeli WB, Van Dierendonck D. Maslach burnout inven- cup Health Psychol 1996;1:9—26. tory: Nederlandse versie [Maslach burnout inventory: Dutch 11. De Jonge J, Dollard MF, Dormann C, Le Blanc PM, Hout- version]. Lisse (The Netherlands): Swets Zeitlinger, 2000. man ILD. The demand-control model: specific demands, spe- 28. Joosten J, Drop MJ. De betrouwbaarheid en vergelijkbaar- cific control, and well-defined groups. Int J Stress Manage heid van drie versies van de VOEG [Reliability and compara- 2000;7, 269—87. bility of three versions of the VOEG]. Gezondheid en Sa- menleving 1987;8:251—65. 12. Karasek RA. Job socialization and job strain: the implica- tions of 2 related psychosocial mechanisms for job design. 29. Mowday RT, Steers RM, Porter LW. The measurement of In: Gardell B, Johansson G, editors. Man and working life. organizational commitment. J Vocat Behav 1979;14:224—47. Chichester (UK): Wiley, 1981:75—94. 30. Cortina JM. What is coefficient alpha? An examination of 13. Landsbergis PA, Schnall PL, Deitz D, Friedman R, Pickering theory and applications. J Appl Psychol 1993;78:98—104. T. The patterning of psychological attributes and distress by 31. Tabachnik BG, Fidell LS. Using multivariate statistics. 2nd ‘job strain’ and social support in a sample of working men. J ed. New York (NY): Harper Collins Publishers, 1989. Behav Med 1992;15:379—405. 32. Sharma S. Applied multivariate techniques. New York (NY): 14. Meijman TF, Ulenbelt P, Lumens ME, Herber RF. Behavio- Wiley, 1996. ral determinants of occupational exposure to chemical agents. 33. Schaufeli WB, Enzmann D. The burnout companion to study and practice: a critical analysis. London: Taylor & Francis, J Occup Health Psychol 1996;1:85—91. 15. De Rijk AE, Le Blanc PM, Schaufeli WB, De Jonge J. Active 1998. coping and need for control as moderators of the job de- 34. Warr PB. Work, unemployment and mental health. Oxford: mand-control model: effects on burnout. J Occup Organ Psy- Oxford University Press, 1987. chol 1998;71:1—18. 35. Buunk BP, de Jonge J, Ybema JF, De Wolff CJ. Psychoso- 16. Demerouti E, Bakker AB, Nachreiner F, Schaufeli WB. A cial aspects of occupational stress. In: Drenth PJD, Thierry model of burnout and life satisfaction among nurses. J Adv H, De Wolff CJ, editors, Handbook of industrial and organi- Nurs 2000;32:454—64. zational psychology: vol 2. 2nd ed. Brighton: Psychology 17. Demerouti E, Bakker AB, Nachreiner F, Schaufeli WB. The Press, 1998:145—82. job demands–resources model of burnout. J Applied Psy- 36. Frese M, Zapf D. On the importance of the objective environ- ment in stress and attribution theory. J Organ Behav chol. In press. 18. Taris TW, Schreurs PJ, Schaufeli WB. Construct validity of 1999;20:761—6. the Maslach burnout inventory — general survey: a two- 37. De Jonge J, Van Breukelen GJP, Landeweerd JA, Nijhuis FJN. Comparing group and individual level assessments of sample examination of its factor structure and correlates. Work Stress 1999;13:223—37. job characteristics in testing the job demand-control model: a 19. Schutte N, Toppinnen S, Kalimo R, Schaufeli WB. The fac- multilevel approach. Hum Relat 1999;52:95—122. torial validity of the Maslach burnout inventory — general 38. McFarlin DB, Sweeney PD. Distributive and procedural jus- survey (MBI-GS) across nations and occupations. J Occup tice as predictors of satisfaction with personal and organiza- Organ Psychol 2000;73:53—66. tional outcomes. Acad Manage J 1992;35:626—37. 20. Maslach C, Jackson SE. Maslach burnout inventory. 2nd ed. Palo Alto (CA): Consulting Psychologists Press, 1986. 21. Schaufeli WB, Leiter MP, Maslach C, Jackson SE. The Received for publication: 7 August 2000 286 Scand J Work Environ Health 2001, vol 27, no 4

Journal

Scandinavian Journal of Work, Environment & HealthUnpaywall

Published: Aug 1, 2001

There are no references for this article.