SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. ENGAGEMENT AND EMOTIONAL EXHAUSTION: OPPOSING PSYCHOLOGICAL PHENOMENA IN THE WORK DOMAIN
  5. THE ROLE OF JOB DEMANDS AND RESOURCES
  6. DEMANDS AND RESOURCES OF THE TEACHING PROFESSION: DISTINGUISHING GENERAL AND DIFFERENTIAL ASPECTS
  7. ASSESSING THE DIFFERENTIAL DEMANDS OF THE TEACHING PROFESSION: TEACHER LEVEL AND SCHOOL LEVEL
  8. THE PRESENT INVESTIGATION
  9. METHOD
  10. RESULTS
  11. DISCUSSION
  12. REFERENCES

Focusing on the teaching profession, this study examines the association between school-specific demands and resources, on the one hand, and engagement and exhaustion, on the other. Individual-level data obtained from 1,939 secondary teachers as well as school-level data from their principals and students, based on 198 German schools, were subjected to multilevel analysis. School-level characteristics accounted for only a small amount of the variance in teachers’ emotional exhaustion. In contrast, teachers’ engagement differed considerably between schools. For the two outcome variables, engagement and exhaustion, specific patterns of predictive effects were observed at the school level: when controlling for individual teacher characteristics, the principal's support in educational matters predicted higher levels of engagement, whereas disciplinary problems in the classroom predicted higher emotional exhaustion. Although school-level data were associated with engagement and exhaustion, results suggest paying particular attention to individual differences between teachers that might predispose them to develop either more engagement or emotional exhaustion.

Axée sur la profession enseignante, cette étude examine le lien entre les exigences et les ressources propres à l’école d’une part, et l’engagement et l’épuisement d’autre part. Des données individuelles obtenues auprès de 1939 enseignants du secondaire ainsi que des données scolaires provenant des proviseurs et élèves de 198 écoles allemandes, ont été soumises à une analyse multi-niveau. Les données scolaires expliquent seulement une petite partie de la variance relative à l’épuisement émotionnel des enseignants. En revanche, l’engagement des enseignants diffère considérablement selon les écoles. Pour les variables dépendantes, engagement et épuisement, les données scolaires permettent de prédire des types spécifiques d’effets: quand on contrôle les caractéristiques individuelles des enseignants, le soutien du proviseur sur des questions éducatives implique un niveau d’engagement plus important, alors que des problèmes de discipline en classe prédisent un épuisement émotionnel plus grand. Bien que les données scolaires soient liées à l’engagement et à l’épuisement, les résultats suggèrent d’accorder une attention particulière aux différences individuelles entre les enseignants qui peuvent les prédisposer à développer soit plus d’engagement, soit plus d’épuisement émotionnel.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. ENGAGEMENT AND EMOTIONAL EXHAUSTION: OPPOSING PSYCHOLOGICAL PHENOMENA IN THE WORK DOMAIN
  5. THE ROLE OF JOB DEMANDS AND RESOURCES
  6. DEMANDS AND RESOURCES OF THE TEACHING PROFESSION: DISTINGUISHING GENERAL AND DIFFERENTIAL ASPECTS
  7. ASSESSING THE DIFFERENTIAL DEMANDS OF THE TEACHING PROFESSION: TEACHER LEVEL AND SCHOOL LEVEL
  8. THE PRESENT INVESTIGATION
  9. METHOD
  10. RESULTS
  11. DISCUSSION
  12. REFERENCES

Work-related emotions and motivations can have important implications for both individuals and organisations. Highly motivated and less stressed employees can increase the productivity of organisations significantly, whereas negative emotions and low levels of motivation are associated with impairments in individual health and with increased costs for the organisation (Brief & Weiss, 2002; Harter, Schmidt, & Hayes, 2002; Salanova, Agut, & Peiro, 2005; Schaufeli & Enzmann, 1998; Wright & Cropanzano, 1998).

The psychological functioning of teachers has recently become a focus of particular attention. Turnover and early retirement rates are high in the teaching profession, and teachers’ emotional and motivational experience may seriously impact their classroom performance (Farber, 1991; OECD, 2005; Skinner & Belmont, 1993). Against this background, research has tended to focus on negative outcomes such as stress and burnout, and on the demanding aspects of teaching. These job demands can be further differentiated into the general demands of the teaching profession (which apply to all teachers) and differential demands (which apply to specific teachers). Research on differential demands has addressed interindividual differences in teachers’ experiences of job demands and their psychological functioning. Between-school differences in demands, resources, and teachers’ psychological functioning have yet to be considered.

Yet some unique features of the teaching profession make between-school differences particularly interesting. First, organisational structures and teachers’ roles are similar across schools, meaning that the general demands of the profession are comparable. Second, within this organisational structure, aspects such as the style of leadership, the climate among colleagues, and characteristics of the student population (e.g. achievement level) can vary considerably between schools. Whereas the effects of the school context on students’ achievement, motivation, and psychosocial development have been studied in detail (e.g. Anderman, 2002; Lüdtke, Köller, Marsh, & Trautwein, 2005; Trautwein, Lüdtke, Marsh, Köller, & Baumert, 2006), empirical data on how school-level characteristics are associated with teachers’ emotional and motivational functioning are scarce (for an exception, see Friedman, 1991).

In the present investigation, we draw on data from a sample of 1,939 teachers from 198 different schools to examine two main research questions. First, are there notable between-school differences in teachers’ motivational and emotional experience, as indicated by their work engagement and emotional exhaustion? Second, to what degree can these differences be explained by characteristics of the school environment, such as work-related demands and resources? Using a multilevel approach, we seek to disentangle the effects of contextual (i.e. school-specific) and individual (i.e. teacher-specific) features.

ENGAGEMENT AND EMOTIONAL EXHAUSTION: OPPOSING PSYCHOLOGICAL PHENOMENA IN THE WORK DOMAIN

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. ENGAGEMENT AND EMOTIONAL EXHAUSTION: OPPOSING PSYCHOLOGICAL PHENOMENA IN THE WORK DOMAIN
  5. THE ROLE OF JOB DEMANDS AND RESOURCES
  6. DEMANDS AND RESOURCES OF THE TEACHING PROFESSION: DISTINGUISHING GENERAL AND DIFFERENTIAL ASPECTS
  7. ASSESSING THE DIFFERENTIAL DEMANDS OF THE TEACHING PROFESSION: TEACHER LEVEL AND SCHOOL LEVEL
  8. THE PRESENT INVESTIGATION
  9. METHOD
  10. RESULTS
  11. DISCUSSION
  12. REFERENCES

Psychological functioning in the workplace has been approached from two main perspectives. One strand of research has traditionally focused on the negative side of motivation and emotion, investigating phenomena such as stress and burnout (Kyriacou, 2001; Schaufeli & Buunk, 2003). The other has looked at more positive aspects, such as occupational and organisational commitment, engagement, involvement, and job satisfaction (Hallberg & Schaufeli, 2006; Judge, Thoresen, Bono, & Patton, 2001; Meyer, Allen, & Smith, 1993).

Burnout has been defined as a psychological syndrome characterised by three symptoms: emotional exhaustion, depersonalisation (cynicism), and reduced personal accomplishment. Its most obvious manifestation and central quality is emotional exhaustion: the feeling of being emotionally drained and depleted of emotional resources (see Lee & Ashforth, 1990). Research has shown that suffering from symptoms of burnout has considerable negative consequences for health, professional careers, and work performance (Maslach, Kackson, & Leiter, 2001; Melamed, Shirom, Toker, Berliner, & Shapira, 2006; Ostroff, 1992).

The conceptual opposite of burnout, work engagement, has recently attracted increased research attention (Hallberg & Schaufeli, 2006). Work engagement is defined as a positive, fulfilling work-related state of mind and, like burnout, is conceptualised multidimensionally, with vigor (high energy levels, willingness to invest effort in one's work), and dedication (as a sense of significance, inspiration, and pride) as its core dimensions (Gonzalez-Roma, Schaufeli, Bakker, & Lloret, 2006; Hakanen, Bakker, & Schaufeli, 2006). High work engagement has been found to promote high organisational commitment, willingness to stay in the organisation, and high performance levels (e.g. Hakanen et al., 2006; Harter et al., 2002).

Some attempts have been made to integrate these two lines of research (Bakker & Demerouti, 2007; Demerouti, Bakker, Nachreiner, & Schaufeli, 2001; Schaufeli & Bakker, 2004). Empirical studies have found burnout and engagement to show moderate negative correlations, suggesting that, for the most part, the two concepts tap different qualities of motivational and emotional experience. In the present study, we take this integrative approach and consider both sides of teachers’ psychological functioning.

THE ROLE OF JOB DEMANDS AND RESOURCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. ENGAGEMENT AND EMOTIONAL EXHAUSTION: OPPOSING PSYCHOLOGICAL PHENOMENA IN THE WORK DOMAIN
  5. THE ROLE OF JOB DEMANDS AND RESOURCES
  6. DEMANDS AND RESOURCES OF THE TEACHING PROFESSION: DISTINGUISHING GENERAL AND DIFFERENTIAL ASPECTS
  7. ASSESSING THE DIFFERENTIAL DEMANDS OF THE TEACHING PROFESSION: TEACHER LEVEL AND SCHOOL LEVEL
  8. THE PRESENT INVESTIGATION
  9. METHOD
  10. RESULTS
  11. DISCUSSION
  12. REFERENCES

A major concern of occupational research is to identify job characteristics that can influence employees’ psychological functioning. Two broad categories of job characteristics have recently been suggested: job demands and job resources (Bakker & Demerouti, 2007). Job demands are “those physical, social, or organizational aspects of the job that require sustained physical and/or psychological (cognitive or emotional) effort and are therefore associated with physiological and psychological costs” (Demerouti et al., 2001, p. 501). In situations that require high effort to sustain an expected performance level, they may become stressors and are therefore associated with high costs and negative outcomes, such as anxiety, depression, and exhaustion. In contrast, job resources are aspects of the job that may enhance motivation and performance. They function in at least one of three ways: by buffering job demands, supporting the achievement of work-related goals, or fostering learning and development (Demerouti et al., 2001).

These two types of job characteristics have recently been examined in combination with two types of psychological outcomes, namely emotional exhaustion and work engagement. The Job Demands–Resources model (JD–R; Bakker & Demerouti, 2007) hypothesises that job demands often lead to emotional exhaustion and health problems, whereas job resources facilitate high work engagement, as well as buffering the effects of work demands on emotional experience (Schaufeli & Bakker, 2004). Studies of different professions have provided empirical support for the JD–R model, identifying demands such as workload, time pressure, unfavorable physical environment, and difficult interactions with customers, as well as resources such as performance feedback, rewards, job control, and social support of colleagues and supervisors (Bakker, Demerouti, & Euwema, 2005; Demerouti et al., 2001; Llorens, Bakker, Schaufeli, & Salanova, 2006; Schaufeli & Bakker, 2004).

Although the JD–R model has proved a valuable framework in several empirical studies, at least two shortcomings of this research approach can be identified. First, although research on stress and burnout has shown that personal characteristics may affect individuals’ appraisals of and ability to cope with demanding situations, research guided by the JD–R model has not yet considered such individual employee characteristics (Semmer, 1996). Second, job characteristics have been assessed among individuals within an organisation (individual level). Although results at the individual level are known to differ from results at the work-unit level, the level of the work unit has not yet been considered (Harter et al., 2002; Ostroff, 1992).

DEMANDS AND RESOURCES OF THE TEACHING PROFESSION: DISTINGUISHING GENERAL AND DIFFERENTIAL ASPECTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. ENGAGEMENT AND EMOTIONAL EXHAUSTION: OPPOSING PSYCHOLOGICAL PHENOMENA IN THE WORK DOMAIN
  5. THE ROLE OF JOB DEMANDS AND RESOURCES
  6. DEMANDS AND RESOURCES OF THE TEACHING PROFESSION: DISTINGUISHING GENERAL AND DIFFERENTIAL ASPECTS
  7. ASSESSING THE DIFFERENTIAL DEMANDS OF THE TEACHING PROFESSION: TEACHER LEVEL AND SCHOOL LEVEL
  8. THE PRESENT INVESTIGATION
  9. METHOD
  10. RESULTS
  11. DISCUSSION
  12. REFERENCES

Emotional and motivational functioning in the teaching profession has been a focus of particular interest (OECD, 2005). When reviewing the literature on teachers’ psychological functioning and the associated work-related demands and resources, it appears to be important to distinguish the general and differential demands and resources of the profession. General demands are challenges faced by all teachers; they may explain differences in psychological functioning between teachers and members of other professions. Differential demands/resources are conditions that may vary between classes and schools; they may explain interindividual differences in teachers’ psychological functioning.

The general demands of the teaching profession relate mainly to teachers’ core tasks of teaching and educating, paperwork, lesson planning, and counseling (OECD, 2005), which tend to be similar across schools. The teacher is responsible for creating opportunities for active learning in the subject within a set curriculum and within the unique social situation of the classroom (Doyle, 1986). However, these general teaching tasks take place in differential contexts, with differential demands and resources. For example, student characteristics such as social background, achievement level, and discipline can vary considerably between schools, making the main task of teaching more or less demanding. These differential characteristics can be located at the teacher level (e.g. number of classes taught, overall workload, number of particularly difficult classes, and demands of the subject) or at the school level (e.g. support of supervisors and colleagues, general achievement level of classes, quality and quantity of interaction with students and parents, physical school environment; Vandenberghe & Huberman, 1999).

ASSESSING THE DIFFERENTIAL DEMANDS OF THE TEACHING PROFESSION: TEACHER LEVEL AND SCHOOL LEVEL

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. ENGAGEMENT AND EMOTIONAL EXHAUSTION: OPPOSING PSYCHOLOGICAL PHENOMENA IN THE WORK DOMAIN
  5. THE ROLE OF JOB DEMANDS AND RESOURCES
  6. DEMANDS AND RESOURCES OF THE TEACHING PROFESSION: DISTINGUISHING GENERAL AND DIFFERENTIAL ASPECTS
  7. ASSESSING THE DIFFERENTIAL DEMANDS OF THE TEACHING PROFESSION: TEACHER LEVEL AND SCHOOL LEVEL
  8. THE PRESENT INVESTIGATION
  9. METHOD
  10. RESULTS
  11. DISCUSSION
  12. REFERENCES

Given the comparability of organisational structures across schools, and the comparability of the general demands of the teaching profession across teachers, the teaching profession provides an ideal opportunity for investigating the effects of differential job demands at both the individual and the contextual level. Surprisingly little previous research has seized this opportunity. In typical questionnaire studies, teachers are nested within schools, resulting in two levels of analysis: the individual level and the school level. At the individual level, teacher ratings represent the individual teacher's perception of job characteristics; thus, differential aspects that vary between single teachers are assessed. At the school level, aggregated teacher ratings or third-person reports reflect job characteristics that might vary across schools.

Previous research on teachers’ psychological functioning has typically relied on teacher self-reports on job characteristics, and has thus reported associations between individual differences in job perception and individuals’ reports on emotional/motivational variables. A typical item used to assess work demands reads as follows: “How great a source of stress for you is the pupils’ lack of respect for teachers?” (from a scale by Kyriacou & Sutcliffe, 1978, often used in research on job characteristics). Clearly, any one teacher's answer describes the inner state of that teacher rather than the working conditions per se. Hence, based on such responses, it is impossible to conclude—as is often (implicitly) done in the literature—that teachers are more exhausted in schools where students show more disciplinary problems and less respect for teachers. Likewise, when job characteristics are assessed at the individual teacher level, it is important to bear in mind that teachers’ perceptions of work demands and resources may be colored by their own emotional and motivational constitution (e.g. Moyle, 1995).

Surprisingly, there has been little research interest in between-school differences in teachers’ emotional and motivational functioning. The findings of the few studies to have taken the school level as the focus of analysis point to systematic between-school differences in teachers’ emotional and motivational experience. Ostroff (1992) examined between-school differences in teachers’ attitudes and emotions in a large Canadian sample, and reported meaningful between-school variance in teachers’ job satisfaction, commitment, and perceived stress. Similarly, Caprara, Barbaranelli, Borgogni, and Steca (2003) found significant between-school variance in teachers’ self-efficacy and job satisfaction in an Italian sample. However, to our knowledge, no previous study has linked between-school differences in the core symptoms of teacher exhaustion and engagement to between-school differences in work demands and resources.

THE PRESENT INVESTIGATION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. ENGAGEMENT AND EMOTIONAL EXHAUSTION: OPPOSING PSYCHOLOGICAL PHENOMENA IN THE WORK DOMAIN
  5. THE ROLE OF JOB DEMANDS AND RESOURCES
  6. DEMANDS AND RESOURCES OF THE TEACHING PROFESSION: DISTINGUISHING GENERAL AND DIFFERENTIAL ASPECTS
  7. ASSESSING THE DIFFERENTIAL DEMANDS OF THE TEACHING PROFESSION: TEACHER LEVEL AND SCHOOL LEVEL
  8. THE PRESENT INVESTIGATION
  9. METHOD
  10. RESULTS
  11. DISCUSSION
  12. REFERENCES

The present study focuses on how differential demands and resources affect teachers’ exhaustion and engagement, on both the individual teacher and the school level. Specifically, we address the following research questions.

First, we explore whether there are meaningful differences between schools in teachers’ emotional exhaustion and engagement. There is an underlying assumption that such differences exist, particularly in burnout research, but empirical data are scarce (for exceptions, see Caprara et al., 2003; Ostroff, 1992). Two theoretical arguments support the assumption that burnout is more likely to occur in some work units than in others (Schaufeli & Buunk, 2003). First, systematic between-school differences in teacher burnout might be caused by school-specific conditions, with some work conditions being more exhausting or motivating than others. Moreover, it has been argued (Edelwich & Brodsky, 1980, p. 25) that symptoms of burnout can be contagious—like “staph infection in hospital: it gets around”. Given that negative emotions are thought to be more contagious than positive ones, the same argument might not apply to teachers’ work engagement (Bakker & Schaufeli, 2000).

Second, we use a multilevel approach to investigate whether differences in teachers’ exhaustion and engagement can be explained by school-specific demands and resources. We go beyond teachers’ self-reports to obtain data from their three major interaction partners: principals, colleagues, and students. Because individual teachers identify problematic student behavior of any kind to be the aspect causing them most stress and exhaustion (Blase, 1986; Evers, Tomic, & Brouwers, 2004; Friedman, 1995; Geving, 2007; Innes & Kitto, 1989), we investigate whether the same holds at the school level. To this end, we conceptualise features of the student population (e.g. achievement level, social background, and discipline) as school-specific demands. Similarly, we investigate whether findings at the individual level that identified the support provided by supervisors or principals to be the most helpful job resource (Browers, Evers, & Tomic, 2001; Burke, Greenglass, & Schwarzer, 1996; Greenglass, Fiksenbaum, & Burke, 1996; Halbesleben, 2006) are replicated at the school level. To disentangle teacher-specific and school-specific demands and resources, we complement school characteristics by individual teacher characteristics, such as individual workload (teaching hours and classes) and the social support of family and friends.

METHOD

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. ENGAGEMENT AND EMOTIONAL EXHAUSTION: OPPOSING PSYCHOLOGICAL PHENOMENA IN THE WORK DOMAIN
  5. THE ROLE OF JOB DEMANDS AND RESOURCES
  6. DEMANDS AND RESOURCES OF THE TEACHING PROFESSION: DISTINGUISHING GENERAL AND DIFFERENTIAL ASPECTS
  7. ASSESSING THE DIFFERENTIAL DEMANDS OF THE TEACHING PROFESSION: TEACHER LEVEL AND SCHOOL LEVEL
  8. THE PRESENT INVESTIGATION
  9. METHOD
  10. RESULTS
  11. DISCUSSION
  12. REFERENCES

Participants and Procedure

Data were provided by the COACTIV study (COACTIV: “Professional Competence of Teachers, Cognitively Activating Instruction, and the Development of Students’ Mathematical Literacy”; Kunter, Klusmann, Dubberke, Baumert, Blum, Brunner, Jordan, Krauss, Neubrand, & Tsai, in press), which was embedded in the German extension to the 2003 cycle of the OECD's Programme for International Student Assessment (PISA). The present sample consisted of teachers, students, and principals from 198 secondary schools that were representative of German secondary schools in terms of state and school track. The final teacher sample consisted of 1,939 secondary mathematics and science teachers (48.7% men; age range: 25–65 years, M= 47.4, SD= 9.4) from 198 schools (on average, 9.8 teachers per school; range: 2–12 teachers). The length of their teaching experience ranged from 1 to 44 years (M= 20.6, SD= 10.6). The 198 principals (73.6% men, age range: 35–65 years, M= 53.95, SD= 5.9) had been in this position for an average of 8.6 years (SD= 5.9) with prior teaching experience ranging from 3 to 39 years (M= 19.0, SD= 6.6). Additionally, students in two 9th grades classes per school were administered achievement tests in several domains, as well as a questionnaire assessing socio-demographic data, motivational measures, and the general class and school situation. On average, 12 students per class completed the questionnaire (range: 10–18). We used the student data from the two classes assessed as an indicator for the respective school's student population.

Measures

In line with our theoretical assumptions, the analyses included both individual- and school-level measures (see Figure 1 for an overview).

image

Figure 1. Overview of the constructs examined in the present study. T = teacher report, P = principal report, S = student report.

Download figure to PowerPoint

Outcome Measures

Engagement was assessed by four subscales from the Occupational Stress and Coping Inventory (AVEM; Schaarschmidt & Fischer, 1997). Prompted by the instruction, “We would like you to describe some of your typical behaviors, attitudes, and habits with respect to your working life”, teachers were asked to rate their agreement with four items per subscale on a 5-point response scale (1 =strongly disagree to 5 =strongly agree). Engagement was tapped by the subscales’ subjective significance of work (sample item: “Work is my main focus in life”), career ambitions (“I have high aspirations for my future career”), exertion (“I spare no effort at work”), and perfectionism (“I always want my work to be faultless”). The internal consistency was .75.

Emotional exhaustion was measured by an established German adaptation (Enzmann & Kleiber, 1989) of the Maslach Burnout Inventory (Maslach, Jackson, & Leiter, 1996). Participants were asked to rate their agreement with five statements (e.g. “I often feel exhausted at school”) on a 4-point response scale (1=strongly disagree to 4 =strongly agree). Internal consistency was good (α= .80).

Teacher-Level Predictors

Individual Workload.  Two indicators of individual workload were assessed. First, the teaching hours measure tapped the number of lesson hours per week; paperwork, lesson planning, and organisational tasks were not included in this measure. Second, the number of classes taught measure assessed the number of different classes taught per week as an indicator for general workload. Both figures tend to remain constant over the school year.

Social Support.  The social support provided by family and friends was assessed by four items (“My family gives me all the support I need”; Schaarschmidt & Fischer, 1997). The internal consistency of the scale was good (α= .85).

School-Level Predictors

We used teacher, principal, and student reports to assess features of students, colleagues, and principals as school-level predictors. Individual teacher and individual student data were aggregated to the school level by averaging the individual responses within schools (the reliability of the aggregated school-level construct is reported in the Results section). As a generic school-level variable, the principals’ reports were not aggregated. All scales are documented in the report on the German PISA 2003 assessment (Ramm, Prenzel, Baumert, Blum, Lehmann, Leutner, Neubrand, Pekrun, Rolff, Rost, & Schiefele, 2006).

Students’ Discipline.  Two measures were used as indicators of students’ discipline. First, teachers rated six items tapping student behavior (“The students behave appropriately when the teacher leaves the classroom”) on a 4-point response format (1=strongly disagree to 4 =strongly agree). Second, principals rated how often the learning process was disrupted by disciplinary problems (e.g. students disrespecting teachers) on a 4-point response format (1 =never to 4 =very often). All items were reverse-coded; thus, higher ratings indicate better discipline among students. The internal consistency was α= .84.

Students’ Basic Cognitive Abilities.  The Figure Analogies subscale from the Cognitive Ability Test 5-12+R was used to tap students’ basic cognitive abilities (Heller & Perleth, 2000). The scale, consisting of 25 figural items in multiple-choice format, is considered a parsimonious test of basic cognitive abilities. Ability scores were averaged per school. Parameters were estimated on the basis of IRT scaling, resulting in weighted-likelihood estimates (WLE; Warm, 1989).

Students’ Social Background.  Parents’ socioeconomic status was assessed using the International Socio-Economic Index (ISEI) developed by Ganzeboom, de Graaf, Treiman, and de Leeuw (1992). We used the highest ISEI score in the family (range: 10–90 points) in the analyses.

Teachers’ Cooperation and Morale.  Two indicators were used to tap cooperation and school climate. First, the frequency of cooperation was assessed by an 11-item scale tapping three aspects: exchange on instructional content and methods (“How often do you talk to your colleagues about instructional methods?”), exchange on class tests and examinations (“How often do you talk to your colleagues about the content of class tests and examinations?”), and exchange on problems with students (“How often do you talk to your colleagues about problems with your class?”). Responses were given on a 4-point scale (1 =never to 4 =very often). Second, principals reported on teachers’ morale. The teachers’ school-related attitudes and behavior were assessed by 4 items (“Morale and work-related attitudes are high in our school”) with a 4-point response format (1 =strongly disagree to 4 =strongly agree). The internal consistency was satisfactory (α= .75).

Principals’ Support.  Teachers rated the perceived pedagogical support provided by their principal on an 11-item scale with the prompt “Do you see your principal as an advisor on educational matters?” and a 4-point response format (1 =strongly disagree to 4 =strongly agree). The scale tapped the principal's perceived availability and competence with respect to pedagogical matters (“The principal gives me targeted advice when educational difficulties arise”).

Statistical Analysis

Statistical Models.  Hierarchical linear modeling was employed to disentangle within- and between-school variance, and to predict emotional exhaustion and engagement by reference to individual and school-level predictors (Raudenbush & Bryk, 2002). All models reported are random-intercept models, which means that the intercept was freely estimated, thus indicating between-school differences in teachers’ engagement and exhaustion. Teacher variables—i.e. age, gender, number of teaching hours, number of classes, and social support of family and friends—were specified at the first level. Simultaneously, we examined whether the school-level variables were related to teachers’ engagement and exhaustion (for a more detailed description of the random-intercept model, see, e.g. Hox, 2002).

The results of hierarchical linear models can be interpreted in a similar way to those of ordinary regression analyses. The advantages of hierarchical linear modeling (HLM) are that it produces correct estimates of standard errors of beta coefficients and that it can provide information on the distribution of variance between the different levels of analysis. Because HLM does not provide standardised regression coefficients, we standardised (M= 0, SD= 1) all continuous predictor variables at the grand mean of the sample in order to enhance the interpretability of the regression coefficients. Dichotomous variables were retained in their original metric. Analyses were performed with the computer program HLM 6 (Raudenbush, Bryk, Cheong, & Congdon, 2004).

Effect Sizes.  Effect sizes are increasingly being used to describe the substantive meaning or real-world importance of an empirical finding beyond its statistical significance. We use three indicators of effect sizes to describe the results of our multilevel models. First, in analogy to the explained variance measure in ordinary linear regression models, we report the proportion of variance explained by the predictor variables at each level. Second, we report easily interpretable regression coefficients for level-1 variables. The coefficients can be interpreted in almost the same way as standardised regression coefficients, because we standardised all continuous predictor and outcome variables before entering them in our multilevel model. Third, we apply a formula described by Tymms (2004) to calculate an effect size for the school-level predictor variables. Effect sizes overcome the interpretational ambiguities associated with regression coefficients at higher levels in multilevel analyses. Tymms’ (2004) formula for continuous level-2 predictors in multilevel models aims to provide an effect size comparable with Cohens's d (Cohen, 1988). It is calculated using the following formula:

  • Δ= 2 ×B×SDpredictor/σe

where B is the unstandardised regression coefficient in the multilevel model, SDpredictor is the standard deviation of the predictor variable at the class level, and σe is the residual standard deviation at the student level. In the present study, we interpret an effect to be “meaningful” if the effect size of a school-level predictor is Δ= .20 or larger.

Missing Data.  The percentage of missing data ranged from 0 per cent to 12 per cent, with an average of 2.9 per cent. There is increasing agreement in the methodological literature that multiple imputation of missing data is superior to pairwise and listwise deletion methods. We thus used the NORM software (version 2.03; Schafer & Graham, 2002) to perform multiple imputation. All data available on the variables considered in this study were used to estimate the missing values. We generated five data sets in which missing values were replaced by estimated values. HLM 6.0 was then used to simultaneously analyse all five imputed data sets.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. ENGAGEMENT AND EMOTIONAL EXHAUSTION: OPPOSING PSYCHOLOGICAL PHENOMENA IN THE WORK DOMAIN
  5. THE ROLE OF JOB DEMANDS AND RESOURCES
  6. DEMANDS AND RESOURCES OF THE TEACHING PROFESSION: DISTINGUISHING GENERAL AND DIFFERENTIAL ASPECTS
  7. ASSESSING THE DIFFERENTIAL DEMANDS OF THE TEACHING PROFESSION: TEACHER LEVEL AND SCHOOL LEVEL
  8. THE PRESENT INVESTIGATION
  9. METHOD
  10. RESULTS
  11. DISCUSSION
  12. REFERENCES

To What Extent do Demands and Resources Differ between Schools?

Before addressing our main research questions, we investigated whether there were indeed meaningful between-school differences in the school-level characteristics assessed (see Tables 1 and 2 for descriptive statistics for the teacher- and school-level predictors). Only in this case it is reasonable to analyse whether teachers’ engagement and emotional exhaustion differ as a function of students’ individual characteristics.

Table 1.  Means, Standard Deviations, and Intercorrelations among Teacher-Level Predictor Variables
 MSD(1)(2)(3)(4)(5)
  1. Note: Correlations > .1 are statistically significant at p < .01; N ranges from 1,734 to 1,900.

(1) Gender (0 = female, 1 = male)0.490.501.00    
(2) Age47.449.390.201.00   
(3) Teaching hours23.874.540.160.021.00  
(4) Number of classes taught5.942.640.06−0.120.291.00 
(5) Social support4.040.770.04−0.04−0.010.021.00
Table 2.  Means, Standard Deviations, and Intercorrelations among School-Level Variables
 M(SD)(1)(2)(3)(4)(5)(6)(7)
  1. Note: SES = socioeconomic status; T = teacher reports, P = principal reports, S = student reports; correlations > .14 are statistically significant at p < .05; N ranges from 187 to 198.

Principal
 (1) Support (T)2.580.401.00      
Teachers
 (2) Morale (P)3.270.41.161.00     
 (3) Cooperation (T)2.740.25.17−.021.00    
Students
 (4) Discipline (T)2.710.33.04.13−.191.00   
 (5) Discipline (P)2.890.56.02.25−.21.521.00  
 (6) SES (S)49.828.20−.21.15−.43.52.381.00 
 (7) Basic cognitive abilities (S)−.190.93−.14.06−.42.61.47.831.00

For the non-aggregated principal data, the standard deviation is a good indicator of the amount of variability between schools. For both principal reports (teacher morale and student discipline), the standard deviation was adequate, at > 0.40 for the 4-point Likert-type scales.

For the other level-2 predictor variables, we calculated the intraclass correlation coefficients ICC1 and ICC2; these coefficients are routinely used to assess the reliability of aggregated level-2 variables (see Lüdtke, Trautwein, Kunter, & Baumert, 2006). The intraclass correlation (ICC1) indicates the amount of variance in a variable that is located between schools relative to the overall variance. Results showed that the between-school variance of the school-level predictors ranged from .23 per cent (teachers’ frequency of cooperation) to .52 per cent (students’ basic cognitive abilities), thus indicating considerable between-school differences in work-related demands and resources. Based on the ICC1 and the average number of units (teachers, students) at the aggregate level, it is possible to calculate the ICC2, which is an indicator of the variable's reliability at the aggregate level (for details, see, e.g. Lüdtke et al., 2006). ICC2 values above .70 indicate that the aggregated ratings provide a reliable assessment of the school-level construct. In the present sample, we found ICC2 values ranging between .67 and .80; only the reliability of teachers’ reports on the frequency of cooperation with colleagues was slightly below .70.

In addition to the reliability of the aggregated data, we examined agreement among teachers on school characteristics (see Lüdtke et al., 2006). To this end, we calculated the average deviation index (ADM), which indicates the average deviation of teachers’ ratings from the school mean. The (ADM) can be interpreted in terms of the original rating metric; for instance, an (ADM) of 1 would mean that, on average, the teachers in a school deviate from the school mean by one unit on the rating scale. It has been suggested that the deviations be seen in relation to the number of response categories. As a rule of thumb, given a 4-point response format, the threshold is .67, with lower values indicating sufficient agreement among the teachers within a school. In the present data, we found ADM values between .47 and .60, indicating very satisfactory agreement among teachers on their work environment.

In sum, our preliminary analyses confirmed that there was considerable between-school variation in the school-level characteristics assessed, and that these characteristics were assessed reliably, with a high level of agreement among the teachers within a school.

To What Extent do Teachers’ Engagement and Exhaustion Differ between Schools?

Concerning the major outcome variables, teachers’ ratings of their work engagement were near the midpoint of the scale (range: 1–5; M= 2.94, SD= .59). In contrast, their emotional exhaustion ratings were below the midpoint of the scale (range: 1–4; M= 2.11, SD= .64). The correlation between teachers’ engagement and teachers’ emotional exhaustion was not statistically significant (r= .01, ns).

Our first main research question concerned between-school differences in teachers’ engagement and emotional exhaustion. To what degree do teachers’ engagement and exhaustion differ between schools? To answer this question, we specified separate empty models (also known as null models or unconditional models) in HLM for engagement and exhaustion. The two models indicate whether the means of teachers’ engagement and exhaustion differed meaningfully across the level-2 unit, in our case between schools. The analysis revealed a significant amount of between-school variance for teachers’ engagement, with 7 per cent of the total variance being located between schools (see Table 3; Model 1). In contrast, only 1 per cent of the variance in teachers’ emotional exhaustion was located between schools (see Table 4; Model 1).

Table 3.  Teachers’ Engagement: Results from Multilevel Modeling
 Engagement
Model 1Model 2Model 3
 BSE(B)BSE(B)
  • Note: Gender: 0 = female, 1 = male; SES = socioeconomic status; T = teacher reports, P = principal reports, S = student reports; Model 1 = intercept-only model; Model 2 = bivariate associations between predictor and engagement; Model 3 = full model.

  • **

    p < .01;

  • *

    p < .05.

School level
 Principal
  Support (T) .10**.03.11**.03
 Teaching colleagues
  Morale (P) −.01.03−.02.03
  Cooperation (T) −.02.03−.01.03
 Students
  Discipline (T) .03.03.02.04
  Discipline (P) .01.03−.04.04
  SES (S) .05.03.05.05
  Basic cognitive abilities (S) .05.03.04.05
Teacher level
 Gender −.16**.05−.15*.05
 Age −.10**.02−.08*.02
 Teaching hours .01.03.04.05
 Number of classes taught .03.03.04.03
 Social support  .03.03−.03.02
Residual variance
 School level.07  .04 
 Teacher level.93  .92 
Explained variance
 School level   .43 
 Teacher level   .01 
Table 4.  Teachers’ Exhaustion: Results from Multilevel Modeling
 Exhaustion
Model 1Model 2Model 3
 BSE(B)BSE(B)
  • Note: Gender: 0 = female, 1 = male; SES = socioeconomic status; T = teacher reports, P = principal reports, S = student reports; Model 1 = intercept-only model; Model 2 = bivariate associations between predictor and engagement; Model 3 = full model.

  • **

    p < .01;

  • *

    p < .05.

School level
 Principal
  Support (T) −.03.02−.04.02
 Teaching colleagues
  Morale (P) −.01.02.01.02
  Cooperation (T) .06*.02.03.02
 Students
  Discipline (T) −.11**.02−.11**.03
  Discipline (P) −.07**.02−.02.02
  SES (S) −.08**.02−.01.04
  Basic cognitive abilities (S) −.08**.02−.01.04
Teacher level
 Gender .03.05−.02.05
 Age .06**.02.07**.02
 Teaching hours .10**.02.02.03
 Number of classes taught .16**.03.19**.03
 Social support  −.18**.02−.18**.03
Residual variance
 School level.01  .00 
 Teacher level.99  .92 
Explained variance
 School level   .75 
 Teacher level   .07 

Predicting Teachers’ Engagement and Exhaustion: The Role of Individual and School Characteristics

Our second research question addressed whether and how school-related demands and resources predict teachers’ engagement and exhaustion, over and above individual teacher characteristics. Beginning with teachers’ engagement as the dependent variable, we first entered just one predictor variable per model to determine the single contribution of each predictor separately. As documented in Table 3 (Model 2), the principal's pedagogical support was the only school-level variable associated with teachers’ engagement; the regression coefficient (B= .10) indicated that higher perceived support in pedagogical matters was associated with higher teacher engagement. Of the teacher-level variables, gender and age were associated with engagement. Male teachers tended to be less engaged than female teachers (B=−.16), and older age was associated with less engagement (B =−.10).

In the next step, we entered all predictors simultaneously. This model helps to understand the specific contributions of variables: the regression coefficients of each predictor can be interpreted as unique predictive effects when all other variables in the model are held constant. As reported in Table 3 (Model 3), the results remained fairly stable relative to the bivariate associations. At the school level, the principal's pedagogical support (B= .11) predicted teachers’ engagement. In other words, when all other school and individual variables were controlled, an increase of one standard deviation in the aggregated perception of the principal's support was associated with an increase of .11 in the school mean of teachers’ engagement. Using the formula described above (Tymms, 2004), we calculated the effect size of this coefficient to be Δ= .24, which we interpret to indicate a small but meaningful effect.

Teachers’ engagement was not statistically significantly predicted by teachers’ morale, the frequency of interaction with colleagues, or student characteristics such as social background, basic cognitive abilities, or discipline. At the teacher level, gender (B=−.15) and age (B=−.08) were statistically significantly associated with teachers’ engagement, but the social support of family and friends and the two indicators of teachers’ workload were not. The predictor variables explained 1 per cent of the between-teacher variance and 43 per cent of the between-school variance in teachers’ engagement.

A parallel series of models was run for teachers’ emotional exhaustion. Again, we started by estimating the multilevel random-intercept model separately for each predictor. As reported in Table 4 (Model 2), the frequency of interaction among colleagues was positively related to teachers’ exhaustion; the regression coefficient was small but statistically significant (B= .06). Moreover, all predictors relating to the school-specific student population had statistically significant regression coefficients. Higher teacher (B=−.11) and principal ratings (B=−.07) of student discipline, higher student socioeconomic background (B=−.08), and higher basic cognitive abilities (B=−.08) were associated with lower levels of emotional exhaustion. On the teacher level, there were statistically significant coefficents for age (B =−.06), number of teaching hours (B = .10), number of classes taught (B = .16), and social support of family and friends (B =−.18).

Next, we estimated the full model including all predictors simultaneously. As documented in Table 4 (Model 3), the students’ general discipline as rated by their teachers was statistically significantly negatively associated with teacher exhaustion (B=−.11). More disciplined student behavior predicted less emotional exhaustion among teachers, when all other variables at the school and the individual level were controlled. Again, using the formula suggested by Tymms (2004), we found an effect size of Δ=−.24, which we consider to be meaningful.

None of the regression weights of the other student characteristics remained statistically significant. At the individual level, teachers’ age positively predicted exhaustion (B= .07). In terms of workload, when controlling for the other predictors, a higher number of classes taught was still associated with a higher level of exhaustion (B= .19), but the number of teaching hours was not. Social support of family and friends was negatively associated with exhaustion (B =−.18). The full model explained 7 per cent of the between-teacher variance and 75 per cent of the between-school variance in teachers’ emotional exhaustion.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. ENGAGEMENT AND EMOTIONAL EXHAUSTION: OPPOSING PSYCHOLOGICAL PHENOMENA IN THE WORK DOMAIN
  5. THE ROLE OF JOB DEMANDS AND RESOURCES
  6. DEMANDS AND RESOURCES OF THE TEACHING PROFESSION: DISTINGUISHING GENERAL AND DIFFERENTIAL ASPECTS
  7. ASSESSING THE DIFFERENTIAL DEMANDS OF THE TEACHING PROFESSION: TEACHER LEVEL AND SCHOOL LEVEL
  8. THE PRESENT INVESTIGATION
  9. METHOD
  10. RESULTS
  11. DISCUSSION
  12. REFERENCES

This study was guided by two main research questions. First, are there between-school differences in teachers’ engagement in their profession and in teachers’ emotional exhaustion? Second, to what extent can teachers’ engagement and exhaustion be explained by characteristics of the school environment and by characteristics of the individual teacher?

Between-School Variability in Teachers’ Engagement and Exhaustion

Our findings showed relatively little between-school variance in teachers’ engagement and even less between-school variance in teachers’ emotional exhaustion. Most of the variability in teachers’ emotional and motivational experience can thus be ascribed to individual rather than school factors. In view of the fact that shared working environments are assumed to have similar effects on employees’ psychological functioning, and given Maslach's conclusion (Maslach et al., 2001) that environmental factors are more relevant than personal factors for work-related behavior, these findings might seem rather surprising. Indeed, two previous empirical studies addressing between-school differences in teachers’ job satisfaction, commitment, and self-efficacy found considerably higher between-school variability (Caprara et al., 2003; Ostroff, 1992).

How, then, can the present findings be explained? First, our study differs from previous research with respect to (a) the teacher outcome characteristics and (b) the teacher populations under investigation. Whereas the two previous studies examined the more cognitive components of teacher attitudes, we focus on the more emotional aspects of teachers’ engagement and emotional exhaustion. Moreover, the school systems of the different teacher samples studied (Canadian and Italian vs. German) are barely comparable. Schools in Germany are likely to be more similar than Canadian or Italian schools in terms of conditions such as teacher pay, tenure, and organisational structures (OECD, 2005). For instance, teachers in Germany are not selected to schools depending on their qualifications (e.g. university grades), but are assigned to schools by the public school administration, which is centralised and fairly independent of local schools. Consequently, selection effects causing teachers with particular characteristics to cluster in the same school due to self-selection (e.g. schools may be more or less attractive for beginning teachers) or school selection are less likely. Hence, it is possible that the higher between-school differences in teacher attitudes found in previous studies, which have been interpreted as “school effects”, might in fact be caused at least partly by “selection effects”.

Second, the small amount of between-school variance may indicate that the general demands of the teaching profession are more important for teachers’ psychological functioning than are school-specific demands and resources. In other words, the low school-specific variability might be attributable to the general demands of the teaching profession, which apply to all teachers. The considerable amount of variance in both engagement and emotional exhaustion that we found within schools contradicts this explanation, however.

Third, in line with the conceptualisation of burnout as a long-term consequence of work-related stress (Maslach et al., 2001), one might argue that particular contextual features come to influence individuals’ emotions and motivations over time, and that between-school differences might be more likely among teachers who have taught in the same school environment for longer. However, this hypothesis is difficult to examine in our study due to reduction of the sample size. Preliminary analyses with the present teacher sample did not confirm the hypothesis. The amount of between-school variance in engagement and exhaustion was not higher in older teachers who had been employed at the same school for longer than in the overall teacher sample.

Fourth, from the perspective of transactional stress theory (Lazarus, 2001), the results can be interpreted as supporting the subjectivity of emotional and motivational behaviors in the working context. Although teachers within schools showed high levels of agreement in their ratings of principal's support, cooperation among colleagues, and student discipline, they differed greatly in their descriptions of exhaustion and engagement. These findings support the hypothesis that personal factors (personality dispositions that influence situational appraisals and coping strategies) moderate the association between environmental conditions and emotional and motivational reactions (Lazarus & Folkman, 1984). The results thus highlight how important it is for models like the JD–R model to take personal prerequisites into consideration.

Contextual and Individual Features Associated with Teachers’ Engagement and Exhaustion

We investigated the associations of school and individual teacher characteristics and teachers’ engagement and exhaustion. We found the teacher-level characteristics of age and gender to predict teachers’ engagement, but these variables explained only 1 per cent of the between-teacher variance. On the school level, the principal's support in pedagogical matters predicted average teacher engagement. In other words, when individual teacher factors were controlled, schools with a more supportive principal had more engaged teachers. This finding emphasises the principal's responsibility and potential influence, and is in line with previous research, which found between-teacher differences in perceptions of the principal's support to be associated with individual engagement (Bakker & Demerouti, 2007).

In contrast, none of the student characteristics were associated with teachers’ engagement. Teachers’ behavior appeared to be independent of student discipline, background, or ability level. These results are in line with the predictions of the job demands–resources model, which states that resources are more strongly related to employee engagement, whereas demands are more strongly related to employee exhaustion.

The findings for teachers’ emotional exhaustion were also in line with the predictions of the job demands–resources model. On the individual level, teachers’ age showed a similar level of association as for engagement, with older teachers tending to report less work engagement. The number of classes taught was associated with higher levels of emotional exhaustion, and higher social support from family and friends was associated with lower exhaustion. Whereas the effect of social support is well established (e.g. Greenglass et al., 1996; Halbesleben, 2006), the finding that the number of different classes taught is more relevant to exhaustion than the number of teaching hours merits discussion. As a first tentative interpretation, we suggest that being involved in different social contexts may require more emotional adaptiveness, which may in turn lead to the depletion of emotional resources. This finding further demonstrates that a more differentiated assessment of the global construct of workload may provide insights into its particularly stressful aspects (see also Blase, 1986, on the distinction between quantitative and qualitative workload).

On the school level, between-school variability in teachers’ exhaustion was predicted only by the teachers’ reports on student discipline. When individual teacher characteristics were controlled, the lower the teachers’ discipline ratings, the higher their average exhaustion. None of the other student characteristics remained statistically significant in the full model, on account of the substantial intercorrelations of students’ ability level, social background, and discipline. In other words, when students’ ability level and social background were controlled, student discipline was the critical aspect for teachers’ exhaustion. Although previous studies have demonstrated the crucial role of students’ discipline in the development of teacher stress and burnout (Geving, 2007), they have focused exclusively on the individual level. We went further by examining between-school differences in students’ behavior. The sizes of our effects were somewhat smaller than those found on the individual level, perhaps because aggregated teacher data provide more objective descriptions of contextual features. Shared perceptions of contextual features may be less colored by idiosyncratic emotional and motivational experience within this context than individual ratings, leading to weaker associations between contextual characteristics and the outcome measures of exhaustion and engagement.

Limitations and Future Research

Some limitations of the present study and avenues for further research should be mentioned. First, the present study relied on cross-sectional data, meaning that the associations reported between individual and school characteristics and teachers’ engagement and exhaustion are correlational. Longitudinal studies are needed to cast light on the causal pathways by which teachers’ emotion and motivation are influenced. Second, the number of teachers and classes per school should be increased in further studies. A design including teachers of different subjects would provide a more detailed picture. Similarly, future studies would benefit from obtaining student ratings from more than two classes per school and in different age groups. Third, as in previous research, our two outcome measures were assessed by teachers’ self-report data. Future research might benefit from including students’ and principals’ ratings of teacher behavior as well as objective indicators of stress and burnout (e.g. days of absence, turnover intentions, and psychosomatic and physical symptoms) to avoid confounding between teacher reports on (a) their emotions and motivations and (b) the characteristics of the school. Finally, on the school level, the present study focused on teachers’ interactions with others in the school. To provide a more comprehensive description of the school environment, future research should also include physical characteristics of the school (e.g. facilities, location, size) and organisational variables (e.g. professional development, school development programs).

Conclusions and Implications

First, in line with previous research, our results indicate that research on occupational well-being should consider positive (engagement) and negative (exhaustion) aspects of psychological functioning separately. From a more practical perspective, we can conclude that an absence of burnout symptoms does not necessarily mean that teachers have high work engagement, and that teachers showing highly engaged teaching behavior might at the same time be experiencing emotional exhaustion. These findings may be important for both prevention and intervention programs (Schaufeli & Bakker, 2004).

Second, research on teachers’ job characteristics should always distinguish between the individual and the school level. Researchers interested in contextual effects should be careful to disentangle objective contextual features from individual interpretations of these features. Hierarchical linear models combining the individual and contextual levels seem a promising and methodologically sound way to enhance the scientific knowledge of contextual effects in the work domain. Only with a deeper understanding of contextual effects will it be possible to buffer employees against context-specific demands and to augment the context with resources that can help them to master the general demands of their workplace.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. ENGAGEMENT AND EMOTIONAL EXHAUSTION: OPPOSING PSYCHOLOGICAL PHENOMENA IN THE WORK DOMAIN
  5. THE ROLE OF JOB DEMANDS AND RESOURCES
  6. DEMANDS AND RESOURCES OF THE TEACHING PROFESSION: DISTINGUISHING GENERAL AND DIFFERENTIAL ASPECTS
  7. ASSESSING THE DIFFERENTIAL DEMANDS OF THE TEACHING PROFESSION: TEACHER LEVEL AND SCHOOL LEVEL
  8. THE PRESENT INVESTIGATION
  9. METHOD
  10. RESULTS
  11. DISCUSSION
  12. REFERENCES
  • Anderman, E.M. (2002). School effects on psychological outcomes during adolescence. Journal of Educational Psychology, 84, 261271.
  • Bakker, A.B., & Demerouti, E. (2007). The job demands–resources model: State of the art. Journal of Managerial Psychology, 22, 309328.
  • Bakker, A.B., Demerouti, E., & Euwema, M.C. (2005). Job resources buffer the impact of job demands on burnout. Journal of Occupational Health Psychology, 10, 170180.
  • Bakker, A.B., & Schaufeli, W.B. (2000). Burnout contagion process among teachers. Journal of Applied Social Psychology, 30, 22892308.
  • Blase, J.J. (1986). A qualitative analysis of sources of teacher stress: Consequences for performance. American Educational Research Journal, 23, 1340.
  • Brief, A.P., & Weiss, H.M. (2002). Organizational behavior: Affect in the workplace. Annual Review of Psychology, 53, 279307.
  • Brouwers, A., Evers, W.J.G., & Tomic, W. (2001). Self-efficacy in eliciting social support and burnout among secondary-school teachers. Journal of Applied Social Psychology, 31, 14741491.
  • Burke, R.J., Greenglass, E.R., & Schwarzer, R. (1996). Predicting teacher burnout over time: Effects of work stress, social support, and self-doubts on burnout and its consequences. Anxiety, Stress and Coping, 9, 261275.
  • Caprara, G.V., Barbaranelli, C., Borgogni, L., & Steca, P. (2003). Efficacy beliefs as determinants of teachers’ job satisfaction. Journal of Educational Psychology, 95, 821832.
  • Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd edn.). Hillsdale, NJ: Erlbaum.
  • Demerouti, E., Bakker, A.B., Nachreiner, F., & Schaufeli, W.B. (2001). The job demands–resources model of burnout. Journal of Applied Psychology, 86, 499512.
  • Doyle, W. (1986). Classroom organization and management. In M.Wittrock (Ed.), Handbook of research on teaching (pp. 392431). New York: Macmillan.
  • Edelwich, J., & Brodsky, A. (1980). Burnout: Stages of disillusionment in the helping professions. New York: Human Sciences Press.
  • Enzmann, D., & Kleiber, D. (1989). Helfer-Leiden: Stress und Burnout in psychosozialen Berufen [Stress and burnout in human service professions]. Heidelberg, Germany: Roland Asanger.
  • Evers, W.J.G., Tomic, W., & Brouwers, A. (2004). Burnout among teachers: Students’ and teachers’ perceptions compared. School Psychology International, 25, 131148.
  • Farber, B.A. (1991). Crisis in education: Stress and burnout in the American teacher. San Francisco, CA: Jossey-Bass.
  • Friedman, I.A. (1991). High- and low-burnout schools: School culture aspects of teacher burnout. Journal of Educational Research, 84, 325333.
  • Friedman, I.A. (1995). Student behavior patterns contributing to teacher burnout. Journal of Educational Research, 88, 281289.
  • Ganzeboom, H.B.G., De Graaf, P.M., Treiman, D.J., & De Leeuw, J. (1992). A standard international socio-economic index of occupational status. Social Science Research, 21, 156.
  • Geving, A.M. (2007). Identifying the types of student and teacher behaviours associated with teacher stress. Teaching and Teacher Education, 23, 624640.
  • Gonzalez-Roma, V., Schaufeli, W.B., Bakker, A.B., & Lloret, S. (2006). Burnout and work engagement: Independent factors or opposite poles? Journal of Vocational Behavior, 68, 165174.
  • Greenglass, E., Fiksenbaum, L., & Burke, R.J. (1996). Components of social support, buffering effects and burnout: Implications for psychological functioning. Anxiety, Stress and Coping, 9, 185197.
  • Hakanen, J.J., Bakker, A.B., & Schaufeli, W.B. (2006). Burnout and work engagement among teachers. Journal of School Psychology, 42, 495513.
  • Halbesleben, J.R. (2006). Sources of social support and burnout: A meta-analytic test of the conservation of resources model. Journal of Applied Psychology, 91, 11341145.
  • Hallberg, U.E., & Schaufeli, W.B. (2006). “Same same” but different? Can work engagement be discriminated from job involvement and organizational commitment? European Psychologist, 11, 119127.
  • Harter, J.K., Schmidt, F.L., & Hayes, T.L. (2002). Business-unit-level relationship between employee satisfaction, employee engagement, and business outcome: A meta-analysis. Journal of Applied Psychology, 87, 268279.
  • Heller, K.A., & Perleth, C. (2000). Kognitiver Fähigkeitstest für 4.–12. Klasse–Revision (KFT 4-12+R) [Basic cognitive abilities test for grades 4 to 12–Revision]. Göttingen, Germany: Hogrefe.
  • Hox, J.J. (2002). Multilevel analysis: Techniques and applications. Mahwah, NJ: Erlbaum.
  • Innes, J.M., & Kitto, S. (1989). Neuroticism, self-consciousness and coping strategies, and occupational stress in high school teachers. Personality and Individual Differences, 10, 303312.
  • Judge, T.A., Thoresen, C.J., Bono, J.E., & Patton, G.K. (2001). The job satisfaction–job performance relationship: A qualitative and quantitative review. Psychological Bulletin, 127, 376407.
  • Kunter, M., & Baumert, J. (2006). Who is the expert? Construct and criteria validity of student and teacher ratings of instruction. Learning Environments Research, 9, 231251.
  • Kunter, M., Klusmann, U., Dubberke, T., Baumert, J., Blum, W., Brunner, M., Jordan, A., Krauss, S., Neubrand, M., & Tsai, Y.-M. (in press). Linking aspects of teacher competence to their instruction: Results from the COACTIV project. In M.Prenzel (Ed.), Studies on the educational quality of schools. The final report on the DFG Priority Programme. Münster, Germany: Waxmann.
  • Kyriacou, C. (2001). Teacher stress: Directions for future research. Educational Review, 53, 2735.
  • Kyriacou, C., & Sutcliffe, J. (1978). Teacher stress: Prevalence, sources, and symptoms. British Journal of Educational Psychology, 48, 159167.
  • Lazarus, R.S. (2001). Conservation of resources theory (COR): Little more than words masquerading as a new theory. Applied Psychology: An International Review, 50, 381391.
  • Lazarus, R.S., & Folkman, S. (1984). Stress, appraisal, and coping. New York: Springer.
  • Lee, R.T., & Ashforth, B.E. (1990). On the meaning of Maslach's dimensions of burnout. Journal of Applied Psychology, 75, 743747.
  • Llorens, S., Bakker, A.B., Schaufeli, W., & Salanova, M. (2006). Testing the robustness of the job demands–resources model. International Journal of Stress Management, 13, 378391.
  • Lüdtke, O., Köller, O., Marsh, H.W., & Trautwein, U. (2005). Teacher frame of reference and the big-fish–little-pond effect. Contemporary Educational Psychology, 30, 263285.
  • Lüdtke, O., Trautwein, U., Kunter, M., & Baumert, J. (2006). Reliability and agreement of student ratings of the classroom environment: A reanalysis of TIMSS data. Learning Environments Research, 9, 215230.
  • Maslach, C., Jackson, S.E., & Leiter, M.P. (1996). Maslach burnout inventory manual (3rd edn.). Mountain View, CA: CPP.
  • Maslach, C., Schaufeli, W.B., & Leiter, M.P. (2001). Job burnout. Annual Review of Psychology, 52, 397422.
  • Melamed, S., Shirom, A., Toker, S., Berliner, S., & Shapira, I. (2006). Burnout and risk of cardiovascular disease: Evidence, possible causal paths, and promising research directions. Psychological Bulletin, 132, 327353.
  • Meyer, J.P., Allen, N.J., & Smith, C.A. (1993). Commitment to organizations and occupations: Extension and test of a three-component conceptualization. Journal of Vocational Behavior, 78, 538551.
  • Moyle, P. (1995). The role of negative affectivity in the stress process: Tests of alternative models. Journal of Organizational Behavior, 16, 647668.
  • OECD (2005). Attracting, developing and retaining effective teachers—Final report: Teachers matter. Paris: OECD.
  • Ostroff, C. (1992). The relationship between satisfaction, attitudes, and performance: An organizational level analysis. Journal of Applied Psychology, 77, 963974.
  • Ramm, G., Prenzel, M., Baumert, J., Blum, W., Lehmann, R., Leutner, D., Neubrand. M., Pekrun, R., Rolff, H.-G., Rost, J., & Schiefele, U. (2006). PISA 2003: Dokumentation der Erhebungsinstrumente. Münster, Germany: Waxmann.
  • Raudenbush, S.W., & Bryk, A.S. (2002). Hierarchical linear models (2nd edn.). Thousand Oaks, CA: Sage.
  • Raudenbush, S.W., Bryk, A.S., Cheong, Y., & Congdon, R.T. (2004). HLM 6: Hierarchical linear and nonlinear modeling. Lincolnwood, IL: Scientific Software International.
  • Salanova, M., Agut, S., & Peiro, J.M. (2005). Linking organizational resources and work engagement to employee performance and customer loyalty: The mediation of service climate. Journal of Applied Psychology, 90, 12171227.
  • Schaarschmidt, U., & Fischer, A. (1997). AVEM—ein diagnostisches Instrument zur Differenzierung von Typen gesundheitsrelevanten Verhaltens und Erlebens gegenüber der Arbeit [AVEM—An instrument for diagnosing different types of work- and health-related behavior and experience]. Zeitschrift für Differentielle und Diagnostische Psychologie, 18, 151163.
  • Schafer, J.L., & Graham, J.W. (2002). Missing data: Our view of the state of the art. Psychological Methods, 7, 147177.
  • 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, 293315.
  • Schaufeli, W.B., & Buunk, B.P. (2003). Burnout: An overview of 25 years of research and theorizing. In M.J.Schabracq, J.A.M.Winnubst, & C.L.Cooper (Eds.), Handbook of work and health psychology (pp. 383425). Chichester: John Wiley & Sons.
  • Schaufeli, W., & Enzmann, D. (1998). The burnout companion to study and practice: A critical analysis. London: Taylor & Francis.
  • Semmer, N. (1996). Individual differences, work stress and health. In M.J.Schabracq, J.A.M.Winnubst, & C.L.Cooper (Eds.), Handbook of work and health psychology (pp. 5386). Chichester: John Wiley & Sons.
  • Skinner, E.A., & Belmont, M.J. (1993). Motivation in the classroom: Reciprocal effects of teacher behaviour and student engagement across the school year. Journal of Educational Psychology, 85(4), 571581.
  • Trautwein, U., Lüdtke, O., Marsh, H.W., Köller, O., & Baumert, J. (2006). Tracking, grading, and student motivation: Using group composition and status to predict self-concept and interest in ninth-grade mathematics. Journal of Educational Psychology, 98(4), 788806.
  • Tymms, P. (2004). Effect sizes in multilevel models. In I.Schagen & K.Elliot (Eds.), But what does it mean? The use of effect sizes in educational research (pp. 5566). London: National Foundation for Educational Research.
  • Vandenberghe, R., & Huberman, A.M. (1999). Understanding and preventing teacher burnout: A sourcebook of international research and practice. Cambridge: Cambridge University Press.
  • Warm, T.A. (1989). Weighted likelihood estimation of ability in item response theory. Psychometrika, 54, 427450.
  • Wright, T.A., & Cropanzano, R. (1998). Emotional exhaustion as a predictor of job performance and voluntary turnover. Journal of Applied Psychology, 83, 486493.