Job demands , job resources , burnout and work engagement of managers at a platinum mine in the North West Province

The objective of this study was to investigate the relationships between job demands, job resources, burnout, and engagement of management staff at a platinum mine in the North West Province. A cross-sectional survey design was used. The study population (N = 310) consisted of managers at the platinum mine. The Maslach Burnout Inventory – General Survey, Utrecht Work Engagement Scale, and the Job Demands-Resources Scale were administered. The results revealed that exhaustion was predicted by workload, job insecurity and a lack of resources, while cynicism was predicted by lack of organisational support and advancement opportunities. Vigour was predicted by organisational support. Dedication was predicted by organisational support and high workload. Engagement was predicted by organisational support.


Introduction
Globalisation and continued international pressure on organisations to perform better with fewer resources are reflected in the changing psychological contracts between employers and employees.Employees are expected to give more in terms of time, effort, skills and flexibility, whilst job security, career opportunities and lifetime employment are diminishing (Maslach, Schaufeli & Leiter, 2001).South Africa and its mining industry are not excluded from these pressures and impacts.The need to improve the country's productivity is reflected in its poor ranking (49 th of 60 countries) in the World Competitiveness Yearbook (http://www01.imd.ch/wcy).The South African mining industry produces 90% of the world's platinum-group metals among other minerals.Its contribution to the country's economic activity and productivity is beyond dispute (Gastrow, 2001).
The key differentiator of competitive advantage in the new world economy is the organisation's employees (Minervini, Meyer & Rourke, 2003).However, employees have to cope with increasing demands from various and diverse roles and organisational stakeholders, often with limited resources (Minervini et al., 2003).Monitoring and improving employee effectiveness in coping with multiple new demands, stimulating their growth and enhancing their wellbeing as well as their organisational performance.In this regard, burnout and engagement are specific research areas (Maslach et al., 2001).Ivancevich and Matteson (1999) believe that managers are responsible for the effectiveness of individuals, groups and organisations.DuBrin (1990) reports that managers who suffer from burnout, harm organisational effectiveness because they spread it to their subordinates.Burnout can thus be 'contagious' and perpetuates itself through the informal interactions on the job.Rothmann (2002) reports that burnout leads to low morale, job dissatisfaction, staff turnover and absenteeism, and that it can bring about deterioration in the quality of service rendered by staff.From these findings it can be deduced that managers can impact directly or indirectly on employee effectiveness and organisational outcomes such as turnover.
According to Jackson, Rothmann and Van de Vijver (2006), empirical studies have confirmed that burnout is related to health problems and turnover intentions, and that it mediates the relationship between job demands and health problems.Also, engagement mediates the relationship between job resources and turnover intentions.
The objectives of this study were, firstly, to determine the relationships between job demands, job resources, burnout, and engagement at a platinum mine in the North West Province where no research of this kind has been conducted before.

Burnout and engagement
Although burnout has originally been conceptualised in the context of the helping professions (Rothmann, 2002), it has recently expanded to all types of professions and occupational groups.Schaufeli and Enzmann (1998: 36) define burnout as 'a persistent, negative, work-related state of mind in normal individuals that is primarily characterised by exhaustion, which is accompanied by distress, a sense of reduced effectiveness, decreased motivation, and the development of dysfunctional attitudes and behaviours at work'.Burnout is characterised by emotional exhaustion, cynicism and reduced professional efficacy (Barkhuizen, 2005;Jackson & Rothmann, 2005;Maslach et al., 2001).
Exhaustion and cynicism constitute the core of burnout (Schaufeli, 2003).Exhaustion represents the individual stress component of burnout (Maslach et al., 2001) and refers to feelings of being overextended and depleted of emotional and physical resources, i.e. incapable of work performance because all energy has been drained.In the development of burnout, exhaustion emerges first in response to an overly demanding work environment (Leiter, 1993).Cynicism entails a general indifferent, callous or cynical attitude towards the work.To cope with excessive job demands and feelings of exhaustion, the individual psychologically withdraws from the work (mental distancing) (Maslach et al., 2001).Professional efficacy refers to an individual's negative self-evaluation of competence, achievement and productiveness, as well as feelings of insufficiency (Schaufeli & Buunk, 1996).Professional efficacy is the weakest burnout dimension in terms of significant relationships with other variables, and is often referred to as the 'least specific' or 'unnecessary' dimension of burnout (Lee & Ashforth, 1996;Schaufeli, 2003).Several authors argue that professional efficacy reflects a personality characteristic rather than a genuine burnout dimension (Cordes & Dougherty, 1993;Shirom, 1989).Seiler and Pearson (1984-5) note that the consequences of dysfunctional stress (burnout) include two forms of withdrawal: the employee may resign (physical withdrawal) or the employee may remain in employment but continue to do the bare minimum (psychological withdrawal).According to Maslach, Jackson and Leiter (1996), burnout is a result of job demands and lack of job resources can lead to negative outcomes such as physical illness, staff turnover and absenteeism.Research has linked burnout to a variety of mental and physical health problems (Lee & Ashforth, 1990), increased absenteeism (Leiter & Harvie, 1998), and decreased quality and quantity of work performance (Blix, Cruise, Mitchell & Blix, 1994).Eventually, individuals may leave the job or profession as a culmination of burnout (Jackson & Simpson, 2001;Watts et al., 1991).
Empirical studies have revealed that some individuals do not develop burnout, regardless of high job demands and long working hours.On the contrary, they seem to find pleasure in working hard and dealing with job demands (Nelson & Simmons, 2003;Schaufeli & Bakker, 2001).This discovery saw the emergence of theoretical and empirical studies on the concept of engagement.Initially, engagement was regarded the direct opposite of burnout (Rothmann, 2002).However, Schaufeli, Salanova, González-Romá and Bakker (2002) have operationalised engagement as a construct in its own right.Research on engagement has adopted a positive psychology perspective that focuses on psychological health and well-being rather than on psychological ill health -as is the case with burnout (Seligman & Csikszentmihalyi, 2000).Schaufeli and Bakker (2004) define engagement as a positive, fulfilling work-related state of mind, characterised by vigour, dedication and absorption.It is not focused on a specific object, event, individual or behaviour (Schaufeli et al., 2002).Vigour refers to high levels of energy and resilience, willingness to invest effort in one's work, and perseverance in the face of difficulties.Dedication refers to strong involvement in one's work, accompanied by feelings of enthusiasm and significance, and a sense of pride and inspiration (Maslach et al., 2001).Absorption refers to a satisfactory state of complete emersion in one's work, which is characterised by focused attention, time distortion, loss of self-consciousness, effortless concentration, absolute control, and intrinsic enjoyment (Csikszentmihalyi,1990).However, absorption seems to be a problematic dimension from a validity perspective.
According to Schaufeli and Bakker (2004), burnout and engagement are indicators of employees' wellness.Therefore, burnout and work engagement can be integrated as one model (Rothmann, 2002).According to Maslach et al. (2001), the study of work-related experiences should include the entire continuum of work-related experiences, ranging from negative (burnout) to positive (work engagement).However, burnout and engagement are best measured with different instruments (Schaufeli et al., 2002).The Maslach Burnout Inventory -General Survey (MBI-GS) measures burnout across occupational settings, whilst the Utrecht Work Engagement Scale (UWES) measures engagement more effectively (Schaufeli et al., 2002).

Job demands and resources
Several theories and models have been developed to explain the effects of job demands (e.g.work overload) and lack of resources (e.g.job control) on burnout.These include the Conservation of Resources (COR) theory (Hobfoll & Freedy, 1993;Lee & Ashforth, 1996), the Job Demands-Resources (JD-R) model (Demerouti et al., 2001) and the Comprehensive Burnout and Engagement (COBE) model, an extension of the JD-R model with engagement, health impairment and organisational withdrawal as additional components (Schaufeli & Bakker, 2004).The theory underlying these models proposes that burnout develops in response to excessive job demands and diminished job resources.
The COBE model assumes two job-related psychological processes, namely an energetic and a motivational process (Jackson, Rothmann & Van de Vijver, 2006).The energetic process links job demands with health problems via burnout.The motivational process links job resources with organisational outcomes via work engagement.The model has been confirmed in the Netherlands by Schaufeli and Bakker (2004) in an empirical study, with job demands being associated with exhaustion, and job resources with work engagement.Burnout is mainly predicted by job demands and lack of resources, it is related to health problems and turnover intentions, and mediates the relationship between job demands and health problems.Engagement is exclusively predicted by availability of job resources, relates only to turnover intentions, and mediates the relationship between job resources and turnover intentions.
The COR theory (Hobfoll & Freedy, 1993) suggests that burnout is likely to develop when valued resources are lost or threatened, or are inadequate to meet the demands.Major demands include role ambiguity, work pressure and workload.Major resources include control, participation in decision-making and job autonomy (Lee & Ashforth, 1996).According to Leiter (1993), demands and resources are related.A work environment which is overly demanding usually also offers insufficient resources.
Taris, Schreurs and Schaufeli (1999) report that a strong correlation exists between burnout and job stress (situational or organisational factors).Schaufeli and Enzmann (1998) categorise organisational stressors as job demands and lack of job resources.Job demands refer to the things that have to be done or activities to be performed, and include the physical, social or organisational aspects of the job that require sustained physical and mental effort (Demerouti, Bakker, Nachreiner & Schaufeli, 2001).Job demands include situational factors such as role ambiguity, role conflict, stressful events, heavy workload and work pressure, pressure to make critical and immediate decisions, being assigned more responsibility, and having to meet deadlines (Rothmann, 2002;Schaufeli & Enzmann, 1998).
Many burnout researchers have studied quantitative job demands (e.g.too much work for the available time).The findings generally indicate that burnout is a response to overload.Heavy workload and time pressure are strongly and consistently related to burnout, particularly the exhaustion dimension (Maslach et al., 2001).Studies of qualitative job demands have focused primarily on role conflict and role ambiguity, both of which consistently show a moderate to high correlation with burnout.Role conflict occurs when conflicting demands at the job have to be met, whereas role ambiguity occurs when there is a lack of adequate information to do the job well (Maslach et al., 2001).
Job resources refer to all aspects (physical, psychological, social and/or organisational) that reduce job demands, facilitate achievement of work goals, and/or stimulate individual growth (Demerouti et al., 2001b;Rothmann, 2002).Job resources include social support (supervisory and collegial), job enhancement opportunities in the form of increased control and autonomy, participation in decisionmaking, reinforcement contingencies (Burke & Richardsen, 1993), as well as recognition, opportunities for advancement and rewards (Rothmann, 2002).
Burnout researchers have also investigated the absence of job resources.Consistent and strong evidence exists of a correlation between lack of social support and burnout, with lack of supervisory support being more important than support from co-workers (Maher, 1983;Maslach et al., 2001).Correlations have also been confirmed between lack of feedback and all three dimensions of burnout, and between lack of autonomy and burnout.People who enjoy little participation in decision-making seem to experience higher levels of burnout (Maslach et al., 2001).
Based on the above discussion, the following hypotheses are formulated: Hypothesis 1: Burnout is predicted by job demands (i.e.work overload) and a lack of job resources.
Hypothesis 2: Work engagement (vigour and dedication) is predicted by job resources.

Research design
A cross-sectional survey design was used.

Participants
Of the study population (N = 310), a sample of 202 management-level employees across the different operational units of a platinum mine in the North West Province was taken.The characteristics of the participants are shown in Table 1.

Measuring instruments
The follow instruments were used in this study: The Maslach Burnout Inventory -General Survey (MBI-GS) (Maslach et al., 1996) was used to measure burnout.In line with our theoretical model, only two subscales of the MBI-GS, namely Exhaustion (five items, e.g.'I feel used up at the end of the workday') and Cynicism (five items, e.g.'I have become less enthusiastic about my work') were used for the purposes of this study.The third scale of the MBI-GS was not used for the purposes of this study.Schaufeli, Van Dierendonck and Van Gorp (1996) reported Cronbach coefficient alphas varying from 0,87 to 0,89 for Exhaustion, and 0,73 to 0,84 for Cynicism.Test-retest reliabilities after one year were 0,65 (Exhaustion), and 0,60 (Cynicism).The items are scored on a seven-point frequency rating scale ranging from 0 (never) to 6 (always).In South African studies, Cronbach alpha coefficients ranged from 0,86 to 0,88 for Exhaustion, and from 0,79 to 0,80 for Cynicism (Coetzer & Rothmann, 2004;Storm & Rothmann, 2003a).
The Utrecht Work Engagement Scale (UWES) (Schaufeli et al., 2002) was applied to measure participants' levels of engagement.In line with our theoretical model, only two subscales of the UWES were used for the purposes of this study, namely Vigour (five items, e.g.'I am bursting with energy in my work'), and Dedication (five items, e.g.'I find my work full of meaning and purpose').Items are scored on a seven-point scale ranging from 0 (never) to 6 (always).The alpha coefficients for the three subscales varied between 0,68 and 0,91 (Schaufeli et al., 2002).Alpha coefficients varied between 0,78 and 0,89 for the two subscales.Storm and Rothmann (2003b) obtained alpha coefficients of 0,78 for Vigour, and 0,89 for Dedication.

Statistical analysis
The statistical analysis was carried out with the SPSS program (SPSS Inc., 2003) and the AMOS program (Arbuckle, 1999).Cronbach alpha coefficients and factor analysis were used to assess the reliability and validity of the measuring instruments (Clark & Watson, 1995).Descriptive statistics (e.g.means and standard deviations) were used to analyse the data.Pearson correlation coefficients were computed to determine the relationships between variables.A cut-off point of p ≤ 0,05 was set for the statistical significance of the results.Effect sizes (Cohen, 1988) were used to decide on the practical significance of the findings.A cut-off point of 0,30 (medium effect, Cohen, 1988) was set for the practical significance of correlation coefficients.
Exploratory factor analyses were carried out to investigate the construct validity of the following a two-step procedure.First, a simple principal components analysis was conducted on the constructs that form part of the measurement model, including burnout and work engagement.The eigenvalues and scree plots were studied to determine the number of factors.Second, a principal axis factor analysis with a direct oblimin rotation was conducted if factors were related, and a principal component analysis with a varimax rotation was used if the obtained factors were not related (Tabachnick & Fidell, 2001).
Structural equation modelling was used to assess the factorial validity of the measuring instruments of burnout, and work engagement.Among the fit indices produced by the AMOS program is the Chi-square statistic (χ 2 ), which is the test of absolute fit of the model.However, the χ 2 value is sensitive to sample size.Therefore additional goodness-offit indices, such as the Goodness-of-Fit Index (GFI), the Adjusted Goodness-of-Fit Index (AGFI), the Normed Fit Index (NFI), the Comparative Fit Index (CFI), the Tucker-Lewis Index (TLI) and the Root Means Square Error of Approximation (RMSEA), were used in this study.
Standard multiple regression analysis was used to determine which combination of job demands and job resources best predict burnout and work engagement (Tabachnick & Fidell, 2001).

Results
Structural equation modelling (SEM) methods, as implemented by AMOS (Arbuckle, 1999), were used to test the factorial models of the MBI-GS and the UWES.Data analysis was conducted in two consecutive steps.Firstly, quick overviews of the model fits were done by inspecting the overall χ² values, together with the degrees of freedom and probability values.Several goodness-of-fit statistics (GFI, AGFI, NFI, TLI, CFI and RMSEA) were used to globally assess the model fits.Secondly, given findings of poorly fitting initially hypothesised models, the focus shifted from model testing to model development (exploratory factor analysis).Exploratory factor analyses were done for the JDRS Subscales.

Hypothesised model: MBI-GS
Two competing models of burnout were tested, namely a one-factor model and a two-factor model were tested.Table 2 presents fit statistics for the test of the original and other models.
Table 2 indicates that the statistically significant χ² value of 179,04 (df = 35; p = 0,00) revealed a relatively poor overall fit of the hypothesised one-factor MBI model.Seen from a practical perspective, Model 1 was not good either.The GFI, NFI, TLI and CFI values lower than 0,95, and the RMSEA values higher than 0,05 are indicative of failure to confirm the hypothesised models.It is thus apparent that some modification in specification is needed in order to determine a model that fits the sample data better.
The fit of the two-factor MBI model was substantially better (χ² = 320,63; df = 34; p = 0,00).However, the standardised regression coefficient of item 13 (which is supposed to measure Cynicism) was low.It was therefore decided to remove this item.The fit of the two-factors of model was acceptable (with all the fit indices higher than 0,90).
Similar procedures were followed to determine the fit statistics for the UWES.
Both one-factor and two-factor models were tested.Table 3 shows that a two-factor model, labelled here as Model 2, better fits the data set with a lower χ² value of 90,99 (df = 43; p = 0,00).The fit indices were all higher than 0,90, while the χ²/df is lower than 5.The RMSEA value of 0,08 was also acceptable compared to the guideline (it should not be higher than 0,08).In order to determine a model that better represents the sample data, modification indices (MI) were examined to identify possible areas of misfit.Item 15 was retained in spite of a relatively low standardised regression weight of 0,38.The subsequent analysis is therefore based on a two-factor model of the UWES.
The five factors that were extracted accounted for 42,16% of the total variance in the data.With a cut-off of 0,35 for inclusion of a variable in interpretation of a factor, seven of the 67 items did not load on the five factors.Items 42, 43, 44 and 61 did not load strongly (< 0,35) on any of the factors and were removed from the questionnaire.Items 4, 21 and 34 could not be grouped into a meaningful factor and were also removed from the questionnaire.
The first factor was labelled Organisational Support.Items loading on this factor relate to managerial support, communication, role clarity, and the extent of work autonomy.The second factor was labelled Workload and encompasses physical, cognitive and emotional load.Items loading on this factor relate to time pressure, attentiveness to many things at the same time, too much work to do, and dealing with power struggles.The third factor was labelled Resources and involves a variety of resources including collegial support, physical resources such as staff and equipment, as well as financial resources.The fourth factor was labelled Advancement Opportunities.Items loading on this factor relate to growth and development, promotion and financial progress.The fifth factor was labelled Job Security.This factor reflects respondents' indications about being secure in keeping their current jobs in the next year, and about keeping their current job levels in the next year.
Table 5 reports the descriptive statistics, Cronbach alpha coefficients and product-moment correlation coefficients of the measuring instruments, namely the MBI-GS, UWES, and JDRS.
Compared to the guideline of 0,70 provided by Nunnally and Bernstein (1994) for Cronbach coefficient alpha levels, Table 5 shows acceptable Cronbach alpha coefficients varying from 0,79 to 0,94 for all the scales.In conclusion, it can be said that all the instruments showed sufficient reliability to be used for the subsequent analysis.Exhaustion is positively related to Workload, and negatively related to Organisational Support (both medium effects).Cynicism is practically significantly negatively related to Organisational Support and Advancement (both medium effects).Vigour (medium effect) and Dedication (large effect) are practically significantly related to Organisational Support.
Multiple regression analyses were carried out with exhaustion and cynicism (as measured by the MBI-GS), vigour and dedication (as measured by the UWES) as dependent variables, and job demands and resources (as measured by the JDRS) as independent variables (see Table 6).The multiple regression analyses were carried out by entering the independent variables in blocks in two steps.In the case of exhaustion, overload and job insecurity (as demands) were entered in the first step, while organisational support, resources and advancement (as job resources) were entered in the second step.Table 6 shows that 20% of the variance in Exhaustion (as measured by the MBI-GS) is predicted by Workload (β = 0,37; p < 0,01) and low Job Security (β = -0,21; p < 0,01), as measured by the JDRS.Furthermore, 33% of the variance in Exhaustion (as measured by the MBI-GS) is predicted by the factors of the JDRS.However, only the regression coefficients of Workload (β = 0,38; p < 0,01), low Job Security (β = -0,20; p < 0,01), and a lack of Resources (β = -0,19; p < 0,01) were statistically significant.Adding the three job resources to the multiple regression significantly increased the R² from 0,19 to 0,31.
Table 6 shows that 21% of the variance in Cynicism (as measured by the MBI-GS) is predicted by low Organisational Support (β = -0,35; p < 0,01) and low Advancement Opportunities (β = -0,20; p < 0,01), as measured by the JDRS.Furthermore, 24% of the variance in Cynicism (as measured by the MBI-GS) is predicted by the factors of the JDRS.However, only the regression coefficients of Organisational Support (β = 0,33; p < 0,01), and a low Advancement Opportunities (β = -0,19; p < 0,01) were statistically significant.Adding Workload and Job Security to the multiple regression did not have any statistically significant effect on the R² in Step 2.
Table 7 shows that 20% of the variance in Vigour (as measured by the UWES) is predicted by job resources (as measured by the JDRS).However, only the regression coefficient of Workload (β = 0,36; p < 0,01) was statistically significant.Furthermore, 21% of the variance in Vigour (as measured by the UWES) is predicted by the factors of the JDRS.However, only the regression coefficient of Organisational Support (β = 0,33; p < 0,01) was statistically significant.Adding Workload and Job Security in Step 2, did not have statistically significant effects on the Vigour of participants.Table 7 shows that 30% of the variance in Dedication (as measured by the UWES) is predicted by job resources (as measured by the JDRS).However, only the regression coefficient of Organisational Support (β = 0,54; p < 0,01) was statistically significant.Furthermore, 34% of the variance in Dedication (as measured by the UWES) is predicted by the factors of the JDRS.However, only the regression coefficients of Organisational Support (β = 0,51; p < 0,01) and Workload (β = -0,20; p < 0,01) were statistically significant.

Discussion
This study set out to investigate the relationships between job demands, job resources, burnout, and work engagement.
In support of the COBE model (Schaufeli & Bakker, 2004), the Pearson correlation analysis confirmed that burnout (consisting of exhaustion and cynicism) was negatively related to engagement (consisting of vigour and dedication).Low levels of burnout were related to high levels of engagement (i.e.high levels of energy related to high levels of identification), confirming the findings of Schaufeli, Martinez, Pinto, Salanova and Bakker (2002).
Exhaustion was positively related to job demands because of workload (which encompassed physical, emotional and cognitive workload), and negatively related to job resources because of insufficient organisational support (which focused mainly on management support, communication, performance feedback, participative decision-making, work autonomy and role clarity).Maslach et al. (2001) found strong correlations between job demands (heavy workload and time pressure) and exhaustion.The findings of this study confirmed the theory underlying the JD-R model (Demerouti et al., 2001) that job demands are primarily related to exhaustion.
Cynicism correlated negatively with job resources because of lack of organisational support and advancement opportunities (which included growth, development and promotional opportunities).Barkhuizen (2005) reported similar findings.Both burnout dimensions were negatively related to organisational commitment, i.e. turnover intentions, and positively related to both physical and psychological ill health.Barkhuizen (2005) also reported exhaustion to be related to ill health.Research has linked burnout with various physical and psychological health problems (Lee & Ashforth, 1990).Maslach (1998) and Wiese, Rothmann and Storm (2003) found a strong correlation between exhaustion and high job demands.Storm and Rothmann (2003b) confirmed the association between exhaustion and lack of job resources.
Regarding the positive outcomes of wellness, vigour and dedication were positively related to job resources because of organisational support and advancement opportunities.Coetzer and Rothmann (2004) found that job demands and a lack of job resources increased levels of burnout, while availability of resources increased levels of engagement.This study further confirmed the theory of the JD-R model (Demerouti et al., 2001)  Existing theoretical relationships between burnout and engagement as well as burnout and ill health were confirmed in this study, and specifically the theory that burnout develops in response to excessive job demands and diminished job resources (Schaufeli & Bakker, 2004).
The results should be interpreted in view of the current transformation process in the organisation and the history of the organisation.Eight months ago the organisation commenced with restructuring aimed at becoming a worldclass organisation and reducing costs.The new organisational design includes centres of excellence and shared business services that will result in job losses and possible demotions.The majority of affected employees are within the management ranks.Although the aims of the restructuring exercise were communicated eight months ago, communication on progress made has been limited and the appointment of employees into their new roles has been extremely slow.It is likely that the situation has increased the anxiety levels of the employees in the organisation, that it has contributed to the current levels of job insecurity experienced by the participants, as well as to a sense of reduced control, participation in decision-making and job autonomy.
Simultaneously, other change initiatives such as continuous improvement, an enterprise resource system and a culture change initiative were implemented.The organisation expects its managers to embrace and champion the change initiatives.According to Maslach and Leiter (1997), major organisational transitions increase managers' workload in three ways: it becomes more intense and more complex and demands more time.It can thus be argued that the workload (physical, emotional and cognitive) of the managers has increased significantly over the past eight months.However, their job resources have not increased.
In conclusion, it can be argued that participants developed exhaustion in response to a significant increase in workload, high levels of job insecurity, and insufficient job resources.This resulted in psychological withdrawal (i.e.cynicism, reduced dedication and vigour) both from the work and the organisation.This is similar to findings cited by Maslach et al. (2001) in a meta-analysis of burnout and engagement research.The results confirmed the underlying theory of the COBE model.Work wellness (burnout and engagement) was predicted by both job demands (workload and job insecurity) and lack of job resources (lack of organisational support and advancement opportunities).Organisational commitment was predicted by work wellness (burnout and engagement).
The present study has certain limitations.The research was a cross-sectional survey design.As a result, no causal inferences could be drawn, even though advanced analytical procedures were employed.Another limitation is that the measurement of this model's variables was based solely on self-reports.Furthermore, the study population was very homogeneous.From a sample of 202 managers, 88,6% were male, and 84,7% white.South Africa's multicultural society necessitates studying the constructs burnout, engagement and organisational commitment for managers from different cultural groups, and proving the construct equivalence and the absence of item bias for these groups.Future studies should include larger sample sizes.

Recommendations
Given the pervasive nature of burnout, organisations should adopt a preventative approach.According to Kompier and Kristensen (2001), interventions may, primarily be directed at the work situation or the coping capacity of employees.Work-oriented interventions aim at improving the fit between an individual and the workplace to the benefit of the individual and the organisational system.Employeeoriented interventions aim at teaching employees effective stress management skills, or skills to modify their appraisals of stressful situations as being less stressful.In the context of the organisation, effective human resource systemsincluding career development and performance management -should be implemented as a matter of priority.
Secondly, interventions may be aimed at eliminating, reducing or altering stressors (primary interventions).Possible interventions include: changes in decision-making processes, work redesign, and provision of a more supportive climate (including constructive performance feedback).Thirdly, secondary-level interventions can be implemented to prevent employees who are already showing signs of stress from getting sick, and to increase their coping capacity.Examples of this strategy include cognitive restructuring, time management, conflict resolution techniques and coping strategies.
More research should be conducted on how to prevent burnout and enhance engagement.Research should also be conducted to evaluate the effectiveness of interventions.

Table 1 : Characteristics of the participants (N = 202)
The Job Demands-Resources Scale (JDRS) was developed for the organisation through focus group interviews.The contextualised questionnaire consists of 67 items and measures job demands and job resources for employees.

Table 7 : Multiple regression analyses with vigour and dedication as dependent variables
that lack of job resources are related primarily to disengagement.