Stress and Health

Stress and Health 24: 3547 (2008) Published online 25 October 2007 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/smi.1163 Received 1 May 2006; Accepted 23 August 2007

A model of work wellness for non-professional counsellors in South Africa

Lukas Fourie,1 Sebastiaan Rothmann1,*, and Fons J. R. van de Vijver1,2
1

WorkWell: Research Unit for People, Policy and Performance, North-West University, South Africa 2 Department of Psychology, Tilburg University, the Netherlands

Summary The objective of this study was to investigate the effects of job demands, job resources and sense of coherence on the burnout and work engagement of non-professional counsellors in South African banks. A cross-sectional survey design was used. The sample consisted of 165 non-professional counsellors doing trauma counselling in three of the major banks in South Africa. The Maslach Burnout InventoryHuman Services Survey, Utrecht Work Engagement Scale, Orientation to Life Questionnaire and Job Characteristics Scale were administered. A good t was found for a model in which sense of coherence was a positive predictor of perceived low job demands and high availability of job resources and work wellness (low burnout and high work engagement). Non-professional counsellors with a stronger sense of coherence experienced more work wellness (low burnout and high work engagement) than those with a weaker level of coherence, presumably because stimuli from the environment are perceived as making cognitive sense, as under control of both the counsellor and signicant others, and as motivationally relevant and meaningful. Copyright 2007 John Wiley & Sons, Ltd.

Key Words work wellness; burnout; engagement; sense of coherence; South Africa; non-professional counsellors

Introduction Over the last few years, South Africa has been characterized by a high incidence of crime and violence (Louw, Shaw, Cameren, & Robertshaw,
* Correspondence to: Sebastiaan Rothmann, WorkWell: Research Unit for People, Policy and Performance, North-West University, Potchefstroom, South Africa 2520. E-mail: Ian.Rothmann@nwu.ac.za Grant sponsor: National Research Foundation. Grant number: 2053344. Copyright 2007 John Wiley & Sons, Ltd.

1998; Schlemmer, 1996). Violence in the broader environment spills over into the workplace. The workplace could therefore become a primary source of violence, such as is evident in the number of armed robberies in nancial institutions. These incidents amounted to a total of 315 for the 2002/2003 nancial year and included armed robberies in branches, cash-in-transit robberies and physical auto teller machine attacks. Furthermore, the modus operandi of armed robberies have become more violent; examples are hostagetaking and threatening staff that they would be set alight after being covered with highly ammable liquids (ABSA Security, 2003).

L. Fourie, S. Rothmann and F. J. R. van de Vijver

A response to workplace violence and trauma has been the introduction of trauma management interventions utilizing trained non-professional trauma counsellors who assist colleagues in coming to terms with the experience of a violent work-based incident (Ortlepp & Friedman, 2002). Beaton and Murphy (1995) see non-professional counsellors as crisis workers and front-line rst responders to whom potential exposure to occupational trauma is a fact of daily life (p. 51). Non-professional counsellors are often volunteers who provide their services on a part-time basis without receiving nancial remuneration. Often they do not even participate in a formal selection procedure (Wilson, 1998). They receive shortterm training in a specic eld such as trauma counselling and do not necessarily have a formal qualication. A growing body of theoretical and empirical literature recognizes that engaging in therapeutic work with trauma survivors can, and does, impact on the counsellor (Figley, 1995; Jenkins & Baird, 2002). Research indicates that helpers are exposed to stressors which can produce an array of psychological, social and physical reactions and even burnout (Cordes & Dougherty, 1993). Although they do not personally experience the physical and emotional distress of the trauma, they do see the destruction it has produced and are intensively exposed to the losses and pain of others. Furthermore, non-professional counsellors are selected from human resource practitioners in various South African banks. This implies that they full counselling duties in addition to their normal job duties and that they might experience conicting demands from counselling and human resource roles. These factors might lead to burnout and low work engagement. Despite extensive research on job stress experienced by caregivers and the impact that engaging in counselling work with traumatized people has on counsellors (Beaton & Murphy, 1995; Figley, 1995), burnout and work engagement in nonprofessional trauma counsellors in a banking environment have not been investigated. While burnout is commonly used to describe a state or process of mental exhaustion and disengagement (Schaufeli & Enzmann, 1998), work engagement is being dened as an energetic state in which the employee is dedicated to excellent performance at work (Schutte, Toppinen, Kalimo, & Schaufeli, 2000). The new focus in the eld on engagement and the positive opposite of burnout promises to yield new perspectives on the interventions to
36 Copyright 2007 John Wiley & Sons, Ltd.

promote mental and physical well-being (Salovey, Rothman, Detweiler, & Steward, 2000) and to alleviate burnout (Maslach, Schaufeli, & Leiter, 2001). According to Nelson and Simmons (2003), meaningful work leads to eustress, which would promote engagement even if the situation is demanding. In an attempt to identify the sources of nonprofessional counsellor stress, with an almost exclusive focus on the negative effects of counselling, little insight is available about the well-being of non-professional counsellors. There is an increasing appreciation in occupational health psychology that both positive and negative health aspects should be considered in order to get a comprehensive view of the health status of workers. It seems, therefore, also necessary to study the work of non-professional counsellors from a positive stance. This is done in the current study by examining both burnout and work engagement. A study by Ortlepp and Friedman (2002) focused on secondary traumatic stress and emotional exhaustion of workplace non-professional trauma counsellors in South African banks. However, their study did not address the role of job demands (other than emotional demands) and resources, which could be related to burnout of non-professional counsellors. Furthermore, positive aspects of their well-being at work (e.g. engagement) were not investigated. Burnout and work engagement According to Maslach and Jackson (1986), burnout consists of three dimensions, which are measured by the Maslach Burnout InventoryHuman Services Survey (MBI-HSS), namely emotional exhaustion, depersonalization and reduced personal accomplishment. Emotional exhaustion refers to the depletion or draining of emotional resources caused by interpersonal demands. Depersonalization involves the development of negative and cynical attitudes towards the recipients of ones services. Finally, reduced personal accomplishment refers to a decline in ones feelings of competence and productivity (Cordes & Dougherty, 1993). The concept of engagement was introduced by Kahn (1990). Engagement is dened by Kahn (1990) as the harnessing of organization members selves to their work roles (p. 694). Personal disengagement is described as the uncoupling of
Stress and Health 24: 3547 (2008) DOI: 10.1002/smi

Work wellness
selves from the work roles. Individuals who are engaged become physically involved in tasks, are cognitively vigilant and become emphatically connected to others in the work they are doing (Kahn, 1990). Maslach and Leiter (1997) argue that engagement is the opposite of burnout and can be adequately measured by the MBI (i.e. by reversing the MBI scores). However, Schaufeli, Salanova, Gonzlez-Rom, and Bakker (2002b) argue that burnout and engagement are conceptually distinct and that their relationship should be studied empirically. Work engagement, as measured by the Utrecht Work Engagement Scale (UWES), consists of three dimensions, namely vigour, dedication and absorption. Vigour is characterized by high levels of energy and mental resilience while working, as well as a willingness to exert effort and to persist even through difcult times. Dedication is characterized by a sense of signicance in ones work, feeling enthusiastic, inspired and proud, and by viewing it as a challenge. Vigour and dedication are the direct opposites of (emotional) exhaustion and mental distance (depersonalization). Absorption comes close to the concept of ow (Nakamura & Csikszentmihalyi, 2002), an optimal state of experience where focused attention, a clear mind, unison of body and mind, effortless concentration, complete control, loss of self-consciousness, distortion of time, and intrinsic enjoyment are experienced (Schaufeli, Martinez, Pinto, Salanova, & Bakker, 2002a). Schaufeli and Bakker (2004) found that the engagement and burnout scales showed moderately negative correlations. Job demands and resources It has been proposed that the severity of clients problems, working with clients, time limitations and inadequate resources are factors that place counsellors at risk for developing burnout (Maslach & Schaufeli, 1993). However, high job resources could contribute to work engagement, even in the face of extraordinary demands (Schaufeli & Bakker, 2004). Therefore, the Job Demands-Resources model (Demerouti, Bakker, Nachreiner, & Schaufeli, 2001) seems to be relevant for investigating the effects of contextual factors on burnout and work engagement among non-professional counsellors. According to Demerouti et al. (2001), job demands refer to those physical, psychological,
Copyright 2007 John Wiley & Sons, Ltd.

social or organizational aspects of the job that require sustained physical and/or psychological effort and are therefore associated with certain physiological and/or psychological costs (e.g. work pressure, role overload and emotional demands). Job resources are those physical, psychological social or organizational aspects of the job that may be functional in achieving work goals, reducing job demands, and stimulating personal growth and development Resources may be located at the level of the organization (e.g. salary and career opportunities), interpersonal relations (e.g. supervisor and co-worker support), the organization of work (e.g. role clarity), and the level of the task (e.g. task signicance and autonomy). According to Demerouti et al. (2001), job demands may preclude the mobilization of job resources. However, high job resources (e.g. social support and autonomy) may reduce the effects of job demands. These authors reported that job demands (e.g. emotionally demanding relationships with clients) are primarily related to the exhaustion component of burnout, whereas a lack of job resources is primarily related to disengagement. Lower levels of burnout could be expected in work situations that provide opportunities to experience challenge, control, autonomy, role clarity, participation in decision making, feedback of results, and support from supervisors and co-workers (Schaufeli & Enzmann, 1998). Whereas burnout of non-professional counsellors seems to be predicted by high job demands and low job resources, Schaufeli and Bakker (2004) showed that work engagement is predicted by the availability of job resources. Although limited information is available regarding the relationship between job demands and work engagement, it seems that individuals could experience work engagement despite high demands. For example, Watts et al. (1991) reported that some employees were very satised with their jobsdespite long working hours and work overload. Kinman and Jones (2003) also found that employees in their study thrived on the fact that their work is stressful. Job resources might moderate the effects of job demands on work engagement. Hakanen, Bakker, and Demerouti (2005, p. 2) found that job resources diminish the effects of job demands on work engagement. They showed that the relationship between job demands and work engagement was weaker for individuals with high (vs. low) job resources.
Stress and Health 24: 3547 (2008) DOI: 10.1002/smi 37

L. Fourie, S. Rothmann and F. J. R. van de Vijver

Sense of coherence Sense of coherence may either alleviate or aggravate stress reactions. It can be described as a pervasive, enduring though dynamic feeling of condence that ones internal and external environments are predictable and that there is a high probability that things will work out as well as can be reasonably expected. The concept includes three dimensions, namely comprehensibility, manageability and meaningfulness (Antonovsky, 1987). Comprehensibility refers to the extent to which one perceives stimuli from the internal and external environment as information that is ordered, structured and consistent. The stimuli are perceived as comprehensible and make sense on a cognitive level. Manageability refers to the extent to which individuals experience events in life as situations that are endurable or manageable, or even as new challenges. Meaningfulness refers to the extent to which one feels that life is making sense on an emotional and not just a cognitive level. Sense of coherence is hypothesized as a fairly stable dispositional personality orientation (Antonovsky, 1987, 1993). During the years of childhood and adolescence, consistency in life experiences enhances comprehensibility, load balances improve manageability, and participation in socially valued decision making promotes meaningfulness. After age 30, sense of coherence is expected to remain relatively stable, as the individual has already made major commitments in his or her life: marriage, a job, a style of life and a set of social roles. All of these commitments provide an individual with a stable set of life experiences, which enables a sense of coherence to become established. To our knowledge, only two studies have investigated job demands and/or resources as predictors of sense of coherence (Albertsen, Nielsen, & Borg, 2001; Feldt, Kinnunen, & Mauno, 2000). In contrast with the view of sense of coherence as an outcome, Feldt, Kivimki, Rantala, and Tolvanen (2004) investigated sense of coherence as a predictor of perceived job characteristics. If sense of coherence is indeed a stable personality construct in adulthood, it is better viewed as a predictor of perceptions at work than as an outcome of such perceptions. Individuals with a strong sense of coherence tend to experience environmental stimuli in a manner sufciently structured to enable them to anticipate events and the resources required to meet the demands
38 Copyright 2007 John Wiley & Sons, Ltd.

imposed on them. Such experiences are likely to lead to favourable perceptions of a persons inuence at work and receiving support from supervisors and colleagues. Antonovsky (1987) also assumed that sense of coherence relates to the probability of exploiting potential resources. In a longitudinal study, Feldt et al. (2004) found that sense of coherence inuenced the ability to mobilize and generate social resources in the workplace but not the ability to produce job control. Amirkhan and Greaves (2003) showed that individuals with a strong sense of coherence were likely to view more life events as having coherence. This perceptual process seems to be subtle: it inuences individuals perceptions of stressful events, but it does so without their conscious awareness. Sense of coherence could also have a direct effect on individuals work-related well-being. Ortlepp and Friedman (2002) found that sense of coherence was strongly related to nonprofessional counsellors reactions to their trauma counselling experiences. Higher levels of sense of coherence were associated with lower levels of secondary traumatic stress. Levert, Lucas, and Ortlepp (2000) reported signicant correlations between two components of burnout (exhaustion and depersonalization) and sense of coherence in a group of psychiatric nurses in South Africa. Gilbar (1998) found signicant correlations between social workers sense of coherence and exhaustion (r = 0.30), as well as their sense of coherence and personal accomplishment (0.34). Rothmann, Malan, and Rothmann (2001) also found that sense of coherence was signicantly correlated with exhaustion (0.56), depersonalization (0.41) and personal accomplishment (0.48). Rothmann, Steyn, and Mostert (2005) showed that sense of coherence is moderately related to work engagement. Aims and hypotheses The objective of this study was to investigate the effects of job demands, job resources and sense of coherence on the burnout and work engagement of non-professional counsellors in South African banks. The following hypotheses were tested: H1: Perceptions of job demands and a lack of resources lead to burnout, while job resources lead to engagement.
Stress and Health 24: 3547 (2008) DOI: 10.1002/smi

Work wellness
H2: Sense of coherence exerts a positive main effect on work wellness (i.e. low burnout and high work engagement) via perceptions of reduced job demands and enhanced job resources. Method Research design A cross-sectional survey design was used. Participants Data were collected among non-professional counsellors employed by three of the major banks in South Africa (n = 295). These employees were from all the provinces in the country and did line or staff work on clerical, supervisory, managerial and specialist levels. A total of 168 nonprofessional counsellors participated in the study. Only 165 questionnaires could be used (due to three questionnaires having missing values). The response ratios for the three banks were 80, 65 and 0 per cent, respectively. Table I presents the characteristics of the participants. The participants were mostly female
Table I. Characteristics of the participants (n = 165). Item Position Category Clerical Supervisory Managerial Specialist Grade 10 Grade 12 Certicate Diploma Degree Postgraduate degree Male Female Single Engaged or in close relationship Married Divorced Separated Widow/widower Remarried Afrikaans English African Language Frequency 16 19 65 65 3 28 19 36 27 52 36 129 25 10 88 21 6 6 9 82 66 17 Percentage 9.7 11.5 39.4 39.4 1.8 17.0 11.5 21.8 16.4 31.5 21.8 78.2 15.2 6.1 53.3 12.7 3.6 3.6 5.4 49.7 40.0 10.3

(78.4 per cent), married and had a high-school education (preliminary statistical analyses indicated that gender was unrelated to any of the target variables of the current study; therefore, gender was not further considered in the analyses). The age of participants ranged from 20 to 60, but only 19.9 per cent were younger than 30. The mean age of participants was 37.97 years [standard deviation (SD) = 8.82], while the mean length of experience as a non-professional counsellor was 5.82 years (SD = 4.85) and the mean number of years of experience in the current working position was 6.13 (SD = 4.57). The mean number of working hours per month spent on counselling was 8.19 (SD = 8.58), while a mean number of 7.09 h (SD = 9.25) was spent outside working hours on counselling in the community. The types of counselling in which the counsellors were engaged were trauma (71.4 per cent), HIV/ AIDS (12.6 per cent), relationships (6.7 per cent), death and bereavement (5.9 per cent), and substance abuse (3.4 per cent). All participants had followed a 4-day training programme for non-professional counsellors before they were deployed by the banks. In eldwork, they operated individually or in teams of two. Each non-professional counsellor had a maximum caseload of ve, and they were only allowed to work with traumatized employees for four

Education

Gender Marital status

Home language

Copyright 2007 John Wiley & Sons, Ltd.

Stress and Health 24: 3547 (2008) DOI: 10.1002/smi

39

L. Fourie, S. Rothmann and F. J. R. van de Vijver

sessions (normally spread over a period of 1 month). Typically they contacted these employees on the day of the incident, the following day, a week later and a month later. The non-professional counsellor usually met employees directly after the incident, listened to their experiences, explained the symptoms of post-traumatic stress, provided support, and referred them to psychologists if they were traumatized. Because employees who were traumatized by the incidents were usually not willing to go back to work on the day following their traumatic experience, nonprofessional counsellors accompanied them. However, because these counsellors were all bank employees, they had to catch up with their own work when they went back to work. A psychologist usually phoned them when they arrived back at work to discuss their experiences and to provide social support. Participants received a certicate annually, but they received no additional compensation for their work as non-professional counsellors. Instruments The MBI-HSS (Maslach & Jackson, 1986) was administered. The scale measures participants relationships with their work on a continuum from engagement to burnout. The MBI-HSS consists of 22 items and has three subscales. All items are scored on a seven-point frequency rating scale, ranging from 0 (never) to 6 (daily). The factoranalysed subscales for the MBI-HSS include emotional exhaustion, depersonalization and personal accomplishment (Maslach, Jackson, & Leiter, 1996). Naud and Rothmann (2004a) found support for the construct validity (i.e. factorial composition) of the MBI-HSS in a South African sample. Fourie (2003) also conrmed a threefactor model of burnout in the current sample. However, only 17 items loaded signicantly on three scales: emotional exhaustion (six items), depersonalization (ve items) and personal accomplishment (six items). Results reported here are based on the set of 17 items. The UWES (Schaufeli et al., 2002b) was used to measure the levels of engagement of nonprofessional trauma counsellors. The UWES measures levels of engagement on a 17-item, seven-point frequency rating scale, ranging from 0 (never) to 6 (every day). Three dimensions can be distinguished, namely vigour, dedication and absorption. Examples of statements relating to the
40 Copyright 2007 John Wiley & Sons, Ltd.

three dimensions are the following: I am bursting with energy in my work (vigour); I nd my work full of meaning and purpose (dedication); and When I am working, I forget everything around me (absorption). Engaged individuals are characterized by high levels of vigour and dedication as well as elevated levels of absorption. Naud and Rothmann (2004b) conrmed the construct validity of the UWES in a South African sample. Fourie (2003) found support for a one-factor model of work engagement in the current sample. The Orientation to Life Questionnaire (OLQ) (Antonovsky, 1987) was used to measure sense of coherence. The 13-item questionnaire was adopted for the present study. The 13-item version of the OLQ includes four items measuring meaningfulness (e.g. How often do you have a feeling that there is little meaning in the things you do in your daily life?), ve items measuring comprehensibility (e.g. Do you have the feeling that you are in an unfamiliar situation and dont know what to do?) and four items measuring manageability (e.g. Has it ever happened that people whom you counted on disappointed you?). Answers have to be given on a seven-point Likert scale on which the extreme answers (e.g. 1 = never and 7 = always) are formulated for each question. Antonovsky (1993) reported alpha coefcients of the OLQ in 29 research studies varying between 0.85 and 0.91. Testretest reliability studies reported coefcients between 0.41 and 0.97 (Antonovsky, 1993). Conrmatory factor analysis in this sample showed that a onefactor model (consisting of 12 items) tted the data better than a three-factor model. After reversing the scores on four items, a total score is obtained, which can range from 12 (low sense of coherence) to 84 (high sense of coherence). The items were reverse scored so that a high score indicates a high sense of coherence. A Job Characteristics Questionnaire (JCQ) with 33 questions was developed, measuring demands and resources associated with nonprofessional counselling, in addition to other duties at work. The items for the JCQ were developed based on experience with the typical job demands and resources reported by non-professional counsellors. The items refer to participants work as non-professional counsellors at the bank. All items are scored on a ve-point rating scale ranging from 1 (strongly disagree) to 5 (strongly agree). Exploratory factor analysis of the JCQ resulted in two factors, which were labelled job demands and job resources (Fourie, 2003). A
Stress and Health 24: 3547 (2008) DOI: 10.1002/smi

Work wellness
simple principal component analysis was carried out on the 33 items of the JCQ. Bartletts test of sphericity showed that the items were factorable (2 = 2923.96, df = 528, p < 0.01). Furthermore, the KaiserMeyerOlkin measure of sampling adequacy was 0.84, which is acceptable compared to the recommended value of higher than 0.60. Examples of items of the Job Demands scale are: I can handle my workloadday-to-day activities, as well as counselling; I have too much paper work to do when counselling; and I am able to make myself available to meet the time commitments involved in counselling). Examples of the Job Resources scale are: I have adequate support from my superiors; I am adequately compensated for my counselling contributions; I have adequate access to debrieng when I need it; and There are clear policies and guidelines to assist me in my role as a counselor. Ethics Considerations regarding ethical issues were addressed by means of active inclusion and consultation with the relevant stakeholders at the banks, as well as with the participants in the study. Participation was voluntary. The objectives of the study were explained to the participants at their place of work, and written consent was obtained from them. Both individual and group feedback were provided for in the design of the study. Condentiality and anonymity were assured. Statistical analysis Structural equation modelling methods as implemented in Analysis of Movement Structures (AMOS) (Arbuckle, 1997) were used to test a structural model of work wellness using the maximum likelihood method. Hypothesized relationships are tested empirically for goodness of t with the sample data. Among the t indices produced by the AMOS programme is the Chi-square statistic (2), which is the test of absolute t of the model. However, the 2 value is sensitive to sample size. Therefore, additional goodness-of-t 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 were used in this study.
Copyright 2007 John Wiley & Sons, Ltd.

Results Cronbachs alpha coefcients of all the measuring instruments are considered to be acceptable, if they are larger than 0.70 (Nunnally & Bernstein, 1994). The percentages of non-professional counsellors who experienced low, moderate or high levels of burnout and work engagement were examined. Reference data were obtained from a South African norm sample (Rothmann, 2005). Compared to the normative data, they reported a below-average level of burnout (emotional exhaustion and depersonalization) and aboveaverage levels of engagement (personal accomplishment and work engagement). Furthermore, compared to the normative data, they reported an above-average level of sense of coherence. To prepare the data for the purpose of testing a structural model of work wellness of nonprofessional counsellors, exploratory factor analyses were carried out on the scales of the MBI-HSS and UWES, as well as the JCQ using SPSS (SPSS Inc., 2003). Firstly, a simple principal component analysis was conducted on the scales of the MBI-HSS and UWES. The scree plot suggested the extraction of two factors (which explained 81.1 per cent of the total variance). The KaiserMeyerOlkin measure of sampling adequacy (0.76) and the Bartletts test of sphericity (2 = 227.96, df = 6, p < 0.01) showed that the items were factorable. Principal component analysis with an Oblimin rotation resulted in two correlated factors (r = 0.52), namely burnout, including emotional exhaustion (loading = 0.91) and depersonalization (loading = 0.88), and extended work engagement, including work engagement (loading = 0.64) and personal accomplishment (loading = 0.99). Next, a simple principal component analysis was conducted on the scales of the JCQ. The scree plot and eigenvalues provided evidence for a one-factor solution, which explained 84.80 per cent of the total variance. This factor was labelled Job Characteristics. The KaiserMeyerOlkin measure of sampling adequacy (0.50) and the Bartletts test of sphericity (2 = 107.67, df = 1, p < 0.01) showed that the items were factorable. Structural model of work wellness A model including the hypothesized relationships was tested in a path model (correlations are given in Table II). The latent variables included burnout
Stress and Health 24: 3547 (2008) DOI: 10.1002/smi 41

L. Fourie, S. Rothmann and F. J. R. van de Vijver

(consisting of two observed variables, namely exhaustion and depersonalization), extended work engagement (consisting of two observed variables, namely work engagement and personal accomplishment), perceived job characteristics (consisting of two observed variables, namely job demands and job resources), and sense of coherence. Three of the four dimensions (namely burnout, work engagement and job characteristics) were covered by at least two scales. For each of these three dimensions, a latent variable was specied on which the corresponding scales loaded, separating random measurement error from true score variance. For sense of coherence, there was only one indicator, meaning that in this case there was a one-to-one correspondence between the manifest variable (scale) and the underlying latent dimension. In this case, usually no distinction is made between random error variance and true score variance, meaning that the correlations among this one-indicator latent variable and other latent variables may be biased (Little, Cunningham, Shahar, & Widaman, 2002). This problem was overcome using a procedure proposed by Bagozzi and Heatherton (1994). Firstly, a one-factor model was tted for all items belonging to the shortened OLQ. Secondly, separate indicators for the scale were formed by selecting items on the basis of their loadings, alternating items with high and low loadings. In this vein, three parcels of items were created for sense of coherence. Our hypotheses stated that job demands and a lack of job resources would lead to burnout, while job resources would lead to work engagement. However, a strong correlation (r = 0.62) was found between burnout (emotional exhaustion and depersonalization) and extended work engagement (work engagement and personal accomplishment). The covariance between these two latent variables could have been modelled by allowing their error terms to correlate. However, this is a controversial technique (Byrne, 2001) and it is unlikely that such a strong correlation is due to methodological artefacts such as commonalities of response formats. In our view, burnout and work engagement should not be modelled here as independent constructs. We modelled their covariance by adding a latent variable (called work wellness). Results indicated that the model tted adequately to the data: 2(22, n = 165) = 39.85, p < 0.01; 2/df = 1.81; GFI = 0.95; AGFI = 0.90; CFI = 0.98; NFI = 0.95, and TLI = 0.96. The model
Stress and Health 24: 3547 (2008) DOI: 10.1002/smi 0.18 0.61*, 0.38*, 0.56*, 0.55*, 0.47*, 0.46*, 0.35*, 1. 2. 3. 4. 5. 6. 7. 8. Emotional exhaustion Depersonalization Personal accomplishment Work engagement Sense of coherence Job demands Job resources Age 13.56 5.00 28.70 66.61 148.37 19.70 32.45 37.97 8.20 4.51 5.05 10.28 19.57 2.87 5.39 8.71 37 19 26 48 112 12 28 40 0.91 0.70 0.75 0.93 0.87 0.71 0.80 0.43*, 0.54*, 0.45*, 0.31*, 0.40*, 0.26*
* Correlation is statistically signicant (p < 0.01). Correlation is practically relevant and has a medium effect size (0.30 < |r| < 0.50). Correlation is practically relevant and has a large effect size (|r| > 0.50). SD: standard deviation.

7 6 5 4 Table II. Descriptive statistics, alpha coefcients and Pearson correlations of the scales. 42 Scale Mean SD Range 1 2 3

Copyright 2007 John Wiley & Sons, Ltd.

0.58*, 0.49*, 0.30*, 0.38*, 0.21*

0.60*, 0.36*, 0.43*, 0.36*,

0.35*, 0.45*, 0.29*

0.70*, 0.33*,

Work wellness
Work Wellness 31

Job Demands -.79

Job Resources .88

Emotional Exhaustion .84 -.89 .38 Work Wellness R2 = .79 .88

Depersonalisation

.75 Burnout R2 = .79

Job Characteristics 2 R = .19

.44

.66

Engagement R2 = .78 .87 .68 Personal Accomplishment

Sense of Coherence
.84 Sense of Coherence Indicator 1 .77 Sense of Coherence Indicator 3

.91 Sense of Coherence Indicator 2

Vigour/ Dedication/ Absorption

Figure 1. Maximum likelihood estimates for the work wellness model, n = 165 (standardized solution). Note: All factors loadings and path coefcients are signicant (p < 0.01).

is presented in Figure 1. As can be seen there, the path from sense of coherence to perceived job characteristics was statistically signicant. A low sense of coherence of non-professional counsellors contributed to perceived high job demands and low job resources. Perceived job characteristics (low job demands and high job resources) had a positive impact on work wellness, which is a second-order factor, comprising of burnout (negative loading) and engagement. Furthermore, sense of coherence had a main effect on work wellness. The direct effect of sense of coherence on work wellness and the indirect effect (through job characteristics) reinforce each other. The model indicates that sense of coherence has a strong inuence on both burnout and engagement and that this effect is both direct and indirect. As could be expected, burnout and extended work engagement had opposite loadings on the work wellness factor. Our rst hypothesis stating that perceived job demands would predict burnout while job resources would predict engagement
Copyright 2007 John Wiley & Sons, Ltd.

was supported, although a caveat is needed. The hypothesis seemingly implies that job demands are not related to engagement. No evidence for this implication was found; quite on the contrary, job demands and job resources showed strong correlations and did not show a differential pattern of relationships with engagement and burnout. Job demands were found to exert a broader inuence than indicated in our rst hypothesis. Non-professional counsellors with low job demands, high job resources and a strong sense of coherence were least likely to display burnout symptoms; the role of sense of coherence suggests that the combination of a low level of job demands and high levels of job resources and sense of coherence make these counsellors less prone to develop problems of burnout and loss of engagement. The second hypothesis (according to which sense of coherence would impact on perceived job demands and job resources and the output variables, viz. burnout and engagement) was also borne out.
Stress and Health 24: 3547 (2008) DOI: 10.1002/smi 43

L. Fourie, S. Rothmann and F. J. R. van de Vijver

Discussion This study set out to investigate the effects of job demands and resources, and sense of coherence on the burnout and work engagement of nonprofessional counsellors in South African banks, using a cross-sectional survey design. The sample consisted of 165 non-professional counsellors doing trauma counselling in three of the major banks in South Africa. The participants reported levels of work wellness that were relatively high compared to those found in other occupations in South Africa. A good t was found for a model in which perceived job characteristics mediate the relation between sense of coherence and output variables (viz. burnout and engagement). Perceived job characteristics (job demands and job resources) predict burnout and work engagement of non-professional counsellors. Sense of coherence played an important role in the link between job characteristics and output variables. It had both a direct and indirect effect (through perceptions of job characteristics) on burnout and work engagement. Non-professional counsellors with a strong sense of coherence were found to experience less burnout and more work engagement, presumably because stimuli from the environment are perceived as making cognitive sense, as under control of both the counsellor and signicant others, and as motivationally relevant and meaningful. The results of this study conrmed previous results (Schaufeli & Bakker, 2004; Schaufeli et al., 2002b), which indicated that burnout consists of emotional exhaustion and depersonalization while engagement includes vigour, dedication, absorption and personal accomplishment. Although this patterning in two factors has a conceptual underpinning, the impact of method factors should not be underrated. Engagement and personal accomplishment are measured by positively worded items, while emotional exhaustion and depersonalization are measured by negatively worded items. This similarity could easily lead to an overestimation of the real correlation of items with similar wordings (e.g. negative statements) and to an underestimation of the real correlation of items with a different format. A relationship between burnout and engagement was also found. Low levels of burnout were related to high levels of engagement. Therefore, the ndings of Schaufeli et al. (2002b) that burnout and engagement are related but distinct concepts were conrmed in this study. However,
44 Copyright 2007 John Wiley & Sons, Ltd.

our results showed that burnout and work engagement are highly related components of the same underlying factor, namely work wellness. We hypothesized that high job demands and a lack of job resources contribute to burnout, while job resources predict work engagement of nonprofessional counsellors. The results showed that both job demands and job resources contributed to burnout and work engagement. However, job demands and job resources correlated strongly in this study. Employees who felt that they had to face high job demands (e.g. having to attend to their jobs in addition to counselling and having too much paper work), also felt that they lacked job resources (e.g. support from their supervisors for their counselling inputs, lack of compensation and inadequate access to debrieng). Although previous studies showed that job demands and job resources are two separate (related) factors (e.g. Demerouti et al., 2001; Schaufeli & Bakker, 2004), these variables were strongly related in the present study. The high correlation between job demands and job resources could be attributed to a third variable, such as stress because of a dual roleas employee in a bank as well as a non-professional counsellor. In the banks, non-professional counsellors work in a matrix type of organizational structure: in their real jobs they report to a line manager, but they also report to a people manager concerning their counselling duties. They often experience stress when they perceive that line managers expect much of them in their jobs in addition to their counselling duties. Furthermore, line managers who expect much might show little consideration for non-professional counsellors because they are not responsible for counsellors performance, rewards and support. Also, crises (e.g. bank robberies) come at unpredictable times, and if they occur, non-professional counsellors might experience that counselling demands (in addition to their usual job demands) are high and that they lack sufcient resources and support. The structural model conrms previous ndings that burnout is caused by job characteristics (i.e. high job demands and low job resources) and a weak sense of coherence (Demerouti et al., 2001; Maslach & Leiter, 1997; Schaufeli & Bakker, 2004). Sense of coherence had main effects on job characteristics and work wellness (low burnout and high work engagement). Therefore, the effect of job demands and resources on burnout depends on the level of sense of coherence of the non-professional counsellor. Job
Stress and Health 24: 3547 (2008) DOI: 10.1002/smi

Work wellness
demands, a lack of resources, and a weak sense of coherence predicted 79 per cent of the variance in the burnout and work engagement of nonprofessional counsellors. Counsellors with a strong sense of coherence experience less burnout and more work engagement, presumably because stimuli from the environment are perceived as making cognitive sense (comprehensible), as under control of both the counsellor and legitimate others (manageable), and as motivationally relevant and meaningful (meaningfulness). This study found that sense of coherence is negatively related to burnout, which conrms the results of Basson and Rothmann (2002) and Wissing, de Waal, and De Beer (1992). Sense of coherence was positively related to work engagement and professional efcacy. It is possible that a strong sense of coherence, as a meaning-providing variable (Strmpfer, 2002), may assist to strengthen work engagement and professional efcacy. Furthermore, counsellors need assurance that they will be provided with the necessary resources and support networks that will enable them to become involved with the issues affecting deeply traumatized people. These needs should be met through a carefully devised, ongoing care plan that supports and informs counsellors. Our results are in line with Ortlepp and Friedman (2002) who found relatively low levels of secondary traumatic stress and burnout of nonprofessional counsellors. It seems that although counsellors might work in unfavourable conditions and are exposed to stress, they still do not display elevated levels of burnout and/or low levels of work engagement. There are two possible explanations for this nding. Firstly, selfselection might play a role; healthier and more resilient individuals with more resources in their personal and professional background may choose to become non-professional counsellors. Secondly, it was clear that the level of sense of coherence of non-professional counsellors in this study was above average. Individuals with a high sense of coherence (compared with those with a low sense of coherence) might nd the nature of stimuli which confront them as more comprehensible, manageable and meaningful, which may have had an impact on burnout and work engagement levels. The present study had various limitations. Firstly, the research design was a cross-sectional survey design, which makes it impossible to prove causal relationships even when, as done here,
Copyright 2007 John Wiley & Sons, Ltd.

advanced analytical procedures such as structural equation modelling techniques are employed. In future studies, longitudinal or experimental designs should be used to investigate causal relationships. Secondly, the sample had a strong overrepresentation of females working in administrative functions. Studying people in a homogeneous group has disadvantages as well as advantages (De Jonge et al., 2001). The disadvantage is not enough variance on the variables of interest might be obtained to allow for hypotheses testing. An advantage of a homogeneous group is that there is little variance in confounding effects such as socio-economic status. Nevertheless, the external validity of our ndings remains to be tested. In future research, the relationships between sense of coherence, job demands and resources, and work wellness should be studied in samples which are representative in terms of gender and occupation. Thirdly, this study exclusively relied on self-reporting, which could lead to common method variance (Schaufeli, Enzmann, & Girault, 1993). Fourthly, methodological aspects such as the sample size as well as the non-probability sampling method could challenge the external validity of the current ndings. Recommendations Job demands and a lack of job resources as a cause of burnout and low work engagement of non-professional counsellors should be addressed. It is probably difcult for counsellors to address demands from their counselling role in addition to their normal workload. Supervisors and counsellors should become aware of the causes and symptoms of burnout, as well as certain management actions that could alleviate the burnout of non-professional counsellors. Counsellors should be trained in the use of stress reduction techniques and coping skills to facilitate their ability to manage stressors. Sense of coherence is another target for intervention to prevent and/or manage burnout of non-professional counsellors. One possibility is to contribute to the development of a strong sense of coherence of non-professional counsellors. They should be able to appreciate their counselling role within the framework of the organization; a good insight in the value of their job may enhance the sense of coherence. With a degree of independence and freedom of choice in the counStress and Health 24: 3547 (2008) DOI: 10.1002/smi 45

L. Fourie, S. Rothmann and F. J. R. van de Vijver

selling role, counsellors will regard their work as meaningful. Participation in decision making may well enhance the counsellors feeling of belonging and contribute to the meaningfulness component of sense of coherence. Future research should focus on sense of coherence, burnout and work engagement in other counselling and caregiver environments in South Africa (e.g. HIV/AIDS counselling).
References ABSA Security (2003). Security incidents: Financial and risk exposure for the nancial year 20022003. Johannesburg, South Africa. Albertsen, K., Nielsen, M.L., & Borg, V. (2001). The Danish psychosocial work environment and symptoms of stress: The main, mediating and moderating role of sense of coherence. Work and Stress, 15, 241253. Amirkhan, J.H., & Greaves, H. (2003). Sense of coherence and stress: The mechanics of a healthy disposition. Psychology and Health, 18(1), 3162. Antonovsky, A (1987). Unraveling the mystery of health: How people manage stress and stay well. San Francisco, CA: Jossey-Bass. Antonovsky, A. (1993). The structure and properties of the sense of coherence scale. Social Science and Medicine, 36, 725733. Arbuckle, J.L. (1997). Amos users guide version 3.6. Chicago, IL: Smallwaters. Bagozzi, R.P., & Heatherton, T.F. (1994). A general approach to representing multifaceted personality constructs: Application to state self-esteem. Structural Equation Modeling, 1, 3567. Basson, M.J., & Rothmann, S. (2002). Sense of coherence, coping and burnout of pharmacists. South African Journal of Economic and Management Sciences, 5, 3562. Beaton, R.D., & Murphy, S.A. (1995). Working with people in crisis: Research implications. In C.R. Figley (Ed.), Compassion fatigue: Coping with secondary traumatic stress disorder in those who treat the traumatized (pp. 120). New York: Brunner/Mazel. Byrne, B.M. (2001). Structural equation modeling with AMOS: Basic concepts, applications and programming. Mahwah, NJ: Erlbaum. Cordes, C.L., & Dougherty, T.W. (1993). A review and integration of research on job burnout. Academy of Management Review, 18, 621656. De Jonge, J., Dormann, C., Janssen, P.P.M., Dollard, M.F., Landeweerd, J.A., & Nijhuis, F.J.N. (2001). Testing reciprocal relationships between job characteristics and psychological well-being: A cross-lagged structural equation model. Journal of Occupational and Organizational Psychology, 74, 2946. Demerouti, E., Bakker, A.B., Nachreiner, F., & Schaufeli, W.B. (2001). The job demands-resources model of burnout. Journal of Applied Psychology, 86, 499512. Feldt, T., Kinnunen, U., & Mauno, S. (2000). A mediational model of sense of coherence in the work context: A one-year follow-up study. Journal of Organizational Behavior, 21, 461476. Feldt, T., Kivimki, M., Rantala, A., & Tolvanen, A. (2004). Sense of coherence and work characteristics: A cross-lagged structural equation model among managers. Journal of Occupational and Organizational Psychology, 77, 323 342. Figley, C.R. (1995). Compassion fatigue as secondary traumatic stress disorder: An overview. In C.R. Figley (Ed.), Compassion fatigue: Coping with secondary traumatic stress disorder in those who treat the traumatized (pp. 1 20). New York: Brunner/Mazel. Fourie, L.A. (2003). Burnout and work engagement of nonprofessional counsellors in South Africa. Unpublished doctoral thesis, Potchefstroom University for CHE, Potchefstroom, South Africa. Gilbar, O. (1998). Relationship between burnout and sense of coherence in health social workers. Social Work in Health Care, 26(3), 3949. Hakanen, J.J., Bakker, A.B., & Demerouti, E. (2005). How dentists cope with their job demands and stay engaged: The moderating role of job resources. European Journal of Oral Sciences, 113, 19. Jenkins, R.A., & Baird, S. (2002). Secondary traumatic stress and vicarious trauma: A validational study. Journal of Traumatic Stress, 15, 423432. Kahn, W.A. (1990). Psychological conditions of personal engagement and disengagement at work. Academy of Management Journal, 33, 692724. Kinman, G., & Jones, F. (2003). Running up and down the escalator: Stressors and strains in UK academics. Quality in Higher Education, 9, 2138. Levert, T., Lucas, M., & Ortlepp, K. (2000). Burnout in psychiatric nurses: Contributions of the work environment and a sense of coherence. South African Journal of Psychology, 30, 3643. Little, T.D., Cunningham, W.A., Shahar, G., & Widaman, K.F. (2002). To parcel or not to parcel: Exploring the question, weighing the merits. Structural Equation Modeling, 9, 151173. Louw, A., Shaw, M., Cameren, L., & Robertshaw, R. (1998, February). Crime in Johannesburg: Results of a city victim survey (Monograph No. 18). Johannesburg: Institute. of Security Studies. Maslach, C., & Jackson, S.E. (1986). Maslach Burnout Inventory (2nd ed.). Palo Alto, CA: Consulting Psychologists Press. Maslach, C., Jackson, S.E., & Leiter, M. (1996). Maslach Burnout Inventory manual (3rd ed.). Palo Alto, CA: Consulting Psychologists Press. Maslach, C., & Leiter, M.P. (1997). The truth about burnout: How organizations cause personal stress and what to do about it. San Francisco, CA: Jossey Bass. Maslach, C., & Schaufeli, W.B. (1993). Historical and conceptual development of burnout. In W.B. Schaufeli, C. Maslach, & T. Marek (Eds), Professional burnout: Recent developments in theory and research (pp. 116). Washington, DC: Taylor and Francis. Maslach, C., Schaufeli, W.B., & Leiter, M.P. (2001). Job burnout. Annual Review of Psychology, 52, 397422. Nakamura, J., & Csikszentmihalyi, M. (2002). The concept of ow. In C.R. Snyder, & S.J. Lopez (Eds), Handbook of positive psychology (pp. 715728). Oxford, UK: Oxford University Press. Naud, J.L.P., & Rothmann, S. (2004a). The validation of the Maslach Burnout InventoryHuman Services Survey for emergency health technicians in Gauteng. South African Journal of Industrial Psychology, 30(3), 2128. Naud, J.L.P., & Rothmann, S. (2004b). The validation of the Utrecht Work Engagement Scale for emergency health tech-

46

Copyright 2007 John Wiley & Sons, Ltd.

Stress and Health 24: 3547 (2008) DOI: 10.1002/smi

Work wellness
nicians in Gauteng. South African Journal of Economic and Management Sciences, 7, 459468. Nelson, D.L., & Simmons, B.L. (2003). Health psychology and work stress: A more positive approach. In J.C. Quick, & L.E. Tetrick (Eds), Handbook of occupational health psychology (pp. 97119). Washington, DC: American Psychological Association. Nunnally, J.C., & Bernstein, I.H. (1994). Psychometric theory (3rd ed.). New York: McGraw-Hill. Ortlepp, K., & Friedman, M. (2002). Prevalence and correlates of secondary traumatic stress in workplace lay trauma counselors. Journal of Traumatic Stress, 15, 213222. Rothmann, S. (2005, September). Work-related well-being in South African organisations: What do we know? Paper presented at the 7th Annual Conference of the Employee Assistance Professionals Association of South Africa, Durban, South Africa. Rothmann, S., Malan, A.M., & Rothmann, J.C. (2001, August). Sense of coherence coping and burnout in a corporate pharmacy group. Paper presented at the 7th Annual PsySSA Conference, Johannesburg, South Africa. Rothmann, S., Steyn, L.J., & Mostert, K. (2005). Job stress, sense of coherence and work wellness in an electricity supply organisation. South African Journal of Business Management, 36(1), 19. Salovey, P., Rothman, A.J., Detweiler, J.B., & Steward, W.T. (2000). Emotional states and physical health. American Psychologist, 55, 110121. Schaufeli, W.B., & Bakker, A.B. (2004). Job demands, job resources, and their relationship with burnout and engagement: A multi-sample study. Journal of Organizational Behavior, 25, 123. Schaufeli, W.B., Martinez, I., Pinto, A.M., Salanova, M., & Bakker, A.B. (2002a). Burnout and engagement in university students: A cross-national study. Journal of Cross-Cultural Psychology, 33, 464481. Schaufeli, W.B., Salanova, M., Gonzlez-Rom, V., & Bakker, A.B. (2002b). The measurement of engagement and burnout: A conrmation analytic approach. Journal of Happiness Studies, 3, 7192. Schaufeli, W.B., & Enzmann, D. (1998). The burnout companion to study and practice: A critical analysis. London: Taylor & Francis. Schaufeli, W.B., Enzmann, D., & Girault, N. (1993). Measurement of burnout: A review. In W.B. Schaufeli, C. Maslach, & T. Marek (Eds), Professional burnout: Recent developments in theory and research (pp. 199215). Washington, DC: Taylor & Francis. Schlemmer, L. (1996, April). National survey on crime. In L. Schlemmer (Ed.), The Nedcor project on crime, violence, and investment (pp. 122). Johannesburg, South Africa: Nedcor Project. Schutte, N., Toppinen, S., Kalimo, R., & Schaufeli, W.B. (2000). The factorial validity of the Maslach Burnout Inventory-General Survey (MBI-GS) across occupational groups and nations. Journal of Occupational and Organizational Psychology, 73, 5366. SPSS Inc. (2003). SPSS 12.0 for Windows. Chicago, IL: Author. Strmpfer, D.J.W. (2002, March). Resilience and burnout: A stitch that could save nine. Paper presented at the 1st South African Burnout Conference, Potchefstroom, South Africa. Watts, W.D., Cox, L., Wright, L.S., Garrison, J., Herkimer, A., & Howze, H.H. (1991). Correlates of drinking and drug use by higher education faculty and staff: Implications for prevention. Journal of Drug Education, 21, 4364. Wilson, K.L. (1998). An exploratory study into secondary traumatic stress in trauma/crisis counsellors and the perceived effectiveness of supervision as a debrieng process and coping mechanism. Unpublished masters dissertation, University of the Witwatersrand, Johannesburg, South Africa. Wissing, M.P., de Waal, M., & de Beer, I. (1992, July). Sense of coherence and stress symptomatology. Paper presented at the 25th International Congress of Psychology, Brussels, Belgium.

Copyright 2007 John Wiley & Sons, Ltd.

Stress and Health 24: 3547 (2008) DOI: 10.1002/smi

47