Impact of Job-Related Well-Being on the Relationship of Self-Efficacy With Burnout

Abstract

In Brazil, professionals engaged in the child and adolescent welfare network (e.g. psychologists, social workers, teachers and social educators) are responsible for meeting the multiple demands of populations in vulnerable situations. Interventions developed by these professionals are generally aimed at preventing acts of violence, protecting victims and promoting healthy development ( Freitas & Habigzang, 2013). Despite the social relevance of these professionals, they are generally exposed to high levels of demand, conflicting demands, time pressure and poor working conditions (e.g. low levels of autonomy, poor work infrastructure). Moreover, they often lack the necessary professional qualifications to develop their work activities properly, nor do they receive the necessary support to address work demands that are emotionally exhausting ( Damasio, Habigzang, Freitas, & Koller, 2014).

Clarissa Pinto Pizarro Freitas is a Ph.D. candidate of the Graduate Program in Psychology at the Universidade Federal do Rio Grande do Sul

Claudia Sampaio Correa da Silva is a Ph.D. candidate of the Graduate Program in Psychology at the Universidade Federal do Rio Grande do Sul

Bruno Figueiredo Damasio is an Assistant Professor of the Instituto de Psicologia at the Universidade Federal do Rio de Janeiro

Silvia Helena Koller is a Full Professor of the Universidade Federal do Rio Grande do Sul

Marco Antonio Pereira Teixeira is a Full Professor of the Universidade Federal do Rio Grande do Sul

Received: Jan 20, 2015

1 Revision: Jul 9, 2015

Approved: Aug 13, 2015

Fundacao de Amparo a Pesquisa do Estado do Rio Grande do Sul / National Council for Scientific and Technological Development (FAPERGS/CNPq PRONEX - Grant # 10/0019.6)

In Brazil, professionals engaged in the child and adolescent welfare network (e.g. psychologists, social workers, teachers and social educators) are responsible for meeting the multiple demands of populations in vulnerable situations. Interventions developed by these professionals are generally aimed at preventing acts of violence, protecting victims and promoting healthy development ( Freitas & Habigzang, 2013). Despite the social relevance of these professionals, they are generally exposed to high levels of demand, conflicting demands, time pressure and poor working conditions (e.g. low levels of autonomy, poor work infrastructure). Moreover, they often lack the necessary professional qualifications to develop their work activities properly, nor do they receive the necessary support to address work demands that are emotionally exhausting ( Damasio, Habigzang, Freitas, & Koller, 2014).

According to the Job Demand and Resources Model (JD-R), occupational conditions may impact the health and well-being of workers ( Schaufeli, Dijkstra, & Vazquez, 2013). The JD-R proposes that occupational conditions can be categorized as demand and resources. The dimension demand refers to organizational aspects that require continuous efforts, such as physical, cognitive and/or emotional exertion. These efforts are associated with physiological and/or emotional costs. In turn, the resources are characterized as aspects that enable the development of the individual and the realization of his/her tasks. These resources can be personal, such as self-esteem and self-efficacy, or organizational, like social support and control over work ( Schaufeli et al., 2013).

The JD-R proposes that the demands and resources can evoke two independent processes, the health impairment process and the motivational process. The health impairment process is related to high levels of demand and low levels of resources. These working conditions may cause lower levels of well-being at work and the development of burnout. On the other hand, the motivation process is related to the availability of resources at work. This process is associated with higher levels of well-being at work and can promote engagement at work ( Schaufeli & van Rhenen, 2006; Schaufeli et al., 2013).

The poor working conditions to which most Brazilian Child and Adolescent Welfare Network professionals are usually exposed can originate the health impairment process, reducing their job satisfaction levels ( Costa & Ferreira, 2014), and may result in the development of burnout ( Carlotto, Dias, Batista, & Diehl, 2015). The emotional and cognitive distancing from work that is characteristic of burnout ( Maslach, Schaufeli, & Leiter, 2001) may lead affected professionals to be negligent regarding important work-related issues. These professionals can improperly value the difficulties experienced by the assisted populations, which might impair the quality of the service provided and reduce the possibility of effective interventions ( Damasio et al., 2014).

Burnout constitutes an occupational syndrome in which professionals have high levels of emotional exhaustion, negative attitudes towards their work activities and a tendency to negatively evaluate their ability to perform their occupational tasks. Emotional exhaustion (EE), depersonalization (Dp) and personal accomplishment (PAc) characterize the dimensions of burnout symptoms ( Maslach et al., 2001; Van Beek, Hu, Schaufeli, Taris, & Schreurs, 2012).

In the health impairment process, professionals who do not have personal and work-related resources may be more prone to experiencing lower levels of job-related affective well-being and to developing burnout. Among professionals' personal characteristics, occupational self-efficacy (OSE) has been found negatively related to burnout ( Carlotto et al., 2015; Consiglio, Borgogni, Alessandri, & Schaufeli, 2013). OSE can be defined as an individual's perception of his/her abilities to perform his/her work tasks and effectively manage conflicts at work ( Bandura, 2000; Rigotti, Schyns, & Mohr, 2008).

Professionals who have high levels of OSE generally adapt better to the work environment, are able to positively evaluate high levels of work demand ( Bandura, 2000; Consiglio et al., 2013), experience positive emotions more often and develop the motivational process ( Salanova, Llorens, & Schaufeli, 2011; Schaufeli et al., 2013). As a result, professionals with high levels of OSE may be at lower risk of experiencing the health impairment process ( Schaufeli et al., 2013) and developing burnout ( Carlotto et al., 2015; Consiglio et al., 2013).

The health impairment process also proposes that professionals who face poor work conditions and experience a lack of personal resources show lower levels of job-related affective well-being. In turn, professionals who show higher levels of self-efficacy would be more prone to developing the motivational process and experiencing higher levels of job-related affective well-being ( Schaufeli et al., 2013). In the same direction, studies have evidenced that OSE can predict the levels of positive and negative affect experienced by professionals ( Salanova et al., 2011; Williams, Wissing, Rothmann, & Temane, 2010). That is, professionals with higher levels of OSE experience higher levels of positive affect and lower levels of negative affect. This result indicates that OSE may protect people from a variety of stressors, including the inability to address work demands, in addition to promoting positive affect, such as feelings of professional accomplishment and competence ( Llorens-Gumbau & Salanova-Soria, 2014).

Besides the role of OSE in the outcomes of health impairment and motivational process, studies have shown that levels of job-related affective well-being ( Schaufeli & van Rhenen, 2006) have an important role as a protective factor against the development of burnout. Job-related affective well-being is characterized by the prevalence of job-related positive affect rather than negative affect. Positive affect (PA) is characterized by pleasant feelings, commitment to and satisfaction in developing work activities. Negative affect (NA) constitutes a state of subjective suffering and the lack of commitment to and pleasure in developing occupational activities ( Van Katwyk, Spector, Fox, & Kelloway, 2000). Higher levels of job-related affective well-being are associated with the development of positive relationships with coworkers, development of motivational process, and a lower risk of developing health impairment process and burnout ( Schaufeli & van Rhenen, 2006).

The presence of positive aspects in the work environment, such as personal resources (e.g. OSE) and PA, may promote the motivational process. In turn, the motivational process is associated with the development of gain spirals at work ( Schaufeli et al., 2013). A gain spiral at work can be defined as a set of cyclical relationships in which different positive dimensions are associated (e.g. OSE and PA), and the benefits associated with these positive dimensions are amplified over time (e.g. improved performance, increased levels of OSE, positive relationships with coworkers) ( Hobfoll & Shirom, 2000; Llorens-Gumbau & Salanova-Soria, 2014). Due to the benefits produced by the gain spirals at work, workers who have significant resources tend to create new resources as well as overcome conflicting situations where resource availability may be hindered ( Hobfoll & Shirom, 2000).

In contrast to the motivational process, the health impairment process may promote loss spirals ( Schaufeli et al., 2013). The loss spirals occur when professionals do not receive the benefits associated with their investments, e.g., an inability to properly recover from exhaustive work shifts due to job insecurity ( Hobfoll & Shirom, 2000). Because resources are connected, the loss of one resource negatively impacts the availability of others. Thus, the loss of one significant resource can lead to a set of losses, hindering the ability to maintain resources or to obtain new ones. This scenario of cyclical losses may result in illness, which is observed in the development of burnout ( Hobfoll & Shirom, 2000).

Studies evidenced the role of self-efficacy in the health impairment and motivational process ( Schaufeli et al., 2013), especially its effect over levels of job-related affective well-being ( Williams et al., 2010) and burnout ( Carlotto et al., 2015). For this reason, the current study aimed to investigate whether the relationships among levels of occupational self-efficacy and the dimensions of burnout (EE, Dp and PAc) are mediated by job-related affective well-being scores (PA and NA). The study hypotheses were that: (a) OSE predicts PA and NA; (b) PA and NA predicts EE, Dp and PAc; and (c) the relationship between PA and NA with EE, Dp and PAc will be fully mediated by OSE.

Method

Participants

A total of 584 professionals (87% women) with a mean age of 37.8 years ( SD= 11 years) participated in this study. The participants were workers in institutions that assist populations in vulnerable situations (e.g. schools, hospitals, social service centers, foster care institutions) in six regions of the Brazilian state of Rio Grande do Sul. Among the participants, 42% worked in the metropolitan region, 5% Northwest, 40% Central-West and 13% South.

The professionals had worked in their institutions for a period of time that ranged from one month to 36 years. Note that 8% of the participants was working in the institution for at most six months, 27% was working for six to 24 months, and 65% for more than 24 months. Among the participants, 15% had completed secondary school, 51% had an undergraduate degree, and 34% had a graduate degree. Men and women presented similar working time (women, 8% was working six months at most, 26% was working for six to 24 months, and 66% for more than 24 moths, men, 5% was working six months at most, 23% was working for six to 24 months, and 45% for more than 24 moths) and educational level (women, 15% had completed secondary school, 50% had an undergraduate degree, and 35% had a graduate degree: men, 23% had completed secondary school, 55% had an undergraduate degree, and 22% had a graduate degree). A total of 21% worked as primary and secondary teachers, 25% worked as guardianship counselors, 2% as nurses, 18% worked as practical nurses, 4% worked as psychologists, 7% worked as social workers, 12% educational supervisors, 9% worked in administrative functions and 2% worked as police officers.

In this study, a non-probability sampling technique was applied. Since convenience sampling was used, including all professionals who agreed to collaborate in the research was preferred, not excluding them by the work time or educational level. This procedure strengthened the ecological validity of the study, despite the fact that this enlarged the range of the work time and educational level among participants.

Instruments

Maslach Burnout Inventory - MBI(Maslach & Jackson, 1986)adapted by Lautert, (1995). This scale assesses burnout levels through the dimensions of EE, Dp and PAc. The original version of the scale exhibited satisfactory Cronbach's alpha ([alpha]) coefficients (EEE, [alpha] = .90, Dp, [alpha] = .77, and PAc, [alpha] = .74) ( Maslach & Jackson, 1986). In the Brazilian version of the MBI, the [alpha] coefficients were similarly satisfactory (EE, [alpha] = .86; Dp, [alpha] = .69; and PAc, [alpha] = .76) ( Lautert, 1995). Each item was answered according to a five-point scale, ranging from 1 ( never) to 5 ( always). The Brazilian version of MBI does not have cut-off point for interpreting the results ( Lautert, 1995). The higher the scores in the EE and Dp dimensions, and the lower the scores in the PAc dimension, the higher the burnout levels. In the current study, the following fit indices of the instrument were: Comparative Fit Index (CFI) = .93; Tucker Lewis Index (TLI) = .92; and Root Mean Square Error of Approximation (RMSEA) (90% confidence interval [CI]) = .074 [.069, .079]. The scale exhibited appropriate internal consistency in the research sample (EE, [alpha] = .87 and PAc, [alpha] = .76), except for Dp ([alpha] = .54). The number of items influences the [alpha] coefficient, as low levels of the [alpha] coefficient in Dp scale may be due to the reduced items numbers of this subscale. Notwithstanding, the confirmatory fit indices suggest the adequacy of the factor structure of the scale.

Occupational Self-Efficacy Scale - Short Form - OSS-SF( Rigotti et al., 2008) adapted by Damasio, Freitas, and Koller (2014). This scale assesses levels of OSE. The original scale has a one-factor structure consisting of six items with satisfactory Cronbach's alpha coefficients ([alpha] = .90) ( Rigotti et al., 2008). In the Brazilian validation study ( Damasio et al., 2014), a one-factor structure with adequate internal consistency ([alpha] = .78) was observed. The items are answered on a five-point scale, ranging from 1 ( completely disagree) to 5 ( completelyagree). In this sample, the Cronbach's alpha coefficients and fit indices of the instrument were satisfactory, suggesting adequacy of the scale: [alpha] = .80; CFI = .97; TLI = .94; and RMSEA (90% CI) = .061 [.036, .089].

Job-Related Affective Well-being Scale - JAWS( Van Katwyk et al., 2000) adapted by Gouveia, Fonseca, Lins, Lima, and Gouveia (2008). This scale assesses the PA and NA experienced at work. The original version of the scale uses 15 items to assess PA and 15 items to assess NA. This instrument exhibits a two-factor structure and demonstrated satisfactory internal consistency (PA, [alpha] = .94, and NA, [alpha] = .87). A short version of the JAWS ( Gouveia et al., 2008) was used that consisted of 12 items, including six items to assess PA and six items to assess NA. The short version maintained a two-factor structure of the instrument, and the internal consistency indices obtained were adequate (PA, [alpha] = .78, and NA, [alpha] = .80). The questions are answered on a five-point scale, ranging from 1 ( never) to 5 ( always). In this study, the expected two solution presented excellent internal consistencies and goodness-of-fit indices: PA, [alpha] = .86; NA, [alpha] = .82; CFI = .98; TLI = .97; RMSEA (90% CI) = .068 [.057, .078]; SRMR = .036.

Procedure

Data collection. The workers were contacted through the study "Impact Assessment of Professional Training for the Child and Adolescent Welfare Network Workers in the Brazilian State of Rio Grande do Sul." In this study, professionals in six cities of the state of Rio Grande do Sul were trained on human rights, the definition and consequences of domestic violence, interventions developed for the assistance of populations in vulnerable situations and conflict mediation. In addition to promoting professional qualifications, this study assessed the workers' characteristics that influenced their decisions to apply the contents learned during training in their professional practices. Nine training courses were developed during the research. The classes of each training course were composed of 50 to 70 professionals. Data collection occurred in the first module. The coordinator of the research and two research assistants collected the data. The questionnaires were collectively applied in auditoriums and classrooms. Participants received the instruments after signing the Informed Consent Form. The mean time for participants to answer the instruments was 30 minutes.

Data analysis. The normality of the data for each scale was initially assessed using the Kolmogorov-Smirnov test. The correlations between the dimensions of burnout (EE, Dp and PAc), PA and NA scores at work and OSE levels were investigated. Because the data were non-normally distributed, Spearman's rho correlations were performed. No multicollinearity was found. The analyses above cited were performed in the Statistical Package for the Social Sciences - Version 21.0 (SPSS 21.0).

To further analyze the correlational results, a multiple-mediation structural equation modeling was implemented to evaluate a theoretical model in which the relationships between OSE and burnout levels would be fully mediated by levels of PA and NA. The model evaluated the causally defined direct and indirect effects in the mediational model. Effects of gender, work time and educational level were controlled in the model.

The structural equation modeling was implemented in Mplus version 7.1, using the Weighted Least Squares Mean and Variance Adjusted (WLSMV) estimation method, which is useful for non-normally distributed ordinal data. To calculate the mediation effects, the bootstrapping technique (5,000 re-samplings) was used to calculate the 95% confidence interval of the mediated effect ( Muthen & Asparouhov, 2015). To assess the global model, the RMSEA, CFI and TLI fit indices were considered. According to the guidelines, RMSEA value should range between .06 and .08 (with the 90% confidence interval not exceeding .10). The CFI and TLI values should be higher than .90 (preferably higher than .95) ( Brown, 2006).

Ethical Considerations

In the first module of the training program, the participants were invited to participate in the study and received a copy of the Informed Consent Form, being informed that their participation was voluntary. Those workers who agreed to participate answered the instruments after signing the Informed Consent Form. This study received approval from the Ethics Committee of the Universidade Ritter dos Reis (UniRitter), in June 11, 2010.

Results

Table 1 shows the descriptive data and correlations between variables. EE and Dp showed positive moderate relationships with the level of NA and negative moderate associations with the level of PA and of low-to-moderate magnitude with OSE. PAc, in turn, showed a positive moderate relationship with PA and OSE, and a negative moderate association with NA. The OSE level presented positive and moderated associations with PA and negative moderated associations with levels of NA.

Table 1; Note; ** Descriptive Statistics and Spearman's Rho Correlations Between Burnout Dimensions, Occupational Self-Efficacy, Positive and Negative Affect at Work

[Table omitted; see PDF]

To further assess the results, a structural equation modeling was performed to assess whether the PA and NA could mediate the relationships between OSE and burnout levels. The fit indices in the final model were satisfactory ([Chi] ² = 1.969,80, df = 836, [Chi] ² / df = 2.35, CFI = .94, TLI = .94, RMSEA (90% CI) = .048 [.045, .051]), suggesting the model's acceptability. The mediation results showed that PA and NA mediate almost all relationships between OSE and indicators of burnout ( Table 2).

Table 2; Note Mediation Effects of Positive and Negative Affect in the Relationship Between Occupational Self-Efficacy and Burnout Indicators (N = 584)

[Table omitted; see PDF]

As shown in Table 2, both PA and NA significantly mediated the relationship between OSE and indicators of burnout. A percentage analysis of the mediated effect showed that the NA had the greatest influence on the model, especially in the relationship of OSE with EE and DP, explaining up to half of the direct relationship between OSE and these indicators (44.41% and 51.39%, respectively). For PAc, however, the levels of NA did not significantly mediate the direct relationship between OSE and PAc. PA also significantly influenced the direct relationship between OSE with EE (37.13%) and PAc (36.59%). Levels of PA did not significantly mediate the direct relationship between OSE and Dp. Furthermore, the relationship of OSE to EE and Dp was fully mediated by PA and NA. This was evidenced in the results, showing that the direct effect between these variables became non-significant after the inclusion of the mediators (OSE...EE, [beta] = .076, n.s.; OSE...Dp, [beta] = .076, n.s.) ( Figure 1). Gender, educational level and work time did not affect the relation among OSE and PA and NA at work. On the other hand, relations between EE with educational level and work time, as well as Dp with educational level were found ( p< .001). However, these relations showed low magnitude, which indicates an absence of relevant effect ( Figure 1).

Figure 1 Relationship between occupational self-efficacy, job-related affective well-being (positive and negative affect) and burnout (emotional exhaustion, depersonalization and personal accomplishment; N = 584). The figure is a representational graph. The items that compose each dimension are not shown to keep the chart simple and parsimonious. The effects between OSE and burnout dimensions are the direct effects (c') after controlling for the mediated effects.

[Figure omitted; see PDF]

Discussion

Low levels of the EE and Dp burnout dimensions, in addition to high levels of PAc, were observed in the present study. Analysis of the levels of OSE, PA and NA of the participants indicated that the participants exhibited high levels of OSE and PA and low levels of NA. The relationships between OSE and PA experienced at work evidenced that OSE and PA can be established as a resource, contributing to the engagement of professionals with their work ( Llorens-Gumbau & Salanova-Soria, 2014; Salanova et al., 2011) and to the development of the motivational process ( Schaufeli et al., 2013). Furthermore, the relationship of OSE with PA and NA indicate that OSE has an impact over professionals' evaluations of their working conditions and levels of job-related affective well-being ( Consiglio et al., 2013; Damasio et al., 2014). These results corroborate the first hypothesis of the study, that: (a) self-efficacy explains PA and NA. Thus, professionals with higher levels of OSE tend to have higher levels of job-related PA and lower levels of NA ( Salanova et al., 2011; Williams et al., 2010). These findings demonstrate the importance of these professionals working in an environment that enables the development of their potential so that greater environmental mastery can be achieved, which constitutes one of the major facets of psychological well-being ( Ryff, 2014).

The protective role of PA and the contribution of NA to the development of burnout were demonstrated in this study, showing the role of job-related affective well-being on the development of motivational process and health impairment process. The results obtained indicate that job-related affective well-being levels (both PA and PA) fully mediate the relationships between OSE levels and the dimensions of EE and Dp. The association of OSE with the dimension of PAc was partially mediated by the levels of PA. These relationships illustrate that affect plays a crucial role in the development of burnout ( Schaufeli & van Rhenen, 2006), as proposed in the second and third study hypotheses: (b) PA and NA explain EE, Dp and PAc; and (c) the relationship between PA and NA with EE, Dp and PAc is fully mediated by OSE.

The moderate magnitude of the associations between NA, EE and Dp found in this study clarify the connection between NA and the negative aspects of work such as exhaustion, emotional distance and negative attitudes toward work demands ( Schaufeli & van Rhenen, 2006). The association between PA and PAc indicates that PA is associated with positive attitudes toward work ( Salanova et al., 2011; Schaufeli & van Rhenen, 2006). This hypothesis is reinforced by the low magnitude of the relationship between NA and PAc because this dimension of burnout refers to the positive aspects of the relationship of the professional with the work ( Schaufeli, Taris, & van Rhenen, 2008). Consistent with these results, empirical evidence ( Schaufeli et al., 2008, 2013) has demonstrated that PAc exhibits a stronger relationship with engagement compared with burnout.

The relationship between OSE and PAc was partially mediated by PA, whereas NA did not contribute to explain this dimension. These relationships were different from the relationships observed between OSE and PA with EE and Dp. The differences found may be associated with the fact that the dimension of PAc develops as a consequence of the interaction of the central dimensions of burnout with the personal characteristics of the professional ( Salanova et al., 2011). In other words, it is possible that the affect experienced at work has a minimal impact on the development of low levels of PAc because such low levels result from the health impairment process, in which a chronic process of exhaustion occurs and emotional distancing between the professional and his/her work. Another aspect that justifies the contribution of OSE to the development of PAc is the similarity between these dimensions, because both refer to the evaluation of professionals regarding their abilities to manage work demands ( Maslach et al., 2001).

Evidence observed in previous studies ( Carlotto et al., 2015; Consiglio et al., 2013) suggests that OSE could be used as a predictor of the dimensions of EE, Dp and PAc. However, the present study demonstrated that the relationships between this personal characteristic (OSE) and the dimensions of burnout were either fully (EE and Dp) or partially (PAc) mediated by PA and NA. These results evidence the complexity of the motivational process and health impairment process ( Schaufeli et al., 2013). For example, professionals with lower levels of self-efficacy may consider their job demands as exhausting, as a consequence of that, they tend to show higher levels of NA. Thus, the prevalence of NA associated to low levels of OSE and poor work conditions increase the likelihood of burnout development ( Salanova et al., 2011; Schaufeli & van Rhenen, 2006). Thus, burnout can be evaluated as a work psychopathology, resulting from exposure to stressful situations and continuous experiences of NA at work ( Schaufeli & van Rhenen, 2006).

The positive correlations between positive dimensions experienced at work (e.g., PA, OSE and PAc) demonstrated the possibility for professionals to develop the motivational process, which involves the experience gain spirals at work. The gain spirals at work occur when workers exhibit the desire to maintain or increase their work resources (e.g., salary, OSE and positive relationships with the coworkers). To avoid impairment of their available resources and maintain the motivational process, professionals invest their resources at work expecting to strengthen or create new resources ( Hobfoll & Shirom, 2000; Salanova et al., 2011).

The relationship between the central dimensions of burnout and NA, in turn, indicate that negative working conditions can cause health impairment process, which include the loss spirals at work ( Salanova et al., 2011). These findings demonstrate how the lack of work resources can reduce the experience of PA at work and increase the presence of NA, thus contributing to the development of the health impairment process, as well as burnout ( Hobfoll & Shirom, 2000; Schaufeli et al., 2013).

The findings of this and other studies ( Salanova et al., 2011; Schaufeli et al., 2013) have indicated that positive aspects of work potentiate the motivational process, which involves positive outcomes and higher levels of well-being among professionals. On the other hand, negative aspects potentiate the health impairment process, which includes negative outcomes and lower levels of well-being among professionals. However, more than that, PA apparently exhibits less potential to promote the motivational process and protect against negative aspects, in the same way as NA exhibits greater potential to exacerbate the health impairment process and negative aspects compared with reducing positive aspects.

This study shows that professionals who frequently experience low levels of PA and high levels of NA are at risk of developing the health impairment process and burnout. In order to prevent professionals from developing the health impairment process, it is important to enable them to experience PA at work. Thus, it is important to develop interventions that promote the development of the motivational process, involving the experience of PA and reduce NA, as a strategy to prevent burnout because high levels of PA at work can be a protective factor against the development of burnout ( Schaufeli & van Rhenen, 2006). Furthermore, professionals who have more positive experiences at work tend to display higher professional engagement ( Schaufeli & van Rhenen, 2006) and invest their professional resources (e.g. task commitment and overtime work) in work activities. Such workers demonstrate higher investments because of the motivational process, as they expect their actions to have a positive impact on their personal development and on the development of the institution ( Damasio et al., 2014).

Despite the contributions presented herein, the current study has some limitations. The exclusive use of self-reported instruments, which were applied due to the characteristics of the evaluated constructs (burnout, OSE and PA and NA) can be considered a limitation of the data collection procedure. Furthermore, another limitation was that most of the participants were female. Nonetheless, it is important to note that the sample included professionals of different ages, education levels and work experience who had worked in the areas of health, education and social assistance. The diversity of the sample evaluated (e.g. different ages, education levels and work experience) may be seen as strengths of the study, since it increases the ecological validity of the results. Other studies that replicate the data presented here should be developed, as well as longitudinal studies to assess the existence of causal relationships between the variables.

Another limitation of this study was the low level of internal consistency of the Dp dimension of MBI Brazilian version. Nevertheless, besides the lower than expected levels of the alpha coefficient, the fit indices of the model as a whole were satisfactory, suggesting that the MBI Brazilian version is a reliable measure for burnout. Future studies should investigate the Dp factor of MBI Brazilian version, in order to make it more accurate.

The findings of this study showed that PA and NA fully mediate the relationships between the burnout dimensions of EE and Dp and OSE and partially mediate the relationship between the burnout dimension of PAc and OSE. The results suggest that more in-depth analyses of the relationships between OSE and burnout should be performed to confirm the role of PA and NA in the development of this occupational syndrome. Accordingly, further studies are needed to expand our knowledge of burnout predictors.

Footnote

Support: Fundacao de Amparo a Pesquisa do Estado do Rio Grande do Sul / National Council for Scientific and Technological Development (FAPERGS/CNPq PRONEX - Grant # 10/0019.6).

References

Bandura A. (2000). Cultivate self-efficacy for personal and organizational effectiveness. In EA Locke (Ed.), Handbook of principles of organization behavior (pp. 120-136). Oxford, UK: Blackwell.

Brown TA. (2006). Confirmatory factor analysis for applied research. New York, NY: Guilford.

Carlotto MS, Dias SRS, Batista JBV, Diehl L. (2015). O papel mediador da autoeficacia na relacao entre a sobrecarga de trabalho e as dimensoes de burnout em professores [The mediational role of self-efficacy in the relation between workload and burnout dimensions in teachers]. Psico-USF, 20(1), 13-23. doi: 10.1590/1413-82712015200102

Costa MFAA, Ferreira MC. (2014). Sources and reactions to stress in Brazilian lawyers. Paideia (Ribeirao Preto), 24(57), 49-56. doi: 10.1590/1982-43272457201407

Consiglio C, Borgogni L, Alessandri G, Schaufeli WB. (2013). Does self-efficacy matter for burnout and sickness absenteeism? The mediating role of demands and resources at the individual and team levels. Work & Stress: An International Journal of Work, Health & Organisations, 27(1), 22-42. doi: 10.1080/02678373.2013.769325

Damasio BF, Freitas CPP, Koller SH. (2014). Occupational Self-Efficacy Scale - Short Form (OSS-SF): Adaptation and evidence of construct validity of the Brazilian version. Revista Brasileira de Orientacao Profissional, 15(1), 65-74.

Damasio BF, Habigzang LF, Freitas CPP, Koller SH. (2014). Can a cognitive-behavioral group-therapy training program for the treatment of child sexual abuse reduce levels of burnout and job-strain in trainees? Initial evidence of a Brazilian model. Paideia (Ribeirao Preto), 24(58), 233-242. doi: 10.1590/1982-43272458201411

Freitas CPP, Habigzang LF. (2013). Percepcoes de psicologos sobre a capacitacao para intervencao com vitimas de violencia sexual [Perceptions of psychologists on training for intervention with victims of sexual violence]. Psicologia Clinica, 25(2), 215-230. doi: 10.1590/S0103-56652013000200013

Gouveia VV, Fonseca PN, Lins SLB, Lima AV, Gouveia RSV. (2008). Escala de bem-estar afetivo no trabalho (JAWS): Evidencias de validade fatorial e consistencia interna [Scale of Job-Related Affective Well-being Scale (JAWS): Evidence of factorial validity and internal consistency]. Psicologia: Reflexao e Critica, 21(3), 464-473. doi: 10.1590/S0102-79722008000300015

Hobfoll SE, Shirom A. (2000). Conservation of resources theory: Applications to stress and management in the workplace. In R Golembiewski (Ed.), Handbook of organizational behavior (pp. 57-80). New York, NY: Marcel Dekker.

Lautert L. (1997). O desgaste profissional: estudo empirico com enfermeiras que trabalham em hospitais [Professional exhaustion: empirical study with nurses who work in hospitals]. Revista Gaucha de Enfermagem, 18(2), 133-144.

Llorens-Gumbau S, Salanova-Soria M. (2014). Loss and gain cycles? A longitudinal study about burnout, engagement and self-efficacy. Burnout Research, 1(1), 3-11. doi: 10.1016/j.burn.2014.02.001

Maslach C, Jackson SE. (1986). Maslach Burnout Inventory (2nd ed.). Palo Alto, CA: Consulting Psychologist.

Maslach C, Schaufeli WB, Leiter MP. (2001). Job burnout. Annual Review of Psychology, 52, 397-422. doi: 10.1146/annurev.psych.52.1.397

Muthen B, Asparouhov T. (2015). Causal effects in mediation modeling: An introduction with applications to latent variables. Structural Equation Modeling: A Multidisciplinary Journal, 22(1), 12-23. doi: 10.1080/10705511.2014.935843

Rigotti T, Schyns B, Mohr G. (2008). A short version of the occupational self-efficacy scale: Structural and construct validity across five countries. Journal of Career Assessment, 16(2), 238-255. doi: 10.1177/1069072707305763

Ryff CD. (2014). Psychological well-being revisited: Advances in the science and practice of Eudaimonia. Psychotherapy and Psychosomatics, 83(1), 10-28. doi: 10.1159/000353263

Salanova M, Llorens S, Schaufeli WB. (2011). "Yes, I can, I feel good, and I just do it!" On gain cycles and spirals of efficacy beliefs, affect, and engagement. Applied Psychology, 60(2), 255-285. doi: 10.1111/j.1464-0597.2010.00435.x

Schaufeli WB, Dijkstra P, Vazquez AC. (2013). Engajamento no trabalho [Engagement at Work]. Sao Paulo, SP: Casa do Psicologo.

Schaufeli WB, Taris TW, 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), 173-203. doi: 10.1111/j.1464-0597.2007.00285.x

Schaufeli WB, van Rhenen W. (2006). Over de rol van positieve en negatieve emoties bij het welbevinden van managers: Een studie met de Job-related Affective Well-being Scale (JAWS) [On the role of positive and negative emotions in the well-being of managers: A study of the Job-related Affective Well-being Scale (JAWS)]. Gedrag en Organisatie, 19(4), 323-344.

Van Beek I, Hu Q, Schaufeli WB, Taris TW, Schreurs BHJ. (2012). For fun, love, or money: What drives workaholic, engaged, and burned-out employees at work? Applied Psychology, 61(1), 30-55. doi: 10.1111/j.1464-0597.2011.00454.x

Van Katwyk PT, Fox S, Spector PE, Kelloway EK. (2000). Using the Job-Related Affective Well-Being Scale (JAWS) to investigate affective responses to work stressors. Journal of Occupational Health Psychology, 5(2), 219-230. doi: 10.1037//1076-8998.5.2.219

Williams SA, Wissing MP, Rothmann S, Temane QM. (2010). Self-efficacy, work, and psychological outcomes in a public service context. Journal of Psychology in Africa, 20(1), 53-60.