Psychology of Personality Research and Well-Being Research Matrix

TheBigFivePersonalityFactorsasPredictorsofChangesAcrossTimeinBurnoutandItsFacets.pdf

Home >Psychology homework help >Psychology of Personality Research and Well-Being Research Matrix

The Big Five Personality Factors as Predictors of Changes Across Time in Burnout and Its Facets

Galit Armon,1 Arie Shirom,2 and Samuel Melamed3 1Haifa University, Israel

2Tel Aviv University, Israel 3Academic College of Tel Aviv-Jaffa, Israel

ABSTRACT We tested the effects of Neuroticism and Conscientious- ness on burnout across time, controlling for age, gender, work hours, and depressive symptoms. Our theoretical model included both global burnout and its physical, emotional, and cognitive facets, consistent with the bifac- tor approach to modeling second-order constructs in structural equation modeling. Data were gathered from 1,105 respondents (63% men) who completed questionnaires at Time 1 (T1) and approximately 24 months later at Time 2 (T2). Neuroticism positively predicted T1 global burnout and negatively predicted T1 and T2 emotional exhaustion. Conscientious- ness negatively predicted T1 global burnout and T1 and T2 cognitive weariness, and positively predicted T1 and T2 emotional exhaustion. Our gender-specific exploratory analysis revealed that for each gender, Neuroti- cism and Conscientiousness predicted different facets of burnout at T1 and T2. We recommend that future research test the possibility that the asso- ciations of Neuroticism and Conscientiousness with global burnout and its facets may be gender specific.

Our major objective in this study was to investigate the across-time effects of personality traits on burnout and its facets. Burnout has most frequently been conceptualized as individuals’ affective responses to the depletion of their energetic resources following exposure to chronic job stress (Halbesleben & Buckley, 2004; Shirom, 2003). We focused on burnout because it has been found to be associated with workers’ health impairment (Melamed, Shirom, Toker, Berliner, & Shapira, 2006) and reduced job performance (e.g.,

Correspondence concerning this article should be addressed to Galit Armon, Department of Psychology, Haifa University, Haifa 31905, Israel. Email: galitarmon@ gmail.com.

Halbesleben & Bowler, 2007). We tested the influence of personality traits, based on the Big Five (BF) personality factors (Extraversion, Agreeableness, Conscientiousness, Neuroticism, and Openness to Experience; McCrae & John, 1992), on burnout, assessed at two time points about two years apart. We used a longitudinal design following Fraley and Roberts (2005), who emphasized the importance of exam- ining the effects of personality traits on affects over time in order to better understand the processes influencing stability and change in affects. We argue that for any burnout prevention effort, knowledge of the individual characteristics implicated in the etiology of burnout is of considerable importance.

Personality predispositions could explain burnout and changes in its level over time for several reasons. First, individuals with certain personality traits may self-select into highly stressful and therefore burnout-conducive occupations and jobs (Garden, 1989). Second, certain personality traits may predispose individual employees to experience stressors more intensely, thus subsequently eliciting burnout. Watson and Clark (1984) described Neuroticism as the disposition to interpret events negatively; plausibly, this predispo- sition could adversely influence levels of burnout over time. Third, in addition to their influence on the experiencing and appraisal of stress, personality traits may influence coping with stressors (Connor-Smith & Flachsbart, 2007), thereby contributing to the etiology of burnout over time. Thus, for example, past studies have shown that Conscientiousness predisposes a person to handle stress more efficiently (e.g., Bruck & Allen, 2003; Wayne, Musisca, & Fleeson, 2004); plausibly, this predisposition could act as a buffer against burnout over time.

In the current study, we conceptualized burnout as a multidi- mensional construct as assessed by the Shirom-Malamed Burnout Measure (SMBM; Shirom & Melamed, 2006; see review of Melamed et al., 2006). The formulation of the burnout construct and its three facets of physical fatigue (i.e., feeling tiredness and low energy at work), emotional exhaustion (i.e., feeling lacking of the energy to display empathy to others at work), and cognitive weariness (i.e., feelings of reduced mental agility on the job) is based on the con- servation of resources (COR) theory. These three facets represent the depletion and draining of three closely interrelated and individu- ally possessed energetic resources (Hobfoll & Shirom, 2000) and therefore share a significant amount of their variance. This shared

Armon, Shirom, & Melamed404

variance reflects what we call global burnout, that is, the underlying latent construction that represents the core meaning of burnout (see Thoresen, Kaplan, Barsky, Warren, & Chermont, 2003). The content of each domain-specific facet of burnout is interpreted as an outcome of depletion of energetic resources specific only to that domain-specific facet and not to the variance shared among all three facets. In structural equation modeling (SEM) terminology, this modeling strategy is referred to as bifactor modeling (Chen, West, & Sousa, 2006).

A meta-analytic study of the relationships between personality variables and burnout concluded that employee personality is con- sistently related to burnout (Alarcon, Eschleman, & Bowling, 2009). With few exceptions (Goddard, O’Brien, & Goddard, 2006; Mills & Huebner, 1998; Miner, 2007; Piedmont, 1993), past studies of burnout and the BF model were based on a cross-sectional design and were, therefore, able to point to a covariation but could not confirm the BF effects over time. In addition, the few longitudinal studies only evaluated a single occupational category or employer using small samples, raising questions about the generalizability of the findings (Reise, Waller, & Comrey, 2000). No prior study con- trolled for the effect of depressive symptoms on burnout, which, for reasons explained below, is necessary in any study on the anteced- ents or consequences of burnout (see review of Melamed et al., 2006). Most past studies on the association between the BF and burnout did not control for age. The present longitudinal study improves upon the earlier longitudinal studies by using a large rep- resentative sample of participants, by excluding respondents whose burnout and depressive symptoms scores were likely to be influenced by a chronic disease or antidepressive medications, and by using analytical methods that simultaneously tested all our hypotheses while controlling for measurement errors. While past studies almost exclusively used the Maslach Burnout Inventory (MBI; Maslach & Jackson, 1986) to assess burnout, we used an alternative, theoreti- cally grounded measure of burnout (for conceptual problems and methodological shortcomings of the MBI, see the review of Kris- tensen, Borritz, Villadsen, & Christensen, 2005). To the best of our knowledge, no prior study has examined the moderating effect of gender on the relationships among personality traits and burnout; for the reasons explained below, the current study attempted to fill this void.

The Big Five and Facets of Burnout 405

Study Hypotheses

Based on the conceptual meaning of each of the BF factors and relevant past studies, we formulated specific hypotheses focusing only on Neuroticism and Conscientiousness for two major reasons. First, Neuroticism and Conscientiousness have been found to predict many work-related outcomes, such as career success and deviance, and represent maladaptive and adaptive coping styles, respectively (Barrick & Mount, 2000; Colbert, Mount, Harter, Witt, & Barrick, 2004; Judge, Higgins, Thoresen, & Barrick, 1999); therefore, they are highly likely to influence the availability and utilization of coping resources (see the meta-analysis of Connor-Smith & Flaschbart, 2007). There is a body of evidence (Schaufeli & Enzmann, 1998) supporting the theoretical argument that when individuals have coping resources to address work-related demands, they are unlikely to manifest strains such as burnout (Hobfoll, 2002). Second, Neu- roticism, the tendency to experience negative affect, has been asso- ciated with negative health outcomes (see the review of Lahey, 2010). Conscientiousness, the tendency to be organized, thorough, and reli- able (Goldberg, 1990; McCrae & John, 1992), has been linked to positive health outcomes (Kern & Friedman, 2008). The health out- comes typically associated with Neuroticism (such as depressive symptoms; see McCrae & John, 1992) and Conscientiousness (such as physical fitness; see Kern & Friedman, 2008) have been linked to burnout (Melamed et al., 2006). The linkages of Agreeableness, Extraversion, and Openness with global burnout and its three facets will be investigated on an exploratory basis.

Neuroticism

Neuroticism consumes resources and is likely to lead to resource depletion or burnout because the higher its level, the more pro- nounced is the tendency to view the world pessimistically and inter- pret many stimuli as threatening (McCrae & John, 1992). Therefore, neurotic individuals are likely to invest resources in dwelling on their internal affective states rather than in addressing work-related demands (Connor-Smith & Flaschbart, 2007). Many past cross- sectional studies found Neuroticism to be positively correlated with all of the MBI’s subscales (see the meta-analysis of Alarcon et al., 2009; Swider & Zimmerman, 2010). Following this accumulated evi- dence, we expected Neuroticism to be a positive predictor of the

Armon, Shirom, & Melamed406

baseline and follow-up levels of global burnout. In addition, follow- ing well-documented effects of Neuroticism on physical (Lahey, 2010), emotional (Judge et al., 1999), and cognitive (Colbert et al., 2004) symptoms, we also expect it to positively predict the physical, emotional, and cognitive facets of burnout.

Conscientiousness

Individuals who score high on Conscientiousness are characterized by careful planning, effective organization, and efficient time man- agement, allowing them to accomplish more in the time available. From a COR perspective, Conscientiousness acts as a resource enabling employees to efficiently deploy their work-related reso- urces, thus conserving energy and reducing the likelihood of burnout (Halbesleben & Buckley, 2004). Conscientious indi- viduals tend to use proactive, rational, problem-focused coping (Connor-Smith & Flaschbart, 2007), further reducing the likelihood of depleting their resources in managing work-related stresses (e.g., Bruck & Allen, 2003; Wayne, Musisca, & Fleeson, 2004). Studies have found Conscientiousness to negatively predict the MBI sub- scales (Alarcon et al., 2009; Swider & Zimmerman, 2010). There- fore, we expected that people rated high on Conscientiousness would experience lower levels of global burnout. Furthermore, as highly conscientious individuals have been characterized as thought- ful, planful, organized, and thorough (Goldberg, 1990), we expected that people who are high on Conscientiousness would experience lower levels of cognitive weariness, the facet of burnout reflecting one’s feeling of having reduced thinking ability. Because Conscien- tiousness is positively associated with customer service quality, social relationship quality, and positive teamwork (Barrick & Mount, 2000), Conscientiousness presumably acts as a buffer against emotional exhaustion, the unique content of the interper- sonal facet of global burnout.

Below we provide the theoretical rationale and empirical evidence supporting our decision to use four major control variables to test our hypotheses: depressive symptoms, age, work hours, and gender. All approaches to conceptualizing burnout include a component of felt fatigue or low levels of physical energy, which also appear among the criteria leading to diagnosis of depressive symptoms or dysthymia, according to the DSM-IV (American Psychiatric Association, 1994).

The Big Five and Facets of Burnout 407

Therefore, there is a conceptual overlap between burnout and depres- sive symptoms (Suls & Bunde, 2005). We maintain, however, that burnout is distinct in that it is dependent on the quality of the workplace social environment (Schaufeli & Enzmann, 1998), whereas depression is a global state that can pervade virtually every aspect of an individual’s environment (Suls & Bunde, 2005). Empirically, depression and burnout have been shown in quantitative (Glass & McKnight, 1996) and qualitative (Melamed et al., 2006; Schaufeli & Enzmann, 1998) reviews to be distinct from each other and to be differentially associated with disease endpoints. Still, because of the conceptual overlap, we control for depressive symptoms in our analy- sis. Of all biographical characteristics, age has most consistently been found to be associated with burnout (Schaufeli & Enzmann, 1998). In addition, studies of adult personality development have suggested that there are noticeable changes in the mean level of all the BF factors from adolescence until the age of 30 or so (McCrae & Costa, 2003): Neuroticism, Extraversion, and Openness decline, whereas Agree- ableness and Conscientiousness increase. Work hours are an impor- tant antecedent of burnout. Many past studies found that employees experience more burnout when they work more hours per week (Schaufeli & Enzmann, 1998). There are indications in the literature that the number of work hours per week is an indicator of work- related stress (e.g., Linzer et al., 2001). Given that burnout can be viewed as a proxy variable reflecting the impact of work-related stressors on one’s energetic resources (Shirom, 2003), we control for it in our analysis. Finally, there is evidence that burnout varies as a function of gender. A recent meta-analysis (Purvanova & Muros, 2010) found that women are somewhat more emotionally exhausted than men. Rather robust gender differences—persisting across a diverse array of measures, data sources, ages, and cultures—have also been found in personality traits (see the recent review of Schmitt, Realo, Voracek, & Allik, 2008). In general, when assessed in terms of the BF model of personality, women report higher levels of Neuroti- cism, Extraversion, Agreeableness, and Conscientiousness than men (Schmitt et al., 2008).

Following the body of evidence presented above on gender differ- ences in burnout levels and in the levels of the BF, we investigated gender differences in the across-time association between the BF and burnout in addition to using gender as a control variable. However, because no past study has focused on gender differences in the effects

Armon, Shirom, & Melamed408

of the BF on burnout, we tested these gender differences on an exploratory basis only.

METHOD

Participants

Study participants (N = 1,930: 1,221 males, 709 females) were all appar- ently healthy, employed adults who were sent by their employer to the Center for Periodic Health Examinations at the Tel Aviv Sourasky Medical Center for a routine health examination at Time 1 (T1) and Time 2 (T2), on average about 24 months apart (M = 737.78 days, SD = 340.69). At T1, they represented 92% of the center’s examinees, and all were voluntary participants in the study. We systematically checked for nonresponse bias at T1 and found that nonparticipants did not differ from participants with regard to any of the sociodemographic or biomedical variables.1 We also tested for attrition bias from T1 to T2;1 those exam- ined at T1 who did not return for a follow-up examination (46%) were more likely to be males, to be older (near retirement age), and to have self-reported a chronic disease at T1. We controlled for these possible sources of attrition bias in our data analyses, as explained below.

We excluded 439 respondents (292 males and 147 females) from the study based on the following criteria: those who self-reported having been diagnosed at T1 or T2 with cardiovascular disease (CVD), diabetes, cancer, or previous stroke or mental crisis. We also excluded participants who reported regularly taking antidepressants or any lipid-lowering drug or steroids because the disease or the medication could impact the level of depression or burnout (see the review of Melamed et al., 2006). We also excluded 376 respondents (195 males, 181 females) who were not actively employed at either T1 or T2, as well as those who worked on a part-time basis (fewer than 3 hours per day), because the assessment of burnout is contextualized in the work domain. Finally, we excluded a few cases with missing data on one or more of the study’s variables. Thus, the final sample consisted of 1,105 apparently healthy employees (799 males, 406 females). In a separate report conducted on the same sample, we found that the respondents’ physiological parameters were not significantly dif- ferent from those obtained in other large-scale studies (Shirom, Melamed, Rogowski, Shapira, & Berliner, in press).

Respondents at T1 had completed a mean of 15.64 (SD = 2.76) years of education, 16.03 (SD = 2.65) for men and 14.99 (SD = 2.82) for women. On

1. Detailed results of the entire sample analysis and the gender-specific analyses are available from the authors upon request.

The Big Five and Facets of Burnout 409

average, 7.8% (6.4% of men and 10% of women) were single, 84.8% (88.7% of men and 78% of women) were married or lived with a partner, 1.1% (1% of men and 1.2% of women) were widowed, and 6.3% (3.9% of men and 10.8% of women) were divorced or separated. They had a mean of 2.44 (SD = 1.22) children: 2.51 (SD = .47) for the men and 2.32 (SD = .09) for the women. In terms of organizational level, an average of 33% of the respondents (24.6% of men and 48.5% of women) were rank-and-file employees, not in charge of other employees; 12% (12% of men and 12.8% of women) were first-level supervisors or forepersons; 27.4% (28.8% of men and 24.6% of women) were middle managers; and 27% (34.6% of men and 14% of women) were managers in charge of other managers.

The study protocol was approved by the ethics committees of the Sourasky Medical Center and the Faculty of Management at Tel Aviv University. Participants were recruited individually by an interviewer while waiting for their clinical examination. The interviewer explained the survey and asked for her or his voluntary participation. In return, partici- pants were promised detailed feedback of the results. Confidentiality was assured, and each participant signed a written informed consent form.

Measures

The study questionnaire covered background, occupational, psychologi- cal, and physical morbidity factors. The multi-item indices constructed for the study had all been included in previous research in Israel (see the review of Melamed et al., 2006), which demonstrated high reliability and construct validity.

Work hours and age. Work hours were represented by the reported average number of work hours per week (one item). Age was reported by the subjects.

The Big Five. Individual differences on personality dimensions were assessed by the Big Five Mini-Marker Scale, Brief Version (Saucier, 1994), consisting of 40 adjectives (eight for each of the five Big Five: Neuroticism, Extraversion, Openness to Experience, Agreeableness, and Conscien- tiousness) that may be worded positively or negatively. The brief version was chosen to maintain the interest of participants and to minimize respondent refusal. It has been shown to have adequate reliability and validity estimates (Saucier, 1994) and has been adopted in other studies on personality and burnout (e.g., Ghorpade, Lackritz, & Singh, 2007) as well as on personality and other emotions (e.g., Wong et al., 2007).

Respondents were asked to indicate how accurately or inaccurately the adjectives describe them. Responses were given on a 9-point Likert scale

Armon, Shirom, & Melamed410

ranging from 1 (extremely inaccurate) to 9 (extremely accurate). We used the back-translation procedure for testing the reliability of our translation of this instrument; three independent judges assessed the adequacy of the translation, with inter-rater reliability of .83. Confirmatory factor analysis (CFA) confirmed the theoretically expected five-factor structure.1 During the testing of the original 40 items by a CFA, we decided to remove two items that did not load as expected, primarily because of differences in the meanings of the original English-language items and the Hebrew- language translation. Other researchers have reported similar difficulties that largely expressed cross-cultural differences in the meanings of the adjectives used in the original version of the Mini-Marker Scale (e.g., Garcia, Aluja, & Garcia, 2004). Following the support we found for the BF factor structure in our final measurement model, we tested each fac- tor’s measurement model. For the BF multi-item latent factors, we con- structed parcels using the radial parceling algorithm (Rogers & Schmitt, 2004) and calculated their internal consistency reliability coefficients (alpha; see Table 1).

Burnout. Burnout was assessed using the Shirom-Melamed Burnout Measure2 (SMBM), which asks respondents to report the frequency of recently experienced loss of energy feelings at work. The SMBM measure has been validated in several studies (e.g., Shirom & Melamed, 2006). The dimensional structure of the SMBM was validated in prior research conducted in several countries (for further details, see Shirom, Nirel, & Vinokur, 2006; Vinokur, Pierce, & Lewandowski-Romps, 2009). All items were scored on a 7-point frequency scale ranging from 1 (almost never) to 7 (almost always). A CFA confirmed the theoretically expected SMBM three-factor structure1 and led to our constructing three subscales, each representing one facet. The three subscales were a six-item subscale of physical fatigue (e.g., “I feel physically drained” and “I feel tired at work”), a five-item subscale of cognitive weariness (e.g., “In my job, I have difficulty concentrating” and “I have difficulty thinking about complex things at work”), and a three-item subscale of emotional exhaustion (e.g., “I feel I am unable to be sensitive to the needs of coworkers”). A previous study documented the superior fit to the data of a second-order model of global burnout as assessed by the SMBM, in which items loaded on the three first-order factors—physical fatigue, emotional exhaustion, and cog- nitive weariness—and the first-order factors loaded on a second-order factor labeled global burnout, relative to one-factor, two-factor, and three-factor models (Shirom et al., 2006). In our research we adopted the

2. The SMBM, its norms, and instructions concerning its use can be downloaded from the following sites: www.shirom.org or www.tau.ac.il/~ashirom/.

The Big Five and Facets of Burnout 411

Ta b

le 1

M e

a n

s, St

a n

d a

rd D

e v

ia ti

o n

s, a

n d

In te

rc o

rr e

la ti

o n

s o

f A

ll St

u d

y V

a ri

a b

le s

1 2

3 4

5 6

7 8

9 10

11 12

13 14

15 16

1. G

lo b

al b

u rn

o u

t, T

2 —

— —

2. G

lo b

al b

u rn

o u

t, T

1 .6

9* —

— —

— 3.

P h

ys ic

al fa

ti gu

e, T

2 .9

1* .6

3* —

— —

4. P

h ys

ic al

fa ti

gu e,

T 1

.6 3*

.9 0*

.6 7*

— —

5. E

m o

ti o

n al

ex h

au st

io n

, T

2 .7

0* .4

4* .7

0* .3

2* —

— —

6. E

m o

ti o

n al

ex h

au st

io n

, T

1 .4

2* .6

7* .3

0* .4

3* .5

2* —

— —

—

7. C

o gn

it iv

e w

ea ri

n es

s, T

2 .8

7* .6

1* .8

7* .4

8* .7

0* .4

2* —

— —

8. C

o gn

it iv

e w

ea ri

n es

s, T

1 .6

0* .8

6* .4

8* .6

2* .3

5* .4

8* .6

5* —

— —

— 9.

N eu

ro ti

ci sm

.2 0*

.2 5*

.1 7*

.1 8*

.2 1*

.2 8*

.1 6*

.2 0*

— —

— 10

. C

o n

sc ie

n ti

o u

sn es

s -.

12 *

-. 12

* -.

10 *

-. 12

* -.

09 *

-. 05

-. 11

* -.

10 *

-. 13

* —

— —

— 11

. A

gr ee

ab le

n es

s -.

05 -.

07 *

-. 01

-. 02

-. 18

* -.

21 *

-. 02

-. 09

* .3

1* —

— —

12 .

E xt

ra ve

rs io

n .2

5* .2

8* .2

1* .2

2* .2

3* .2

2* .2

3* -.

44 *

.0 9*

.0 3

— —

13 .

O p

en n

es s

.0 5

.0 6

.0 5

.0 7*

.0 8*

.0 1*

.0 1

-. 30

* .2

8* .1

2* .2

7* —

— —

14 .

D ep

re ss

iv e

sy m

p to

m s

.5 0*

.5 5*

.5 0*

.5 2*

.2 7*

.2 8*

.4 2*

.4 9*

-. 15

* -.

10 *

-. 02

.2 0*

.0 2

— —

15 .

A ge

-. 11

* -.

10 *

-. 11

* -.

15 *

-. 07

* -.

03 -.

07 *

-. 03

-. 06

.0 3

.0 2

.0 3

.0 1

.0 2

— —

— 16

. W

o rk

h o

u rs

-. 08

* -.

11 *

-. 09

* -.

08 *

-. 02

-. 01

-. 08

* -.

14 *

.0 5

.0 6

-. 05

.0 3

.1 1*

-. 10

* -.

10 *

— —

17 .

G en

d er

.1 9*

.2 5*

.2 3*

.0 2

-. 01

.1 3*

.1 5*

.0 2

.1 1*

.0 8*

-. 13

* -.

26 *

.0 4

-. 37

* —

C ro

n b

ac h

’s al

p h

a .9

3 .9

2 .9

1 .9

0 .9

1 .8

9 .9

1 .9

0 .7

3 .8

0 .7

1 .7

4 .7

6 .7

9 —

— —

M 2.

03 2.

19 §

2. 32

2. 47

§ 1.

80 1.

92 §

1. 82

1. 99

§ 4.

90 5.

56 5.

74 4.

41 5.

30 1.

24 45

.6 7

49 .7

3 .3

7 S

D .8

3 .8

2 1.

05 1.

03 .9

4 .9

2 .9

5 .9

4 .9

1 .5

4 .6

4 .9

9 .7

8 .3

2 9.

64 10

.6 8

.4 8

N ot

e. N

= 1,

10 5.

§ S

ig n

ifi ca

n tl

y d

if fe

re n

t fr

o m

th e

co rr

es p

o n

d in

g T

2 va

lu es

, at

p <

.0 5.

A fu

ll co

rr el

at io

n m

at ri

x is

av ai

la b

le fr

o m

th e

au th

o rs

u p

o n

re q

u es

t. *p

< .0

Armon, Shirom, & Melamed412

bifactor model, which has several major advantages over the standard second-order factor (see Chen et al., 2006).

Depression. Depression was measured using the Patient Health Ques- tionnaire (PHQ9), the depression section of a patient-oriented, self- administered instrument derived from the PRIME-MD (Kroenke et al., 2009). The PHQ9 includes nine potential symptoms of depression (e.g., feeling down, depressed, or hopeless; having little interest or pleasure in doing things). Respondents were asked to rate the frequency of experienc- ing each of the nine potential symptoms during the previous 2 weeks on a scale ranging from never (1) to almost always (4). This measure has been used in many studies; it has been extensively validated for major depression and is often used in screening populations (see Kroenke et al., 2009).

Statistical Analyses

We used structural equation modeling (SEM) with AMOS 18 software (Arbuckle & Wothke, 1999) to test our hypotheses. Following widely accepted SEM practices (Kline, 2004), we allowed the exogenous variables (the BF factors) to correlate, included them in the model correlations among the measurement errors of the same indicators at T1 and T2, and constrained the factor loadings of the same indicators at T1 and T2 to be equal. To facilitate understanding of the results, we did not include these correlations in Figure 1.1 The BF latent factors and the latent factor of depression were each indicated by a single indicator, namely the mean of the items included in it, and we fixed the error variance of each of these latent factors using the well-known formula of multiplying variance by (1 − reliability). We did not include the above single indicators in Figure 1. In an initial version of our analyses, we also modeled yet another control variable, representing the exact T1–T2 lag time in days for each respon- dent. Because it did not have any significant effect on either of our T2 latent factors,1 we removed it from Figure 1. A major advantage of SEM lies in its ability to estimate a model’s parameters while correcting for the biasing effects of random measurement error. We report on three goodness-of-fit indices—the normed fit index (NFI), the non-normed fit index (NNFI, also known as TLI, which is recommended for large samples), and the comparative fit index (CFI)—and two misfit indices: the standardized root mean square residual (SRMR) and the root mean square error of approximation (RMSEA) with a 90% confidence interval (McDonald & Ho, 2002). It has been suggested (Hu & Bentler, 1999) that fit indices close to or above .95, combined with SRMR and RMSEA below .06, can be considered indicative of good fit of the tested model to the data. These threshold values for approximate fit indices have been

The Big Five and Facets of Burnout 413

N C

OEACN

A E O

T 1 D

e p r e s s io

R 2

= . 3 2

N C

O A

T 1 G

lo b a l

B u r n o

u t

R 2 =

. 5 7

T 2 G

lo b a l

B u r n o

u t

R 2 =

. 5

T 1

h y s ic

a l

F a

ti g

u e

R 2 =

. 0

T 1

m o

t io

n a l

E x h

a u

s t io

R 2 =

. 3

T 1

o g n it iv

W e a r in

e s s

R 2 =

. 1

T 2

o g n it iv

W e a r in

e s s

R 2 =

. 3

T 2

m o

t io

n a l

E x h

a u

s t io

R 2 =

. 3

T 2

h y s ic

a l

F a

ti g

u e

R 2 =

. 2

A g e

- . 2 3 *

- . 2 1 *

- . 4 4 *

- . 8 3 *

. 3 3 *

- . 4 8 *

. 2 2 *

. 1 6 *

- . 1 6 *

W o r k

h o u r s

- . 2 9 *

. 5 0 *

. 2 1 *

n . s

. 5

3 *

* 8

4 .

* 6

4 .

* 4

4 .

* 0

6 .

* 6

5 *

3 2 .

. 7

5 *

. 6

9 *

. 6

8 *

. 6

8 *

* 0

8 .

* 5

6 .

* 9

6 .

* 2

8 .

* 1

7 .

. 6

6 *

. 5

1 *

* 3

7 .

* 5

7 .

* 9

6 .

* 4

7 .

* 1

5 .

* 8

6 .

. 6

6 *

. 6

6 *

. 7

2 *

. 5

8 *

. 5

8 *

. 6

6 *

. 5 6 *

. 5 4 *

. 3

0 *

. 5

5 *

. 6

3 *

. 5

6 *

. 4

1 *

. 4

5 *

. 4

1 *

. 6

3 *

. 6

3 *

. 7

3 *

. 3 7 *

n . s

- . 1 7 *

n . s

- . 1 8 *

- . 2 1 * . 1 8 *

n . sn . s

. 3

9 *

. 8

1 *

. 6

6 *

. 5

1 *

n . s

G e n d e

rn . s

. 2 7 *

* 0

5 .

* 5

3 .

* 2

5 .

Fi g

u re

1 T

h e

B ig

Fi v

e ef

fe ct

s on

g lo

b a

l b

u rn

ou t

a n

d it

s fa

ce ts

a cr

os s

ti m

e. N

= N

eu ro

ti ci

sm ;

C =

C on

sc ie

n ti

ou sn

es s;

A =

A g

re ea

b le

n es

s; E

= Ex

tr a

v er

si on

; O

= O

p en

n es

s; n

.s .

= n

ot si

g n

if ic

a n

t. To

fa ci

li ta

te u

n d

er st

a n

d in

g of

th e

ef fe

ct s

of th

e B

ig Fi

v e

on th

e la

te n

t fa

ct or

s in

th e

m od

el ,

w e

re p

ea t

th e

a b

ov e

fi v

e fa

ct or

s on

th e

fi g

u re

's le

ft a

n d

ri g

h t

si d

es ,

a n

d a

ls o

a t

th e

b ot

to m

of th

e fi

g u

re . So

li d

a n

d b

ro k

en a

rr ow

s re

p re

se n

t si

g n

if ic

a n

t (p

< .0

5) a

n d

in si

g n

if ic

a n

t ef

fe ct

s, re

sp ec

ti v

el y

. A

ll co

ef fi

ci en

ts a

re st

a n

d a

rd iz

ed re

g re

ss io

n co

ef fi

ci en

ts . T

h e

co rr

el a

ti on

s a

m on

g th

e ex

og en

ou s

fa ct

or s

a re

n ot

re p

re se

n te

d .

Er ro

r v

a ri

a n

ce s

of ea

ch in

d ic

a to

r of

th e

la te

n t

fa ct

or s

w il

l b

e se

n t

u p

on re

q u

es t.

Armon, Shirom, & Melamed414

questioned by recent studies (e.g., Yuan, 2005). In the current study, we used fit indices primarily to screen out unacceptable models; moreover, we focused on comparing nested models using the chi-square (c2) differ- ence test (see Yuan, 2005).

Testing the measurement models. We first tested the measurement model of the SMBM at T1 and T2 and found an acceptable fit, respectively: c2 (33, n = 1,105) = 89.65, 87.82, with normed, non-normed, and compara- tive fit indices all above .99, .99, RMSEA = .04, .04 (the 90% confidence interval for RMSEA ranged from .03 to .05, .03 to .05). Next, we tested the measurement model for all the variables presented in Figure 1 and found an acceptable fit: c2 (373, n = 1,105) = 1234.35 with normed, non- normed, and comparative fit indices all above .95, RMSEA = .04 (the 90% confidence interval for RMSEA ranged from .04 to .05).

Because we conducted exploratory analyses of gender differences in the structural model, we also tested the measurement model for all the vari- ables separately for men and women. The results demonstrated an accept- able fit.1 We tested the equivalence of the male-female models by including equality constraints across the groups for factor loadings, paths of direct influence, and the variances of the exogenous factors, (i.e., age, work hours, and the BF), testing the constrained model against the alternative model in which these parameters were freely estimated. The results demonstrated a good approximate fit to our data: c2 (791) = 1699.36 with normed, non-normed, and comparative fit indexes all above .93, RMSEA = .03 (the 90% confidence interval for RMSEA ranged from .03 to .04). Additionally, when we tested the constrained versus the uncon- strained model using the c2-difference test, the two models were not significantly different from one another, thus providing empirical support for our exploratory analysis of gender differences.

RESULTS

Descriptive Results

The intercorrelation matrix of the variables included in our analysis,1

and their means and standard deviations, are depicted in Table 1. Global burnout’s consistency coefficient was found to be .69, and the consistency coefficient of all its facets ranged between .30 and .91. However, the correlations of each T1 burnout facet with its T2 level (after controlling for the variance it shared with the other two facets, represented by global burnout at T1 or T2) ranged between .35 and .52 (see Figure 1), considerably lower than the coefficients reported

The Big Five and Facets of Burnout 415

above. On average, there was a significant decrease in the global burnout level and also in the levels of each of its facets from baseline to follow-up. The Cronbach internal consistency reliabilities of global burnout and its facets at T1 and at T2 were high (a range = .89–.93), and those of the BF factors (a range = .71–.80) were largely on par with those reported in a meta-analysis (Viswes- varan & Ones, 2000). Global burnout and its facets were negatively associated with age, in congruence with the findings of a meta- analytic study on correlations between age and facets of burnout (Brewer & Shapard, 2004). In addition, slightly negative associations were found between work hours and global burnout and its facets. Presumably, these associations were a result of the strong correlation between organizational level and work hours, since organizational level was negatively associated with burnout. The significant differ- ence in the percentage of men and women across organizational levels, c2(1,105) = 87.60, p < .05, that we found provided yet further justification for the inclusion of gender as a control variable and analyzing BF-burnout relations separately by gender. In congruence with a past review (Glass & McKnight, 1996), burnout was found to be associated with depression. The correlations between the BF and depression were consistent with those reported in past studies (Bagby, Quilty, & Ryder, 2008).

Testing the Structural Models

We tested our hypothesized structural model in which burnout is modeled by the bifactor model as described in Figure 1, where T1 and T2 global burnout account for the commonality of all the burnout items. Additionally, three domain-specific factors of global burnout (e.g., physical fatigue, emotional exhaustion, and cognitive weariness) account for the unique variance in the burnout items they predict, independently of global burnout. The results demonstrated a good approximate fit to our data (see Table 2, the bifactor model). The R-squares (disturbances) of the three facets of burnout at T1 and T2 indicate that the BF explained, on average, 19% and 33% of the unique variance of the three facets of burnout at T1 and T2, respec- tively (see Figure 1). Thus, there was sufficient unique variance to be accounted for over and above the contribution of global burnout. We used the c2 difference to test our decision to use the bifactor model instead of the second-order factor model. The results

Armon, Shirom, & Melamed416

Ta b

le 2

R e

su lt

s o

f th

e St

ru c

tu ra

l M

o d

e ls

(S E

M )

fo r

th e

E ff

e c

ts o

f th

e B

ig F

iv e

o n

G lo

b a

l B

u rn

o u

t a

n d

It s

Fa c

e ts

a t

T im

e 1

a n

d T

im e

M o

d el

S am

p le

c2 df

N F

I T

L I

C F

I R

M S

E A

C I

(9 0%

) fo

r R

M S

E A

S R

M R

C o

m p

ar is

o n

o f

M o

d el

D df

D c2

T h

e b

if ac

to r

m o

d el

F u

ll sa

m p

le 10

39 .5

7 38

2 .9

5 .9

6 .9

7 .0

4 .0

3– .0

4 .0

3 19

52 3.

10 *

W o

m en

70 9.

05 35

9 .9

2 .9

5 .9

6 .0

5 .0

4– .0

5 .0

4 19

14 0.

68 *

M en

70 3.

29 35

9 .9

5 .9

7 .9

8 .0

4 .0

3– .0

4 .0

4 19

23 1.

02 *

T h

e se

co n

d -o

rd er

m o

d el

F u

ll sa

m p

le 15

62 .6

7 40

1 .9

3 .9

4 .9

5 .0

5– .0

6 .0

4 W

o m

en 84

9. 73

37 8

.9 1

.9 3

.9 4

.0 6

.0 5–

.0 6

.0 5

M en

93 4.

31 37

8 .9

3 .9

5 .9

6 .0

5 .0

4– .0

5 .0

N ot

e. N

= 1,

10 5.

df =

d eg

re es

o f

fr ee

d o

m ;

N F

I =

n o

rm ed

fi t

in d

ex ;

T L

I =

T u

ck er

-L ew

is In

d ex

as n

o n

-n o

rm ed

fi t

in d

ex ;

C F

I =

co m

p ar

at iv

e fi

t in

d ex

;R M

S E

A =

ro o

t m

ea n

sq u

ar e

er ro

r o

f ap

p ro

xi m

at io

n ;C

I =

co n

fi d

en ce

in te

rv al

;S R

M R

= st

an d

ar d

iz ed

ro o

t m

ea n

sq u

ar e

re si

d u

al ;

D =

th e

am o

u n

t o

f ch

an ge

b et

w ee

n th

e tw

o n

es te

d m

o d

el s

co m

p ar

ed .

*p <

.0 5.

The Big Five and Facets of Burnout 417

indicated that the bifactor model fit the data better than the second- order model, thus providing strong support for our decision (see Table 2). Specifically, the second-order model represented a gain of 19 degrees of freedom for an excessive cost of c2 = 523.10. Following these results, we report below the results of the tests of the specific predictions linking Neuroticism and Conscientiousness with global burnout and its facets at T1 and T2 (Hypotheses 1 and 2).

Testing Specific Predictions

Figure 1 depicts the detailed results of the tests of the specific hypoth- eses. Hypothesis 1, which expected Neuroticism to positively predict global burnout and its three facets at baseline and follow-up, was partially supported. Neuroticism was a positive predictor of global burnout, but only at T1 (b = .21, p < .05). In addition, Neuroticism was found to significantly predict emotional exhaustion at both T1 and T2 (b = -.18, -.21, respectively; p < .05). However, contrary to our expectation, the path coefficient was negative rather than posi- tive. The expected associations of Neuroticism with cognitive weari- ness and with physical fatigue at both T1 and T2 were not supported. The results also partly supported Hypothesis 2, which expected Con- scientiousness to negatively predict global burnout and its cognitive and emotional facets at baseline and follow-up. Conscientiousness was found to negatively predict global burnout, but only at T1 (b = -.21, p < . 05), and to negatively predict T1 and T2 cognitive weariness (b = -.44, -.18, respectively; p < .05). However, contrary to our expectation, it was found to positively predict emotional exhaustion at both T1 and T2 (b = .34, .18, respectively; p < .05).

Some interesting findings emerged in our exploratory analysis on the other factors of the BF. Agreeableness was found to positively predict global burnout at T1 (b = .16, p < .05) and to negatively predict emotional exhaustion at both T1 and T2 (b = -.83, -.48, respectively; p < .05). Extraversion negatively predicted global burnout and only at T1 (b = -.16, p < .05).

Gender Differences: Exploratory Analysis

We tested on an exploratory basis whether gender moderated the hypothesized associations between the study variables. In order to test whether there were gender differences in the associations between Neuroticism and burnout and its facets, we used the

Armon, Shirom, & Melamed418

c2-difference test, constraining the relevant paths to be equal for men and women and comparing this model with one in which these paths were freely estimated. We found that the unconstrained model fit the data significantly better; it represented a gain of two degrees of freedom for an excessive cost of c2 = 13, thus indicating significant gender differences in the paths between Neuroticism and global burnout and its facets. Specifically, for women, Neuroticism was found to negatively predict emotional exhaustion at both T1 and T2 (b = -.68, -.48, respectively; p < .05). For men, Neuroticism was found to predict cognitive weariness positively at T1 and negatively at T2 (b = .16, -.25, respectively; p < .05).

Next, we used the c2 difference to test whether there were gender differences in the associations between Conscientiousness and burnout and its facets, constraining the relevant paths to be equal for men and women and comparing this model with one in which these paths were freely estimated. We found that the uncon- strained model fit the data significantly better; it represented a gain of six degrees of freedom for an excessive cost of c2 = 60, thus point- ing to significant differences. Specifically, for women, Conscien- tiousness negatively predicted T1 and T2 global burnout (b = -.35, -.17, respectively; p < .05) and negatively predicted cognitive weari- ness at T1 and T2 (b = -.17, -.28, respectively; p < .05), whereas it positively predicted emotional exhaustion at T1 and T2 (b = .42, .36, respectively; p < .05). We found the same pattern of association for men, but only at T1.

DISCUSSION

We investigated the across-time effects of the BF on global burnout and its facets, controlling for the effects of age, depressive symptoms, work hours, and gender. We focused on two BF factors, Neuroticism and Conscientiousness, and tested two hypotheses concerning these relationships in a large representative sample of apparently healthy employees. Overall, we did not find any support for our predicted effects of the BF on T2 level of global burnout. Moreover, most of the hypothesized associations of the BF with the T2 levels of burn- out’s facets did not obtain support in our SEM model. All tests of the hypothesized effects on T2 global burnout and its facets were con- ducted after controlling for the effect of T1 global burnout and the control variables.

The Big Five and Facets of Burnout 419

Some interesting results emerged with regard to the effects of the BF on the specific facets of global burnout. Specifically, the emo- tional facet of burnout was found to be significantly associated with Neuroticism, at both T1 and at T2, but the path coefficient was negative rather than positive in sign. Our failure to support the across-time association of Neuroticism with global burnout and its cognitive and physical facets could be due to an underrepresentation in the scale of the measure we used (Saucier, 1994), of items assessing the affective traits of depression, anger, and hostility, which have been found to be strong antecedents of physical and cognitive symp- toms (see, e.g., the reviews of Eckhardt, Norlander, & Deffenbacher, 2004; Miller, Smith, Turner, Guijarro, & Hallet, 1996). In other popular measures of the BF factors, such as the NEO-PI-R (Costa & McCrae, 1992), depression, anger, and hostility constitute important facets of Neuroticism.

With regard to Conscientiousness, as expected, a significant nega- tive association was found with the cognitive facet of burnout at both T1 and T2, indicating that Conscientiousness acts as a buffer against reduced thinking ability. Furthermore, Conscientiousness was sig- nificantly associated with the emotional facets of burnout at both T1 and T2, but in contrast to our expectation, this path coefficient was positive rather than negative in sign. It is plausible that for individu- als high on Conscientiousness, maintaining the effort to work hard and be efficient may lead in the long run to depletion of their emo- tional resources and result in emotional exhaustion.

We argue that an additional explanation for our failure to support most of our hypotheses regarding the associations of Neuroticism and Conscientiousness with global burnout and its facets across time could be due to the moderating effect of gender on these associations. Indeed, when we tested these associations on an exploratory basis separately for men and women, a different and interesting picture emerged. For each gender, Neuroticism was found to be associated across time with different facets of burnout. While for men, Neuroti- cism predicted the cognitive facet of burnout at both T1 and T2, for women, Neuroticism predicted the emotional facets of burnout at both T1 and T2. With regard to Conscientiousness, our gender- specific exploratory analyses revealed a significant association across time with both global burnout and its emotional and cognitive facets, but only for women. We may speculate that different social roles and challenges for men and women (Wood & Eagly, 2002) may account

Armon, Shirom, & Melamed420

for these results. As indicated above, the two BF factors of Neuroti- cism and Conscientiousness have consistently been found to be sig- nificant predictors of burnout, but almost no past study has tested the possibility that these effects differ by gender. Future research might address this issue.

An additional contribution of our study is that we were able to support the suggestion (Brouwer, Meijer, Weekers, & Baneke, 2008) that the bifactor model allows researchers modeling multidimen- sional constructs, such as burnout, to investigate the pathways leading to or from domain-specific factors while controlling for the general factor, represented in our study by global burnout. For example, the use of the bifactor model enabled us to find effects of Neuroticism and Conscientiousness on emotional exhaustion (representing the unique variance of the interpersonal dimension of burnout) that were qualitatively different from their effects on global burnout and the facet of cognitive weariness. This finding may have practical implications for occupations in which the interpersonal dimension of burnout is important in determining employee effec- tiveness, such as the caring professions.

Strengths, Limitations, and Implications

The present study has several strengths. First, it investigated the across-time relations between personality and burnout in a large sample and explored gender differences in these relationships. Second, we carefully excluded individuals who self-reported chronic illness or taking medications that could influence their burnout scores. No past study has applied these strict exclusion criteria. A third strength comes from using T1 global burnout, depressive symp- toms, age, and work hours as controls in our analyses of the effects of the BF on T2 global burnout and its facets. It is often argued that findings of studies that interrelate self-reported attitudes or affects to the BF are biased because of their failure to consider the influence of “third variables” on these relationships. Examples of these third variables are occupation-specific environmental conditions and life circumstances such as educational attainment. Presumably, these third variables, to the extent that they predate the T1 global burnout levels, had already influenced them before our T1 wave of measure- ment. Finally, all T1 and T2 variables were assessed using the same standard questionnaire at a single center.

The Big Five and Facets of Burnout 421

Our study has certain limitations. First, our findings could be biased because of the well-known “healthy worker effect,” which refers to the possibility that employees with elevated levels of burnout decided to change their place of work or stopped working, leaving their healthier colleagues to participate in our research. We assume that this had probably already occurred before our T1 (Gold- berg, 1990). Yet another limitation is that we did not test the possi- bility that the factors of the BF may interact in affecting burnout levels (Richardson, Wing, Steenland, & McKelvey, 2004), primarily because of insufficient evidence that such interactions could add significantly to the prediction of burnout. Third, in the current study, for reasons described above, we used the Mini-Marker Scale, which allows only a total score for each of the BF factors. We recommend that future studies consider using personality scales that allow an assessment of different facets of each of the BF factors, such as the NEO-PI-R.

Additionally, our design was based on only two waves of measure- ment. It has obvious advantages over testing cross-sectional relations because it controls for the confounding influence of time-invariant common causes (Grant & Langan-Fox, 2006). However, because the BF were not assessed during our T2 wave of measurement, we could not test and disconfirm the possibility of reverse causation. A design based on only two waves of measurement cannot provide information on the precise nature of intra-individual change over time (Dormann, 2001). For plotting individual growth curves, which may include linear as well as curvilinear terms, at least three waves of data mea- surement are needed. We recommend that future research in this area use multiwave repeated measurement within a longitudinal design to allow for the assessment of the unfolding of individual change over time. As noted, men and women were found to significantly differ in their distribution across organizational levels. Because of the sta- tistical difficulty of including a multinomial predictor violating the assumption of multivariate normality in our SEM analysis, we did not use organizational level as a control variable. Future research may investigate the possibility that gender, organizational level, and per- sonality traits interact in the prediction of burnout across time.

Finally, although T1 and T2 burnout scores are moderately highly correlated, we still found a slight but significant decline in the mean levels of global burnout and its facets across time. The decline in burnout level across time that is often found in longitudinal studies

Armon, Shirom, & Melamed422

(e.g., Houkes, Winants, & Twellaar, 2008) could possibly be partly due to employees adopting more effective strategies to cope with chronic stresses across time, to their acquiring additional coping resources with the passage of time such as ties of friendship and social support from others, and to age-related processes. Future research might investigate potential predictors of these decreases over time.

If our findings are corroborated by future studies, researchers testing structural equation models that include burnout should con- sider the first-order factors representing the construct, in addition to the higher-order representation of burnout, since these domain- specific factors could make a unique contribution to explaining outcome variables and they could be differentially related to anteced- ent variables. Our results suggest that personality traits are involved in the etiology of burnout over time in gender-specific ways. It follows that researchers testing the association between personality traits and burnout should formulate gender-specific hypotheses. An important next step in this field of research is to examine possible mechanisms underlying the longitudinal effects uncovered by our study.

REFERENCES

Alarcon, G., Eschleman, K. J., & Bowling, N. A. (2009). Relationships between personality variables and burnout: A meta-analysis. Work and Stress, 23, 244– 263.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author.

Arbuckle, J. L., & Wothke, W. (1999). AMOS 4.0 user’s guide. Chicago: Small- waters.

Bagby, R. M., Quilty, L. C., & Ryder, A. C. (2008). Personality and depression. Canadian Journal of Psychiatry/Revue Canadienne de Psychiatrie, 53, 14–25.

Barrick, M. R., & Mount, M. K. (2000). Select on conscientiousness and emo- tional stability. In E. A. Locke (Ed.), Handbook of principles of organizational behavior (pp. 15–28). Malden, MA: Blackwell.

Brewer, E. W., & Shapard, L. (2004). Employee burnout: A meta-analysis of the relationship between age or years of experience. Human Resource Development Review, 3, 102–123.

Brouwer, D., Meijer, R. R., Weekers, A. M., & Baneke, J. J. (2008). On the dimensionality of the Dispositional Hope Scale. Psychological Assessment, 20, 310–315.

Bruck, C. S., & Allen, T. D. (2003). The relationship between Big Five personality traits, negative affectivity, type A behavior, and work-family conflict. Journal of Vocational Behavior, 63, 457–472.

The Big Five and Facets of Burnout 423

Chen, F. F., West, S. G., & Sousa, K. H. (2006). A comparison of bifactor and second-order models of quality of life. Multivariate Behavioral Research, 41, 189–225.

Colbert, A. E., Mount, M. K., Harter, J. K., Witt, L. A., & Barrick, M. R. (2004). Interactive effects of personality and perceptions of the work situation on workplace deviance. Journal of Applied Psychology, 89, 599–609.

Connor-Smith, J. K., & Flachsbart, C. (2007). Relations between personality and coping: A meta-analysis. Journal of Personality and Social Psychology, 93, 1080–1107.

Costa, P. T., & McCrae, R. R. (1992). NEO-PI-R. Professional manual. Odessa, FL: Psychological Assessment Resources.

Dormann, C. (2001). Modeling unmeasured third variables in longitudinal studies. Structural Equation Modeling: A Multidisciplinary Journal, 8, 575–598.

Eckhardt, C., Norlander, B., & Deffenbacher, J. (2004). The assessment of anger and hostility: A critical review. Aggression and Violent Behavior, 9, 17–43.

Fraley, R. C., & Roberts, B. W. (2005). Patterns of continuity: A dynamic model for conceptualizing the stability of individual differences in psychological con- structs across the life course. Psychological Review, 112, 60–74.

Garcia, O., Aluja, A., & Garcia, L. F. (2004). Psychometric properties of Gold- berg’s 50 personality markers for the Big Five model: A study in the Spanish language. European Journal of Psychological Assessment, 20, 310–319.

Garden, A. M. (1989). Burnout: The effect of psychological type on research findings. Journal of Occupational Psychology (Vol. 62, pp. 223). British Psy- chological Society.

Ghorpade, J., Lackritz, J., & Singh, G. (2007). Burnout and personality. Journal of Career Assessment, 15, 240–256.

Glass, D. C., & McKnight, J. D. (1996). Perceived control, depressive symptoma- tology, and professional burnout: A review of the evidence. Psychology and Health, 11, 23–48.

Goddard, R., O’Brien, P., & Goddard, M. (2006). Work environment predictors of beginning teacher burnout. British Educational Research Journal, 32, 857– 874.

Goldberg, L. R. (1990). An alternative “description of personality”: The Big Five factor structure. Journal of Personality and Social Psychology, 59, 1216– 1229.

Grant, S., & Langan-Fox, J. (2006). Occupational stress, coping and strain: The combined/interactive effect of the Big Five traits. Personality and Individual Differences, 41, 719–732.

Halbesleben, J. R. B., & Bowler, W. M. (2007). Emotional exhaustion and job performance: The mediating role of motivation. Journal of Applied Psychology, 92, 93–106.

Halbesleben, J. R. B., & Buckley, M. R. (2004). Burnout in organizational life. Journal of Management, 30, 859–879.

Hobfoll, S. E. (2002). Social and psychological resources and adaptation. Review of General Psychology, 6, 307–324.

Hobfoll, S. E., & Shirom, A. (2000). Conservation of resources theory: Applica- tions to stress and management in the workplace. In R. T. Golembiewski (Ed.),

Armon, Shirom, & Melamed424

Handbook of organization behavior (2nd revised ed., pp. 57–81). New York: Dekker.

Houkes, I., Winants, Y., & Twellaar, M. (2008). Specific determinants of burnout among male and female general practitioners: A cross-lagged panel analysis. Journal of Occupational and Organizational Psychology, 81, 249– 276.

Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit in covariance structure analysis. Structural Equation Modeling, 6, 1–55.

Judge, T. A., Higgins, C. A., Thoresen, C. J., & Barrick, M. R. (1999). The Big Five personality traits, general mental ability, and career success across the life span. Personnel Psychology, 52, 621–652.

Kern, M. L., & Friedman, H. S. (2008). Do conscientious individuals live longer? A quantitative review. Health Psychology, 27, 505–512.

Kline, R. B. (2004). Principles and practice of structural equation modeling (2nd ed.). New York: Guilford Press.

Kristensen, T. S., Borritz, M., Villadsen, E., & Christensen, K. B. (2005). The Copenhagen Burnout Inventory: A new tool for the assessment of burnout. Work & Stress, 19, 192–208.

Kroenke, K., Strine, T. W., Spitzer, R. L., Williams, J. B. W., Berry, J. T., & Mokdad, A. H. (2009). The PHQ-8 as a measure of current depression in the general population. Journal of Affective Disorders, 114(1–3), 163–173.

Lahey, B. B. (2010). Public health significance of neuroticism. American Psycholo- gist, 64, 241–256.

Linzer, M., Visser, M. R. M., Oort, F. J., Smets, E. M. A., McMurray, J. E., & de Haes, H. C. J. M. (2001). Predicting and preventing physician burnout: Results from the United States and the Netherlands. American Journal of Medicine, 111, 170–175.

Maslach, C., & Jackson, S. E. (1986). The Maslach burnout inventory. Palo Alto, CA: Consulting Psychologist Press.

McCrae, R. R., & Costa, P. T., Jr. (2003). Personality in adulthood: A five-factor theory perspective (2nd ed.). New York: Guilford Press.

McCrae, R. R., & John, O. D. (1992). An introduction to the five factor model and its applications. Journal of Personality, 60, 175–215.

McDonald, R. P., & Ho, M. R. (2002). Principles and practices in reporting structural equation analyses. Psychological Methods, 7, 64–82.

Melamed, S., Shirom, A., Toker, S., Berliner, S., & Shapira, I. (2006). Burnout and risk of cardiovascular disease: Evidence, possible causal paths, and prom- ising research directions. Psychological Bulletin, 132, 327–353.

Miller, T. Q., Smith, T. W., Turner, C. W., Guijarro, M. L., & Hallet, A. J. (1996). A meta-analytic review of research on hostility and physical health. Psycho- logical Bulletin, 119, 322–348.

Mills, L. B., & Huebner, E. S. (1998). A prospective study of personality charac- teristics, occupational stressors, and burnout among school psychology practitioners. Journal of School Psychology, 36, 103–120.

Miner, M. H. (2007). Burnout in the first year of ministry: Personality and belief style as important predictors. Mental Health, Religion and Culture, 10, 17–29.

The Big Five and Facets of Burnout 425

Piedmont, R. L. (1993). A longitudinal analysis of burnout in the health-care setting—The role of personal dispositions. Journal of Personality Assessment, 61, 457–473.

Purvanova, R. K., & Muros, J. P. (2010). Gender differences in burnout: A meta-analysis. Journal of Vocational Behavior, 77, 168–185.

Reise, S. P., Waller, N. G., & Comrey, A. L. (2000). Factor analysis and scale revision. Psychological Assessment, 12, 287–297.

Richardson, D., Wing, S., Steenland, K., & McKelvey, W. (2004). Time-related aspects of the healthy worker survivor effect. Annals of Epidemiology, 14, 633–639.

Rogers, W. M., & Schmitt, N. (2004). Parameter recovery and model fit using multidimensional composites: A comparison of four empirical parceling algo- rithms. Multivariate Behavioral Research, 39, 379–412.

Saucier, G. (1994). Mini-Markers: A brief version of Goldberg’s unipolar Big-Five Markers. Journal of Personality Assessment, 63, 506–516.

Schmitt, D. P., Realo, A., Voracek, M., & Allik, J. (2008). Sex differences in Big Five personality traits across 55 cultures. Journal of Personality and Social Psychology, 94, 168–182.

Schaufeli, W. B., & Enzmann, D. (1998). The burnout companion to study and practice: A critical analysis. Washington, DC: Taylor & Francis.

Shirom, A. (2003). Job-related burnout. In J. C. Quick & L. E. Tetrick (Eds.), Handbook of occupational health psychology (pp. 245–265). Washington, DC: American Psychological Association.

Shirom, A., & Melamed, S. (2006). A comparison of the construct validity of two burnout measures in two groups of professionals. International Journal of Stress Management, 13, 176–200.

Shirom, A., Melamed, S., Rogowski, O., Shapira, I., & Berliner, S. (2009). Work- load, control, and social support effects on blood lipids: A longitudinal study among apparently healthy employed adults. Journal of Occupational Health Psychology.

Shirom, A., Nirel, N., & Vinokur, A. D. (2006). Overload, autonomy, and burnout as predictors of physicians’ quality of care. Journal of Occupational Health Psychology, 11, 328–342.

Suls, J., & Bunde, J. (2005). Anger, anxiety, and depression as risk factors for cardiovascular disease: The problems and implications of overlapping affective dispositions. Psychological Bulletin, 131, 260–300.

Swider, B. W., & Zimmerman, R. D. (2010). Born to burnout: A meta-analytic path model of personality, job burnout, and work outcomes. Journal of Voca- tional Behavior, 76, 487–506.

Taylor, S. E. (2006). Tend and befriend bio behavioral bases of affiliation under stress. Current Directions in Psychological Science, 15, 273–277.

Thoresen, C. J., Kaplan, S. A., Barsky, A. P., Warren, C. R., & Chermont, D. K. (2003). The affective underpinnings of job perceptions and attitudes: A meta- analytic review and integration. Psychological Bulletin, 129, 914–945.

Vinokur, A. D., Pierce, P. F., & Lewandowski-Romps, L. (2009). Disentangling the relationships between job burnout and perceived health in a military sample. Stress & Health, 25, 355–363.

Armon, Shirom, & Melamed426

Viswesvaran, C., & Ones, D. S. (2000). Measurement error in “Big Five Factors” personality assessment: Reliability generalization across studies and measures. Educational and Psychological Measurement, 60, 224–235.

Watson, D., & Clark, L. A. (1984). Negative affectivity: The disposition to experience negative emotional states. Psychological Bulletin, 96, 465–490.

Wayne, J. H., Musisca, N. & Fleeson, W. (2004). Considering the role of person- ality in the work-family experience: Relationships of the Big Five to work- family conflict and facilitation. Journal of Vocational Behavior, 64, 108–130.

Wong, S. S., Oei, T. S. P., Ang, R. P., Lee, B. O., Ng, A. K., & Leng, V. (2007). Personality, meta-mood experience, life satisfaction, and anxiety in Australian versus Singaporean students. Current Psychology, 26, 109–120.

Wood, W., & Eagly, A. H. (2002). A cross-cultural analysis of the behavior of men and women: Implications for the origins of sex differences. Psychological Bul- letin, 128, 699–727.

Yuan, K. H. (2005). Fit indices versus test statistics. Multivariate Behavioral Research, 40, 115–148.

The Big Five and Facets of Burnout 427

This document is a scanned copy of a printed document. No warranty is given about the accuracy of the copy.

Users should refer to the original published version of the material.