Engagement vs. Burnout: An examination of the relationships between the two concepts within the framework of the JDR model

Transcription

1 ESADE WORKING PAPER Nº 234 July 2012 Engagement vs. Burnout: An examination of the relationships between the two concepts within the framework of the JDR model Scott Moodie Simon L. Dolan Ronald J. Burke

2 ESADE Working Papers Series Available from ESADE Knowledge Web: ESADE Avda. Pedralbes, E Barcelona Tel.: ISSN Depósito Legal: B

3 Engagement vs. Burnout: An examination of the relationships between the two concepts Scott Moodie PhD Candidate in Management Science ESADE Business School, Ramon Llull University, Spain Simon L. Dolan Professor and Future of Work Chair ESADE Business School, Ramon Llull University, Spain (corresponding author) Ronald J. Burke Professor Emeritus, Schulich School of Business, York University (Canada) and Associate Professor at the Future of Work Chair (ESADE Business School) July 2012 Abstract The precise relationship between the positive psychological state of work (i.e. engagement) and the negative psychological state (i.e. burnout) has been receiving an increased attention. Some view these as opposite states on the same or similar continuum, while others take the position that they represent different biobehavioral spheres. Both states exhibit significant correlations to job demands and resources, elements of physical and mental wellbeing, and to each other. This study expands our knowledge of the phenomena of engagement and burnout by analyzing their separate and joint manifestations. Using a large sample of 2,094 nurses, respondents were segmented into quadrants that represent a 50/50 (median split) of Engagement and Burnout. The four resulting quadrants were examined in a series of analyses including logistic regression and ANOVAs. This configurational approach allowed us to examine both inverse and concurrent states of Engagement and Burnout. The findings suggested that engagement and burnout were generally inversely related (67% of the sample) but could be manifested concurrently at either extreme (33% of the sample). Burnout was chiefly driven by work demands as both quadrants of low burnout had lower demands and both quadrants of high burnout had higher demands. Engagement was primarily driven by resources and affinity. Social support acted independently by aligning with states of burnout. Worker health was primarily driven by burnout wherein both states of low burnout exhibited better health and both states of high burnout exhibited poorer health. Keywords: Engagement, Burnout, JD-R, Affinity, Health, Social Support Short Title: Exploring the multiple linkages between Work Engagement and Burnout 3

4 Notes and Acknowledgements 1. A version of this paper has been presented at the Academy of Management meeting in Boston (August 2012). 2. This work has been supported in part by the MEC (Spanish Ministry of Education and Science) SEJ and AGAUR - Generalitat de Catalunya. We also wish to express our gratitude to the Col legi Oficial d'infermeres de Barcelona for their instrumental assistance. Corresponding author contact: Simon L. Dolan ESADE Business School, Av. Torre Blanca, 59 Sant Cugat, Spain simon.dolan@esade.edu Tel

5 Introduction Researchers in the fields of organizational psychology and organizational behavior have examined ways in which positive or negative perceptions and attitudes are linked to performance at work. The subjective or perceptual measurement of indicators such as satisfaction and motivation have a long history of use in organizational research but increasingly have been criticized for being insufficient to predict performance outcomes on both conceptual or methodological grounds (Brief & Weiss, 2002; Latham & Pinder, 2005; Wright, 2006; van Saane, Sluiter, Verbeek, & Frings-Dresen, 2003). They have been replaced by measurements attempting to capture more objective, work-related states of mind. Two of the latter include the constructs of burnout and engagement (Bakker, Schaufeli, Leiter, & Taris, 2008). Burnout is a negative construct contributing to decreased job satisfaction and organizational commitment, and increasing undesired outcomes such as turnover and absenteeism (Lee & Ashforth, 1996). Engagement is a positive indicator characterized by vigor, dedication, and absorption (Schaufeli, Salanova, González-Romá, & Bakker, 2002). Vigor is characterized by high levels of energy and mental resilience at work. Dedication is a strong sense of involvement with one s work along with a sense of significance, enthusiasm, and challenge. Absorption occurs when an individual is fully concentrated and happily engrossed in their work to the extent that time passes quickly and they have difficulty detaching themselves from their work (Bakker & Demerouti, 2008; Schaufeli & Bakker, 2004). Workers that scored high on engagement have been shown to exhibit high energy and self-efficacy (Schaufeli et al., 2001). The positive nature of work engagement leads workers to create their own positive feedback in terms of appreciation, recognition, and success (Bakker & Demerouti, 2008). Bakker and Demerouti (2008) found that engaged workers carried their enthusiasm and energy with them outside of the organization and felt a sense of accomplishment at the end of the work day. Schaufeli & van Rhenen, (2006) reported a strong connection between positive emotions and engagement. Engagement has also been shown to be positively related to health (Schaufeli, Taris, & van Rhenen, 2008). Furthermore, 5

6 engaged workers reported fewer psychosomatic complaints than co-workers who displayed low levels of engagement (Demerouti, Bakker, De Jonge, Janssen, & Schaufeli, 2001). Schaufeli & Bakker, (2004) observed that engaged workers suffered from fewer self-reported headaches, cardiovascular problems, and stomach aches. Different approaches to the conceptualization of burnout have at their core low levels of physical and emotional energy (Shirom, 2005). Burnout represents a chronic condition that remains stable over time (Taris et al., 2005; Kristensen et al., 2005; Halbesleben & Demerouti, 2005). Several studies indicated that aspects of the job environment were stronger predictors of burnout than were personality factors (Lee & Ashforth, 1996; Schaufeli & Enzmann, 1998). However, research also suggested that there was a link between emotional exhaustion, as operationalized within a burnout framework, and family-related factors (Bakker et al., 2005). This connection between low physical and emotional energy was primarily dependent upon social aspects of the work environment and was distinct from chronic states of depression that represented a propensity towards negative affective states (Shirom, 2005; Schaufeli & Enzmann, 1998). Burnout has been also linked to a variety of health complaints including circulatory and heart problems, muscular pains, sleep disturbances, headaches, and gastro-intestinal problems (Gorter et al., 2000; Kahill, 1988; Westman and Bakker, 2008) The predominant formulations of burnout and engagement bear a number of similarities which can create confusion concerning the manner in which they relate to each other. At the most basic level, constructs evaluating engagement have been linked to a set of positive emotions (Schuafeli & Van Rhenen, 2006) whereas burnout constructs wwere closely linked to a set of corresponding negative emotions (Schaufeli & Enzmann, 1998). Some researchers have positioned burnout and engagement at opposite ends of a single dimension (Demourouti and Bakker, n.d). For example, The Oldenburg Burnout Inventory (OLBI; Demerouti et al., 2001, 2003) is a tool which measures burnout by categorizing respondent s levels of exhaustion and disengagement thereby effectively placing burnout and engagement at opposite ends of the same scale. 6

7 The Dutch UWES model connected with the pioneering work of Bakker and Schaufeli conceptualizes engagement according to states of vigor, dedication, and absorption. The UWES measures of vigor and dedication are regarded as polar opposites of the most frequently used measure used in burnout research, the Maslach Burnout Inventory -MBI (Maslach & Jackson 1981). The central measurers of the MBI are exhaustion and cynicism. However, the UWES ignores the MBI dimension of reduced personal efficacy and replaces it with the concept of absorption (Schaufeli & Bakker, 2003). According to Schaufeli & Bakker, (2003), this modification was made in part because studies had demonstrated that lack of professional efficacy was less determinate of burnout than exhaustion and cynicism (Maslach et al., 2001; Shirom, 2002). A second reason for the addition of absorption to the UWES resulted from qualitative findings suggesting that engagement was characterized by being immersed and happily engrossed in one s work (Schaufi & Bakker, 2003). Shirom, (2003) argued that vigor (engagement) and burnout were obliquely related and were thus not positioned at opposite ends of the same continuum. One of his arguments ass that vigor and burnout are subjective components of different biobehavioral systems (Watson et al., 1999). By adapting this theory, Shirom, (2003) posited that burnout may be an antecedent of an internal withdrawaloriented behavioral inhibition system whereas vigor is a component of an approach-oriented behavior facilitation system. This argument presented the possibility that both burnout and vigor could be operationalized in the same context since many job situations present both demands and opportunities for rewards. In this way, demanding situations may evoke burnout, but the anticipated rewards of completion may simultaneously evoke vigor (Dweck & Legget, 1988). Joint and Separate Determinants and consequences of Work Engagement and Burnout. The origins of both burnout and engagement are multi-factorial and multi-faced (Shirom, 2005). Possible sources can include individual differences and personalities, organizational demands and environmental factors, and cultural 7

8 factors that interact in specific configurations (Diez-Pinol, Dolan, Sierra, & Cannings, 2008). To develop a broader and more complete view of the relationship between work engagement and burnout, it is important to explore them together to determine the joint influence that antecedents have on separate and concurrent states of Engagement and Burnout. Thus, this study employs frequently used measures of Engagement and Burnout and uses a configurational approach dividing respondents into quartiles based on the presence of Low Engagement and Low Burnout (LELB), Low Engagement and High Burnout (LEHB), High Engagement and Low Burnout (HELB), and High Engagement and High Burnout (HEHB). These quartiles were then examined using personal and organizational determines nested within the larger framework of the J-DR model (Bakker & Demerouti, 2007; Demerouti et al., 2001) in an attempt to best predict the quartiles along with the comparative states of physical and mental health of its members (Table 1 about here) Study objectives and research questions This study expands on our knowledge of the phenomena of Engagement and Burnout by analyzing their separate and joint manifestations. Thus, as explained below, the respondents were segmented into quadrants that represented a 50/50, or median split, on both Engagement and Burnout.The four resulting quadrants include those with A. Low Engagement and Low Burnout (LELB); B. Low Engagement and High Burnout (LEHB); C. High Engagement and Low Burnout (HELB); and D. High Engagement and High Burnout (HEHB). This classification enabled the empirical testing that shows which combination of antecedents can best predict each manifestation, and what consequences are exhibited on individual health and wellbeing. This is the essence of what has been referred to in this study as a configurational approach. 8

9 More specifically, the following research questions represent the core for the present study: 1. Are Engagement and Burnout inversely or concurrently related? 2. What is the configuration of individual and organizational variables that predict states of A. Low Engagement and Low Burnout? B. Low Engagement and High Burnout? C. High Engagement and Low Burnout? D. High Engagement and High Burnout? 3. What is the effect of each joint configuration of Engagement and Burnout on Physical and Mental Health? To address these questions, the study was positioned into the broader J-DR research model (Bakker & Demerouti, 2007; Demerouti et al., 2001). Respondents were nurses employed in multiple regions and multiple hospitals in Spain.. Nurses have been studied extensively due to the worsening of their respective working conditions and the alarming rates of burnout reported worldwide (Gilbert et al, 2010, Leiter et al, 2009; Leiter et al, 2010). Studying the nursing population can be instrumental for comparative reasons and enhance the external validity of the study. Personal and organizational factors linked to engagement and burnout were divided into four categories: Job Demands, Job Resources, Social Support, and Affinity. Job demands and resources have a long established connection to both engagement and burnout. The JD-R model proposes that burnout arises from situations where many demands are made without the provision of sufficient resources to meet those demands (Demerouti, Bakkar, Nachreiner, & Schaufeli, 2001). The conservation of resources theory suggests that burnout increases when valued resources are lost, are unable to meet job requirements, or are insufficient (Hobfoll & Freedy, 1993). Structural equation modeling in the Schaufeli et al, (2009) study found that increases in demands and decreases in 9

10 resources predicted burnout and that increases in resources predicted engagement. This study also concluded that resources and engagement were reciprocal in that work engagement lead to an increase in resources which subsequently increased the level of engagement. A meta-analysis by Crawford et. al (2010) found that demands and burnout were positively associated and that resources and burnout were negatively associated. With respect to engagement, they found that resources and engagement were positively associated but the association between demands and engagement varied according to the nature of the demand. Demands that were considered to be hindrances were negatively associated with engagement whereas demands considered to be challenges had a positive association with engagement. Another meta-analysis conducted by Halbesleben et al, (2010) found that engagement and burnout constructs were negatively associated albeit with a few exceptions. Engagement was positively related to resources and negatively related to demands with resources having a stronger relationship. Within the JD-R model, the commonly used measure of social support measures cooperation among colleagues. For the present study, we replaced this measure with a measure of social support that assesses support at three levels: supervisor support, colleague support, and spousal support. A meta-analysis by Halbesleben (2011) found that work-related social support was more closely associated with the exhaustion component of burnout than non-work sources of social support. Shaufeli et al.,(2008) reported that co-worker support was negatively associated with the reduced professional efficacy dimension of the MBI burnout measure and positively associated with the dedication dimension of Engagement and that supervisor support was negatively associated with the exhaustion and cynicism dimensions of burnout. Affinity refers to the long-term affective evaluation of an employee to their work and includes the aspects of Affective Commitment and Job Involvement. Organizational Commitment and Job Involvement have been found to be discriminately different constructs from work engagement (Hallberg & Shaufelli, 10

11 2006). Affective commitment is one of the dimensions of organizational commitment as proposed by Allen & Meyer (1990). Affective organizational commitment is more closely related to job characteristics than to personal factors and therefore related to extrinsic circumstances of the work (Morrow, 1983, Hallberg & Shaufeli, 2006). Organizational commitment has been found to be inversely associated with the cynicism and reduced professional efficacy dimensions of burnout and positively associated with the dedication and absorption dimensions of work engagement (Schaufeli et al., 2008). Job involvement has been defined as a cognitive belief state of psychological identification with one s job (Kanungo, 1982, p. 80). This relatively stable attitude (Kunel et al., 2009) is different from short-term states of engagement which can change from day to day (Dalal, Brummel, Wee, & Thomas, 2008). Job Involvement has been shown to have a direct positive effect on Engagement among nurses (Kunel et al., 2009). Griffen et al, (2010) found that job involvement was positively related to the emotional exhaustion dimension of burnout. Present states of physical and mental health were evaluated by assessing general physical health, anxiety and depression. As noted above, both engagement and burnout are measures of employee wellbeing and both have been extensively linked to health. It is common for people suffering from stress to engage in behaviors such as smoking, alcohol abuse, or eating unhealthy food that may alleviate short term effects, but have long-term negative health consequences (Scwarzer & Fuchs, 1995). Several studies have confirmed that health is closely correlated with burnout. (Kahill, 1988; Soderfeldt et al., 2000). Engagement is linked to positive emotions and better health (Bakker et al., 2008). While there is a need for further research on the specific paths between engagement and health, a recent study found that patients who demonstrated positive affect had a reduction in 10-year incidents of coronary heart disease (Davidson et al., 2010). A summary of the general hypothesized relationships of these categories with regards to Engagement and Burnout can be found in Table 2 11

12 (Table 2 about here) Sample Methods and procedures Respondents Nurses must be registered and qualified within a regional association to practice nursing n Spain.. This study was carried out with the support of these associations in selected regions in Spain. An online survey was prepared, pre-tested and prevalidated, and sent to the regional associations for distribution among their members. The total recipient number could not be determined but 2,115 surveys were completed online. Of these, 21 surveys were removed from the dataset due to spurious data or omissions leaving 2094 complete responses. The majority of responses were received from the provinces of Catalunya and Gipuzkoa. The respondents were 90.3% female, had a mean age of 39 years and had worked as a nurse for an average number of 16 years. Of these, 72% reported living with a partner and 85% of respondents worked full time as a nurse. Measures The central measures of the study were the shortened (Schaufeli et al., 2006) Utrecht Work Engagement Scale (UWES) developed by Schaufeli and Bakker (2003) and the Shirom-Melamed Burnout Measure (SMBM, Shirom et al., 2005). The UWES is based upon a three dimension model of engagement which includes vigor, dedication, and absorption (Schaufeli et al., 2002) whereas the SMBM assesses the depletion of an individual s energetic resources at work by measuring the three dimensions of physical fatigue, emotional exhaustion, and cognitive weariness. 12

13 Job Demands and Resources were investigated by employing measures used in the Job Demands Resources (JD-R) model of Bakker et al. (2003) which has been employed in many other studies on engagement (i.e. Xanthopoulou et al., 2007a, b) and burnout (Bakker et al., 2004). While the model seems to be dynamic and in a continued state of evolution ( Demerouti & Bakker, 2011), some core components were present across various versions of it. For example, job demands refer to physical, psychological, social, or organizational aspects of the job that require sustained psychological or physical effort (Bakker et al, 2004). This model examines job demands according to three categories: workload, emotional demands, and work-home conflict. Job Resources refer to physical, psychological, social, or organizational aspects of the job that are necessary to achieve goals, reduce job demands, and stimulate personal growth and development (Bakker et al., 2004). Of the original JD-R measures, the scales for Autonomy and Self Development Opportunities were used. The JD-R measure of Social Support that measures a single factor of social support was replaced with the scale of Dolan et al (1992a, 1992b) which assesses support along the axes of Supervisor Support, Colleague Support, and Spousal Support. Affinity towards work was assessed using measures of Affective Commitment and Job Involvement. The inclusion of these variables permits an assessment of perceived affiliation and identification with the work and workplace. Affective Commitment was assessed using the measures of Meyer, Allen and Smith, Job Involvement was assessed using a 4 item version from Frone and Rice (1987) which was based on the original measure by Kanungo (1982). Employee health and wellbeing was measured by assessing General Health, Anxiety, and Depression. General health was assessed through a single item which asked respondents to evaluate their general state of health varying from very poor and very good (Benyamini & Idler, 1999). The single-item construct has been found to be a valid predictor of all-cause morbidity and mortality (DeSalvo et al., 2006a) and has demonstrated comparable reliability and validity when 13

14 compared to other multi-item measures used by health professionals (DeSalvo et al., 2006b). Mental health was measured via the anxiety and depression scales developed by Dolan and Arsenault (1983). The variables used in the study, including their respective psychometric properties are presented in Table (Table 3 about here) Analysis Strategy The data was analyzed in stages which included zero-item correlations and ANOVAs to determine their independence and suitability for predicting states of Engagement and Burnout. The Engagement and Burnout scores were binned using a 50/50 median split and then merged to reflect four distinct groups representing Low Engagement and Low Burnout, Low Engagement and High Burnout, High Engagement and Low Burnout, and High Engagement and High Burnout. A series of Binary Logistic Regression analyses were undertaken to determine the relative influence of the independent variables as predictors of each state of Engagement and Burnout after controlling for Age, Gender, and Part Time/Full Time work status. Explained variance was measured with the Nagelkerke Pseudo R 2 score. A series of ANOVA analyses were then conducted to determine whether the independent variables demonstrated a significant difference within each group when compared to the remaining population. Results The purpose of this study was to identify whether the states of Engagement and Burnout are manifested inversely or can be manifested concurrently. An effort was also made to illustrate these conditions by identifying the primary antecedents of each condition. In addition, levels of physical and mental health found among each configuration were analyzed. 14

15 Segmentation of respondents according to levels of Engagement and Burnout revealed that 67% of the population experienced states of either Low Engagement and High Burnout or High Engagement and Low Burnout. The remaining 33% of the population experienced states of either Low Engagement and Low Burnout or High Engagement and High Burnout. Thus while the majority of workers exhibited one or the other dominant states, a sizable percentage of workers were functioning in one of the extremes (Tables 4 and 5 about here) The group defined by states of Low Engagement and Low Burnout was best predicted by all three Demands and Job Involvement (see results of Logistic Regression in Table 4). Results of the ANOVA analysis (see Table 5) indicated that when compared to the remaining population, these nurses exhibited significantly lower Demands and Affinity and had higher levels of Colleague Support. These nurses demonstrated better overall health on all measures. This group has been labeled Loafers to indicate their relative lack of demands, normal levels of resources and support and their apathy towards the work. The group defineded by states of Low Engagement and High Burnout was best predicted by Emotional Demands, Work Interfering with Home, Self Development Opportunities, Colleague and Spousal Support, and the two measures of Affinity. Results of the ANOVA analysis indicated that this group of nurses demonstrated significantly higher Demands, lower Resources, lower Social Support, and lower Affinity. They reported a significantly poorer level of health on all measures. This group has been labeled Slaves, to indicate their increased demands and lack of resources, support or personal affinity to the work. The group defined by states of High Engagement and Low Burnout was predicted using all variables except for Autonomy. The ANOVA results revealed that, compared to the remaining population, this group exhibited significantly lower Demands, higher Resources, higher Social Support, and higher Affinity. They also demonstrated significantly better health on all measures. This group has been 15

16 labeled Apprentices to indicate their lower demands and higher levels of resources, support, and affinity to the work. Finally, the group defined by states of High Engagement and High Burnout was best predicted by all measures of Demands, Self Development, Supervisor Support and Job Involvement. The results of the ANOVA analysis indicate that these workers experience higher Demands, higher Self Development Opportunities, lower Social Support, and higher Affinity than their colleagues. They were also marked by comparatively lower health on all measure when compared to their colleagues. We have termed as Lone Rangers to reflect their higher demands, resources, and affinity to the work but lack of social support. Discussion The findings suggested that engagement and burnout were generally inversely related but could be manifested concurrently at either extreme. The explained variance of the dominant quadrants is very satisfactory assuming individual level of analysis (LEHB R 2 =.36, HELB R 2 =.35), but the extreme quadrants are lower (LELB R 2 =.17, HEHB R 2 =.12). This suggests that the primary determinants of these states are external to the working conditions. This may reflect personality types or social and cultural factors that influence their participation in the work environment or their ability to manage resources, demands, or social relationships. Results show that burnout was chiefly driven by work demands as both quadrants of low burnout had lower demands and both quadrants of high burnout had higher demands. This supports the findings of other studies (Xanthopoulou et al, 2007). Engagement was found to be primarily driven by resources and affinity. The relationship between resources and engagement has been amply documented in other studies (Bakker et al, 2007, Crawford et al, 2010). While both job involvement and organizational commitment have been independently examined along with burnout and engagement, this is the first time that they have been combined as dimensions of a single construct we called affinity. This provides 16

17 insight into the long-term attitude toward the work and the identification that workers feel with their job which may be a useful construct in future studies. The relationship of affinity to engagement can be partially explained in the conceptual similarity of these attitudinal states with the UWES construct of dedication which asks respondents several questions related to their identification with their work. Findings pertaining to the influence of social support on burnout and engagement provided an interesting insight. Social support has historically been considered a resource as evidenced by its inclusion as a resource in the JD-R model. Nonetheless, findings in this study do not confirm this assertion. Social support did not behave like the other resources which were associated with states of engagement. Instead, social support acted independently by aligning with states of burnout. Both states of low burnout evidenced higher social support whereas both states of high burnout evidenced lower social support. This suggests that social support should perhaps be regarded as a distinct construct not to be considered as a traditional resource. Worker health was primarily influenced by burnout wherein both states of low burnout exhibited better health and both states of high burnout exhibited poorer health. While there has been an abundance of research linking engagement to positive health (Kanste 2011) and burnout to negative health (Ahola, 2007, Shirom, 2009), the findings of this study suggest that burnout is the greater determinant of health. However, our measures of burnout, anxiety, and depression were highly correlated which may be indicative of an inherent susceptibility of this type of psychometric research to common method variance. The difficulty of effectively isolating these phenomena has been noted by Shirom (2003). Practical Implications There are a number of practical implications that follow from this study. There has been a recent debate about the relative merits of focusing on positive states of engagement or negative states of burnout. Our study suggests that these states 17

18 are not mutually exclusive and that focusing too strongly on either state may distract from a more fundamental need to create a positive and balanced work environment. The strongest predictors in this study were demands and self development opportunities. If management would focus on balancing demands and resources and follow that by fostering social support and encouraging affinity through training and recruitment, the result should be positive regardless of whether it is measured according to burnout or engagement. We are not suggesting that measuring either engagement or burnout is irrelevant. To the contrary, they are both useful measures for assessing an employee s current state as reflective of the work environment. However, if the objective is to improve working conditions and enhance employee wellbeing, it would at times be more appropriate to directly assess demands and resources and find ways to optimize their balance. Beyond that, it is in the best interest of companies and managers to continually seek ways to foster social support and feelings of affinity to the work and workplace. Strengths and Limitations The purpose of this study was to use a straightforward configurational approach to classify the inverse and concurrent manifestations of engagement and burnout. The recent growth in popularity of the concept of work engagement and the JD-R model testify to the importance of these constructs and their relationships. Much of the current research on this topic considers engagement and burnout to be linear dimensions and focuses on building structural models of the precise relationships between variables. That approach is to be encouraged, but there is also a need to jointly deconstruct dimensions and relationships in a tactile manner that can inform future structural models. While the majority of respondents fell in the expected categories of HELB and LEHB, the 33% that exhibited one of the concurrent states strongly suggests that new approaches to the study of these variables should be considered. The secondary benefit of this approach is that these findings (Table 5) can be submitted directly to managers to provide an easily understood approach for assessments and interventions. 18

19 This quantitative simplification is useful in understanding the underlying nature of the dimensions, but it is also prone to numerous susceptibilities. The study was conducted at a single point in time using purely psychometric sources. It is therefore prone to common method variance (Podsakoff et al., 2003). Further, the variables of burnout, anxiety and depression have been previously noted to exhibit multicolinearity (Shirom, 2003). While the sample size was rather large for this type of studies (N=2094), the findings may not be generalizable beyond nurses working in Spain. Future Research Direction This study presents the first step towards the classification of specific joint states of work engagement and burnout. By demonstrating that the clusters have unique compositions of antecedents, it would be useful to study whether the contemporary structural models of variables interact differently in each of the four quadrants. It would also be useful to examine whether the extreme quadrants (HEHB, LELB) are influenced by other variables which could include individual personality types, social functioning ability, values, clear instructions, job control or compensation, among others. One of the primary goals of future research should be the identification of paths between the quadrants, or more specifically, to identify paths to HELB from the other three quadrants. If we can increase the accuracy of the worker classification and the paths between groups, we will be able to provide managers with specific tools for maximizing the well-being and productivity of their workers. 19

20 Burnout High Low Table 1 Configurational segmentation of sample (N = 2094) Engagement Low High (LELB) (HELB) N = 342 (16.3%) N = 731 (34.9%) (LEHB) (HEHB) N = 666 (31.6%) N = 355 (17.0%) 20

21 Table 2 Summary of expected relationships of Engagement and Burnout with study constructs Engagement Burnout Demands - + Resources + - Affinity + - Mental Health + - Physical Health + - Note: + = positive relationship; - = negative relationship 21

22 Table 3 Means (M), Standard Deviations (SD), Internal Consistencies (Cronbach's α) and Zero-Order Correlations of the Study Variables (N = 2094) Variable M SD Items Range α Engagement Burnout ** - 3. Work Overload ** - 4. Emotional Demands **.50 **.51 ** - 5. Work Int. w/ Home **.47 **.31 **.40 ** - 6. Self Dev. Opp ** -.39 ** -.12 ** -.20 ** -.19 ** - 7. Autonomy ** -.31 ** -.21 ** -.22 ** -.19 **.50 ** - 8. Supervisor Support ** -.34 ** -.19 ** -.23 ** -.17 **.36 **.39 ** - 9. Colleague Support ** -.32 ** -.13 ** -.23 ** -.19 **.30 **.29 **.36 ** Spousal Support ** -.17 ** -.04 * -.11 ** -.12 **.11 **.10 **.17 **.21 ** Affective Commitment ** -.30 ** -.12 ** -.16 ** -.10 **.38 **.30 **.34 **.21 **.05 * Job Involvement ** **.06 *.23 **.19 **.08 **.11 ** *.39 ** Anxiety **.58 **.27 **.39 **.42 ** -.25 ** -.25 ** -.19 ** -.20 ** -.13 ** -.12 **.09 ** Depression **.65 **.21 **.37 **.37 ** -.32 ** -.25 ** -.22 ** -.22 ** -.14 ** -.19 ** ** Physical Health n.a..24 ** -.50 ** -.13 ** -.23 ** -.25 **.18 **.17 **.15 **.18 **.18 **.07 ** -.08 ** -.37 ** -.40 ** Note: * p <.05; ** p <.01; *** p <.001.

23 Table 4 Binary Logistic Regression of Engagement and Burnout Quartiles on the Study Variables (β-values) (N= 2094) LELB LEHB HELB HEHB Demands 1. Work Overload -.43***.00.17*.20* 2. Emotional Demands -.37***.56*** -.52***.35*** 3. Work Int. w/ Home -.48***.57*** -.61***.23** Resources 4. Self Development ***.72***.39*** 5. Autonomy Social Support 6. Supervisor Support ** -.21* 7. Colleague Support **.21* Spousal Support *.30** -.06 Affinity 9. Affective Commitment ***.32*** Job Involvement -.33*** -.30***.29***.32*** Explained Variance (R 2 ) Note: * p <.05; ** p <.01; *** p <.001. Explained Variance is measured with the Nagelkerke Pseudo R 2 23

24 High Burnout Low Burnout Table 5 Synoptic ANOVA analysis for each quadrant compared to the remaining population (N= 2094) Low Engagement Loafers N = 342 (16.3%) Work Overload *** Emotional Demands *** Work Int. w/ Home *** Colleague Support * Demands: Resources: Avg. Social Support: ~ Affective Commitment ** Affinity: Job Involvement *** Anxiety *** Wellbeing: Depression *** Health *** Slaves N = 666 (31.6%) High Engagement Apprentices N = 731 (34.9%) Work Overload *** Demands: Emotional Demands *** Work Int. w/ Home *** Self Development *** Resources: Autonomy *** Supervisor Support *** Social Support: Colleague Support *** Spousal Support *** Affective Commitment *** Affinity: Job Involvement *** Anxiety *** Wellbeing: Depression *** Health *** Lone Rangers N = 355 (17.0%) Notes: Work Overload *** Demands: Emotional Demands *** Work Int. Home *** Self Development *** Resources: Autonomy *** Supervisor Support *** Social Support: Colleague Support *** Spousal Support *** Affective Commitment *** Affinity: Job Involvement *** Anxiety *** Wellbeing: Depression *** Health *** (A) Levels of Significance: * p <.05; ** p <.01; *** p <.001 Work Overload *** Demands: Emotional Demands *** Work Int. w/ Home *** Self Development * Resources: ~ Supervisor Support ** Social Support: Colleague Support * Spousal Support * Affective Commitment * Affinity: Job Involvement *** Anxiety *** Wellbeing: Depression *** Health *** (B) Arrows: indicates significantly greater levels than the remaining population; indicates significantly lesser levels than remaining population 24

25 References Bakker, Arnold, Schaufeli, W., Leiter, M., & Taris, T. (2008). Work engagement: An emerging concept in occupational health psychology. Work Stress, 22(3), doi: / Bakker, A.B., & Demerouti, E. (2008). Towards a model of work engagement. Career Development International, 13(3), Emerald, 60/62 Toller Lane, Bradford, West Yorkshire, BD 8 9 BY, UK,. Brief, A. P., & Weiss, H. M. (2002). Organizational behavior: affect in the workplace. Annual review of psychology, 53, doi: /annurev.psych Demerouti, E., Bakker, A., De Jonge, J., Janssen, P., & Schaufeli, W. (2001). Burnout and engagement at work as a function of demands and control. Scandinavian Journal of Work, Environment & Health, 27(4), Diez-Pinol, M., Dolan, S. L., Sierra, V., & Cannings, K. (2008). Personal and organizational determinants of well-being at work: The case of Swedish physicians. International Journal of Health Care Quality Assurance, 21(6), doi: / Latham, G. P., & Pinder, C. C. (2005). Work motivation theory and research at the dawn of the twenty-first century. Annual review of psychology, 56, doi: /annurev.psych Lee, R. T., & Ashforth, B. E. (1996). A meta-analytic examination of the correlates of the three dimensions of job burnout. Journal of Applied Psychology, 81(2), 123. Schaufeli, W.B., Salanova, M., González-Romá, V., & Bakker, A. B. (2002). The measurement of engagement and burnout: A two sample confirmatory factor analytic approach. Journal of Happiness studies, 3(1), Springer. 25

26 Schaufeli, W., Taris, T., Le Blanc, P., Peeters, M., Bakker, A., & De Jonge, J. (2001). Maakt arbeid gezond? Op zoek naar de bevlogen werknemer ( Does work make happy? In search of the engaged worker ). De Psycholoog, 36, Schaufeli, Wilmar 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(3), doi: /job.248 Schaufeli, Wilmar B., & Van Rhenen, W. (2006). Over de rol van positieve en negatieve emoties bij het welbevinden van managers: Een studie met de Jobrelated Affective Well-being Scale (JAWS) [About the role of positive and negative emotions in managers well-being: A study using the Jobrelated Affect. Gedrag & Organisatie, 19, Schaufeli, Wilmar B., Taris, T. W., & van Rhenen, W. (2008). Workaholism, Burnout, and Work Engagement: Three of a Kind or Three Different Kinds of Employee Well-being? Applied Psychology, 57(2), doi: /j x Wright, T. a. (2006). The emergence of job satisfaction in organizational behavior: A historical overview of the dawn of job attitude research. Journal of Management History, 12(3), doi: / van Saane, N., Sluiter, J. K., Verbeek, H. A. M., & Frings-Dresen, M. H. W. (2003). Reliability and validity of instruments measuring job satisfaction--a systematic review. Occupational Medicine, 53(3), doi: /occmed/kqg038 26

27 27

28 Campus Barcelona Pedralbes Av. Pedralbes, Barcelona (España) T F Campus Barcelona Sant Cugat Av. de la Torreblanca, Sant Cugat del Vallés Barcelona (España) T F Campus Madrid Mateo Inurria, Madrid (España) T F Campus Buenos Aires Av. del Libertador, Becar-San Isidro (B1643CRD) Buenos Aires (Argentina) T Global Center Munich Global Center Sao Paulo PSEUDO MAXIMUM LIKELIHOOD ESTIMATION OF STRUCTURAL CREDIT RISK MODELS WITH EXOGENOUS DEFAULT BARRIER. 3