918297

research-article20202020
SGOXXX10.1177/2158244020918297SAGE OpenPark and Shin

Original Research

SAGE Open

A Meta-Analysis of Special Education

April-June 2020: 1­–18
© The Author(s) 2020
DOI: 10.1177/2158244020918297
https://doi.org/10.1177/2158244020918297

Teachers’ Burnout journals.sagepub.com/home/sgo

Eun-Young Park1 and Mikyung Shin2

Abstract
This meta-analysis verifies associations between three dimensions of special education teachers’ (SET) burnout (emotional
exhaustion, depersonalization, and personal accomplishment) and student-, teacher-, and school-related variables. Altogether,
28 peer-reviewed English articles and 13 dissertations (total sample of teachers = 6,623) published between 1983 and
December 2018 were analyzed. The degree of correlation effect sizes between special education teachers’ burnout and
its related variables was extensive. Results revealed distinct relations by each burnout dimension: emotional exhaustion,
depersonalization, and personal accomplishment and student-, teacher-, and school-related variables. Student age (Fisher’s Z =
.316) was significantly associated with SET depersonalization. Self-efficacy (Fisher’s Z = −0.390, emotional exhaustion; −0.321,
depersonalization; 0.633, personal accomplishment), stress (0.366, emotional exhaustion; 0.340, depersonalization; −0.110
personal accomplishment), and support from school personnel (−0.119, emotional exhaustion; −0.140, depersonalization;
0.172, personal accomplishment) were also significantly related to each burnout dimension. Support programs to relieve SET
burnout must consider these variables.

Keywords
burnout, depersonalization, emotional exhaustion, personal accomplishment, special education teachers

The problem of teacher’s burnout has been the steady subject as a psychological syndrome involving three dimensions:
of research. Special education teachers have been reported to emotional exhaustion, depersonalization, and a diminished
be particularly vulnerable to burnout (Emery & Vandenberg, sense of personal accomplishment (Maslach, 1982). Those
2010; Y. L. Lee et al., 2011). Burnout was identified as a experiencing burnout have frustration, cynicism, and other
major contributor to attrition in previous research (Wong negative emotions toward themselves and their work, which
et al., 2017). Even in America, this problem has produced a causes emotional depletion, impersonality, and an overall
sustained reduction in special education teachers (Emery & diminishment of their sense of achievement (J. H. J. Lee &
Vandenberg, 2010; Fore et al., 2002; Leko & Smith, 2010). Ok, 2012). Teacher burnout is a serious concern, because it
The U.S. Department of Education, Office of Postsecondary threatens psychological well-being. Unlike stress, it is diffi-
Education (2017) reported that 46 states were drastically cult for those afflicted to manage burnout in the work context
short of special education teachers. Burnout is closely related where it manifests. It is therefore important to identify pre-
to the physical and psychological health and work perfor- ventive measures and promotional variables.
mance of teachers, as well as attrition, and it may affect teach-
ing and the way teachers interact with students in educational
Key Variables Associated With Special
settings (Hall, 2013). Teachers’ burnout gradually develops
over an extended period, becoming chronic and worse (Fernet Education Teachers’ Burnout
et al., 2012). Teachers’ burnout is reported to be related to Various student-, teacher-, and school-related variables
student motivation; teachers’ burnout can undermine student related to the onset of special education teachers’ burnout
motivation (Shen et al., 2015). Therefore, research on special exist. These variables closely align with the realities of
education teachers’ burnout has become a major concern and
is extensively studied (Brunsting et al., 2014; Leung & Lee, 1
Jeonju University, College of Education, Jeonju University, Jeonju, South Korea
2006); however, effective intervention strategies for burnout 2
West Texas A&M University, Canyon, USA
do not yet exist (Hastings & Brown, 2002).
Corresponding Author:
Burnout has a long-term relationship with individuals
Mikyung Shin, Department of Education, West Texas A&M University,
who are constantly exposed to fatigue, hostility, discourage- 2501 4th Ave, Canyon, TX 79016, USA.
ment, maladjustment, discomfort, and restraint. It is defined Email: [email protected]

Creative Commons CC BY: This article is distributed under the terms of the Creative Commons Attribution 4.0 License
(https://creativecommons.org/licenses/by/4.0/) which permits any use, reproduction and distribution of
the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages
(https://us.sagepub.com/en-us/nam/open-access-at-sage).
2 SAGE Open

special education teachers’ daily work experiences; therefore, capacities (Boujut et al., 2016) supported special education
it is critical to provide them with the means to alleviate burn- teachers’ experiences in schools, reducing teacher burnout.
out before it produces negative outcomes (Brunsting et al., However, teacher stress was positively associated with
2014). teacher burnout variables of emotional exhaustion and deper-
sonalization, yet negatively related to personal accomplish-
ment (Ruble & McGrew, 2013).
Student-Related Variables
Student-related variables include age, disability type, behav- School-Related Variables
ior problems, grade, student numbers, socioeconomic status
(SES), and setting. Special education teachers demonstrated School-related variables include work hindrances, emotional
different responses based on students’ disability type, behav- experiences, number of support personnel, available
ior problems, and the severity of students’ challenges. Banks resources, and support from school personnel. Recently, with
and Necco (1990) discovered that teachers teaching students the emphasis on teacher accountability, special education
with emotional disorders demonstrated higher burnout than teachers’ increased workloads (e.g., managing classrooms,
those teaching students with intellectual disabilities. Nichols working as case managers for Individualized Education
and Sosnowsky (2002) reported that the proportion of stu- Program meetings, student progress monitoring) are posi-
dents with emotional disorders who exhibited challenging tively associated with their burnout levels (Embich, 2001;
behaviors was associated with higher intensity of teacher Ruble et al., 2011). Studies also support that work hindrances
burnout in self-contained classrooms. Student measures of (e.g., role conflict, role ambiguity) are positively associated
severity (e.g., intelligence quotient [IQ], language, adaptive with special education teachers’ burnout levels (Adera &
behavior) negatively affected teacher satisfaction and Bullock, 2010; Embich, 2001). However, emotional experi-
required greater support in schools (Hastings & Brown, ences in schools such as feeling secure, autonomous (Carlson
2002; Ruble & McGrew, 2013). Moreover, students’ SES & Thompson, 1995), and supported by other teachers and
and service delivery models could affect teachers’ teaching administrators (Bataineh & Alsagheer, 2012; Boujut et al.,
experience and burnout levels. Kvande et al. (2018) identi- 2016) were negatively associated with special education
fied the mediated effects of students’ SES status on special teachers’ burnout. Zarafshan et al. (2013) also revealed that
education services. Specifically, students with low SES were positive relations with coworkers and opportunities for pro-
more likely to receive special education services, and teach- motion at schools decreased the level of emotional exhaus-
ers reported increased feelings of helplessness when these tion and depersonalization. Unfortunately, however, Kaff
students demonstrated low academic achievement. (2004) noted that numerous special education teachers did
not feel they received sufficient support from their principals
and resources in schools to manage their academic
Teacher-Related Variables responsibilities.
Teacher-related variables include teachers’ age, gender, teach-
ing experience, education level, self-efficacy, satisfaction, Exploring Special Education
coping, and stress. Gong et al. (2013) noted that special edu- Teachers’ Burnout in Previous
cation teachers aged between 26 and 30 years displayed
greater emotional exhaustion than those above 30 years.
Reviews and Syntheses
Brewer and McMahan (2004) also reported that women were Literature has focused on issues such as stress and burnout
more likely to experience job stress and burnout. Gong et al. among special education teachers for more than three decades
(2013) stated that female teachers experienced higher per- (Banks & Necco, 1990; Embich, 2001; R. H. Zabel & Zabel,
sonal accomplishment, yet male teachers were more prone to 2002). Given, especially, the increasing demands on special
depersonalization. Years of teaching experience were also education teachers through education quality and educa-
associated with teacher burnout. Williams and Dikes (2015) tional accountability policies (e.g., the Every Student
found that the number of years of teaching was positively cor- Succeeds Act of 2015 in America and the Special Education
related with burnout; thus, the greater the number of years at Act for Individuals with Disabilities and Others of 2008 in
school, the higher the stress and burnout level. Carlson and South Korea), they are prone to increased burnout, a high
Thompson (1995), however, indicated that teacher experience risk factor associated with teacher attrition (Hagaman &
did not significantly contribute to degree of teacher burnout. Casey, 2018; Ryan et al., 2017). Special education teachers’
Teachers’ education levels were also connected to burnout burnout is associated with their personal and classroom
issues. As Embich (2001), in particular, indicated, younger, issues, as well as their increasing educational accountability.
less experienced, and less educated team-teaching teachers Despite these ongoing concerns, in previous years, only two
experienced higher burnout levels than other teachers. Studies studies (Brunsting et al., 2014; Wisniewski & Gargiulo,
also reported that teachers’ self-efficacy (Sariçam & Sakiz, 1997) conducted a synthesis and a review, respectively,
2014), job satisfaction (Platsidou, 2010), and coping investigating the associations between special education
Park and Shin 3

teachers’ burnout and related variables. Other two recent Ruble & McGrew, 2013). These studies commonly reported
studies (Aloe et al., 2014; Iancu et al., 2018) targeted school high special education teacher burnout levels compared with
teachers in general and examined the intervention programs that of general education teachers, and related factors
that aim to reduce teacher burnout levels. included age, gender, number of teaching years, degree of
Wisniewski and Gargiulo (1997) conducted a review of self-efficacy, stress level, and type of disabilities. Therefore,
studies of occupational stress and burnout among special to generalize the findings across nations and update the syn-
education teachers. They reviewed mediating variables that thesis to cover the global context, we targeted studies pub-
affect teacher stress. For example, they stated that role con- lished in English and extended our research scope to include
flict and role ambiguity were the primary sources of teacher countries beyond the United States. Furthermore, Brunsting
stress. Specifically, special education teachers’ stress et al. (2014) descriptively reviewed issues related to special
increased when organizations provided information about education teachers’ burnout, while this study utilized a meta-
their roles and responsibilities that conflicted with their jobs analytic method to quantitatively examine variables associ-
and when they had insufficient information to perform their ated with special education teachers’ burnout. Meta-analysis
teaching responsibilities. Furthermore, Wisniewski and serves as a statistical method for synthesizing the findings of
Gargiulo also indicated that professional interactions with primary studies; it possesses a more advanced statistical
other school personnel (e.g., teachers, administrators, par- capacity to detect effects than primary studies. Through
ents) and instructional assignments (e.g., the highly stressful meta-analysis, we can compare effect sizes across studies
condition of classrooms of students with emotional and that differ in experimental rigor and other methodological
behavioral disorders) influenced special education teachers’ factors (Lipsey, 2003). Thus, we conducted a meta-analysis
burnout. to reach an overall conclusion regarding special education
Brunsting et al. (2014) conducted a synthesis of special teachers’ burnout associated with teacher-, student-, and
education teacher burnout from 1979 to 2013. They reviewed school-related variables. It provides the magnitude and direc-
23 studies, altogether, which measured teacher burnout in tion of effect of related variables on special education teach-
terms of their emotional exhaustion, depersonalization, and ers’ burnout. The research questions are as follows: (a) Is
lack of personal accomplishment. In addition, they analyzed there a relation between burnout score and student-related
teacher burnout at individual, classroom, school, district, and variables? (b) Is there a relation between burnout score and
state levels. Findings revealed that stress management and teacher-related variables? (c) Is there a relation between
emotional regulation affected a decrease in burnout. Students’ burnout score and school-related variables?
challenging behavior (e.g., percentage of students with emo-
tional disorders) was identified as a variable predominantly
affecting teachers’ degree of burnout. Student age was Method
another risk factor demonstrating that special education
teachers in secondary schools were exposed to higher levels
Literature Search
of burnout. Such school variables as role conflict, role ambi- A comprehensive search of the literature was performed
guity, and administrative support were strong factors in through a three-step process (Cooper, 1998). First, an online
teacher burnout. database search of EBSCOhost (n = 912), PsycINFO (n =
Although Aloe et al. (2014) also quantitatively analyzed 989), Education Source (n = 72), and Education Resources
teacher burnout, they only included one study (Sedgwick, Information Center (ERIC) (n = 98) was conducted to locate
1998) that combined general and special education teachers, a total of 2,071 studies.
thus lacking disaggregated findings for special education
teachers’ issues. More recently, Iancu et al. (2018) conducted Online search terms and their combinations. In Field 1, terms
a meta-analysis on the effects of interventions that aim to related to disabilities such as mental retardation, intellectual
reduce teacher burnout in general. Only two out of 23 studies disability, autism, emotional and/or behavioral disorder,
focused on special education teacher as their participants emotional disturbance, health impairment, or disorder were
(Breeman et al., 2016; Cooley & Yovanoff, 1996). Therefore, used. Terms related to special education teachers were input-
exclusive focus on special education teacher burnout is ted in Field 2. Terms related to burnout such as emotional
required. Furthermore, as Brunsting et al. (2014) were the exhaustion, depersonalization, personal accomplishment,
last to review special education teacher burnout, targeted and burnout were inserted in Field 3. A total 2,071 studies
studies were conducted only in the United States. However, were identified in the initial search. Of these, 114 duplicated
the issue of special education teacher burnout is gaining studies were excluded.
research attention in various countries around the world, Second, the two authors screened the titles and abstracts of
such as the United Kingdom (e.g., Hastings & Brown, 2002), the 1,957 studies across four inclusion criteria: (a) Participants
the Netherlands (e.g., Hopman et al., 2018), Greece (e.g., should be special education teachers, (b) studies should be
Platsidou, 2010), Iran (e.g., Zarafshan et al., 2013), Turkey published in English peer-reviewed journals and disserta-
(e.g., Sariçam & Sakiz, 2014), and the United States (e.g., tions, (c) studies should measure and report disaggregated
4 SAGE Open

56 dissertations searched through Total of 2,071 articles searched using

ProQuest and PsycINFO keywords
912 articles identified through EBSCOhost
114 articles duplicated
989 articles identified through PsycINFO 1,712 articles excluded by checking the
43 dissertations excluded by reviewing the title
72 articles identified through Education
whole paper Source
- Not related to burnout (n = 30)
- Did not meet the research design criteria 98 articles identified through ERIC
(n = 3)
- Did not provide sufficient information for
data analysis (n = 7)
- Duplicated with included article (n = 3) 245 articles screened for abstract reviews 133 articles excluded by reviewing the
abstract

89 articles excluded by reviewing the

whole paper
112 articles selected for full-text reviews - Did not target special education
teachers (n = 28)
- Not related to burnout (n = 15)
- Did not meet the research design
criteria (n = 43)
- Did not provide sufficient information
13 dissertations selected 23 articles selected for data analysis (n = 1)
- Did not report subcategory measures
(n = 2)

5 articles added through an inverse

search

41 studies included

Figure 1. A flowchart describing the selection of the 41 studies after multiple phases of the screening process.
Note. ERIC = Education Resources Information Center.

data on special education teachers’ burnout (i.e., emotional step, a search for dissertations was completed. In ProQuest
exhaustion, depersonalization, and personal accomplishment) and PsycINFO, 56 dissertations were initially searched; 13
in schools, and (d) studies should provide the correlation data. dissertations meeting the four inclusion criteria (see above)
According to recommendations of meta-analysis article were finally included. In these research procedures, 28 arti-
searching procedures, studies which employed similar meth- cles and 13 dissertations were selected for this
odology were included in this meta-analysis (Gopalakrishnan meta-analysis.
& Ganeshkumar, 2013). We included studies with correlation, Out of 41 included studies in the current meta-analysis, 11
regression, and longitudinal research designs and excluded studies (26.8%) overlapped with Brunsting et al. (2014);
intervention studies. After assessing the titles, 1,712 studies although Brunsting et al. (2014) included 23 studies in their
were excluded. Of the 245 screened studies, 133 were synthesis, 12 of the 23 studies (e.g., Frank & McKenzie,
excluded after reviewing the abstracts. 1993; M. K. Zabel et al., 1984) did not meet our criteria.
Third, the full-text copies of 112 studies were retrieved Figure 1 describes the screening process.
and two authors reviewed them to meet the inclusion criteria;
89 noneligible studies were excluded for the following rea-
sons: (a) 28 (31.5%) did not target special education teach-
Coding Procedure and Inter-Rater Reliability
ers, (b) 15 (16.9%) were not related to burnout, (c) 43 Using Microsoft Excel, we coded the following variables:
(48.3%) did not meet the research design criteria, (d) one design, number of special education teachers, age, gender,
(1.1%) did not provide sufficient information for data analy- years of teaching, burnout measures, related variables, and
sis, and (e) two studies (2.2%) did not report subcategory correlation coefficients between their burnout and related
measures. Twenty-three articles were selected through this variables. Regarding burnout measures, we focused on emo-
review. In December 2018, we updated the included articles tional exhaustion, depersonalization, and personal accom-
through database and manual searches. Statistically nonsig- plishment following the three dimensions of the Maslach
nificant results are less likely to be published and included in Burnout Inventory (MBI; Maslach & Jackson, 1996): deper-
meta-analyses. To prevent this file-drawer effect, two authors sonalization, that is, development of negative and cynical
conducted an inverse search on the reference pages of the 23 attitudes toward service users; emotional exhaustion, that is,
included articles and found five additional articles. As a last staff feeling that they have little to offer at a psychological
Park and Shin 5

level to their work; and personal accomplishment, that is, certain cases, studies reported more than one related mea-
staff evaluating themselves and their accomplishments posi- sure; here, we coded more than one category whenever
tively. Given the meaning discrepancy underlying these three applicable and displayed orthogonal contrasts to capture the
dimensions, the positively higher scales in depersonalization relation between two variables (Gersten et al., 2009). Fol-
and emotional exhaustion denote negative attitudes, while lowing Borenstein et al.’s (2009) recommendations, we con-
the positively higher scales in personal accomplishment verted the correlation coefficients to Fisher’s Z scale using
denote positive attitudes. We reversely coded eight studies the following transformation formula: Fisher’s Z = 0.5 *
(Chatlos, 2016; Crane & Iwanicki, 1986; Dickerson, 2017; Log (1 + Corr)/(1 − Corr). Subsequently, we performed data
Embich, 2001; Freed, 1994; Goodall, 1986; Hastings & analysis using the transformed Fisher’s Z values.
Brown, 2002; LaMonica, 1983) that analyzed personal
accomplishment as reduced personal accomplishment in Statistical analyses. For statistical analyses such as publica-
their original survey. When coding the data from the longitu- tion bias, overall effect sizes, and Q statistics, data were ana-
dinal research design, we independently considered each lyzed employing Comprehensive Meta-Analysis (CMA)
correlation effect size at each different time point. version 3.0 (Biostat, 2017) and aggregated effect sizes across
Furthermore, for burnout-related variables, we referred to studies. First, before applying the meta-analytic approach,
Bronfenbrenner’s (1977) ecological model and organized we examined the publication bias using Duval and Tweedie’s
variables as student-, teacher-, and school-related (see Trim and Fill (Egger et al., 1997). The possibility of publica-
Brunsting et al., 2014). For example, we included student tion bias can indicate an overestimation of meta-analysis
age, disability, and setting as student-related variables; results (Rosenberg, 2005; Rosenthal, 1979).
teacher age, gender, and years of teaching as teacher-related Second, we disaggregated effect sizes by emotional
variables; and resources, support from school personnel, and exhaustion, depersonalization, and personal accomplish-
the number of support personnel as school-related variables. ment. Evaluating heterogeneity of variance using the Q sta-
Variables that could not be categorized as specific variables tistic for each effect size, we applied a random-effects
were coded as others. Others in teacher-related variables model for the current study when p value was less than .05,
included instructional strategies, teaching quality, teacher indicating significant between-studies variance; subse-
race, and general health. Others in student-related variables quently, we applied a fixed-effects model when p value was
included language, IQ, and student engagement. equal or greater than .05, indicating nonsignificant between-
We received training before initiating the coding. Four studies variance (Borenstein et al., 2009; Lipsey & Wilson,
(14.8%) articles were coded independently. We initially 2001).
achieved an inter-rater reliability of 97.3%, using a percent
agreement method: “agreements/(agreements + disagree-
Results
ments) × 100% = P%” (Araujo & Born, 1985, p. 208). The
disagreement was resolved through discussion of the defini- Table 1 displays the summary of study features of 41 included
tion of variables. In the second coding, there was 100% studies (28 articles, 13 dissertations) that examined special
agreement. After the training, two coders independently education teachers’ burnout issues related to student-,
coded all articles and cross-checked the completed coding. teacher-, and school-related variables. The publication years
Between the two coders, the average inter-rater reliability of 41 studies ranged from 1983 to 2018: there were seven
was 98% for emotional exhaustion (range = 97%–100% for between 1983 and 1989, 12 between 1990 and 2009, and 22
each variable), 99% for depersonalization (range = 99%– between 2010 and 2018. Of the 41 studies, 29 (70.7%) were
100% for each variable), and 99% for personal accomplish- published in the United States. A total of 6,623 teachers were
ment (range = 99%–100% for each variable). Two coders included in the meta-analysis. Sample sizes ranged from 21
reviewed the dissimilar coding items by rechecking the full to 443 special education teachers. Out of 41 studies, 32
text together to ensure 100% agreement on all the coding reported information regarding the samples’ gender: out of
variables. 4,070 reported samples, 881 (21.6%) were male and 3,189
(78.4%) were female special education teachers. In 24 stud-
ies reporting samples’ age, participants’ ages ranged from 21
Data Analysis to 71 years. In 30 studies reporting samples’ teaching experi-
Calculation of effect sizes. We employed Pearson’s r as the ence, participants possessed an extensive range of years of
effect size index by examining the correlation between spe- teaching from less than 1 year to at least 30 years.
cial education teachers’ burnout and related variables. When Figures 2 to 4 showed the forest plots of weighted effect
studies reported regression data results, we extracted r with sizes on each dimension of burnout. The overall effects as
the following conditions: The regression should be linear, well as specific confidence intervals of correlation effects of
indicating one predictor variable (df = 1); or we could extract each variable were described. Each figure demonstrates dif-
R (or R2) from the regression table, and the correlation direc- ferent ranges of effect sizes for each student-, teacher-, and
tion should be the same as the nonstandardized B value. In school-related variables.
6
Table 1. Summary of Study Characteristics.
Author (year) N Age years (n) Gender Years of teaching (n) Related variables Country Type

Ambrosini (2013) 94 29–42 (25), 43–55 (46), 56–69 (23) 19 M, 75 F 1–8 (50), 9–12 (24), 13–33 (20) T: Age, EL, stress, TE United States D
Banks and Necco 181 NR NR 9.36 T: Age, TE United States A
(1990)
Bataineh (2009) 83 24–48 43 M, 40 F 3–15 Sc: SSP Jordan A
Bataineh and Alsagheer 300 22–45 50 M, 250 F 4–17 Sc: SSP UAE A
(2012)
Biglan et al. (2013) 42 NR NR NR T: Coping United States A
Bloom (1992) 33 37.6 2 M, 31 F 12.7 T: Age, TE; Sc: EE, NSP, WH United States D
Boujut et al. (2016) 103 41.58 15 M, 88 F 14.84 T: Coping; Sc: EE, R, satisfaction, stress, SSP France A
Caputo and Langher 276 30–62 52 M, 224 F 17.6 Sc: SSP Italy A
(2015)
Chatlos (2016) 31 NR 5 M, 23 F, 3 NR 0–10 (12), 11–20 (14), >20 (5) T: Self-efficacy, O United States D
Coman et al. (2013) 53 NR 1 M, 52 F 9.83 St: Disability, SN; T: E; Sc: SSP United States A
Crane and Iwanicki 443 NR NR NR T: O; Sc: WH United States A
(1986)
Dickerson (2017) 46 NR 8 M, 37 F, 1 NR NR Sc: Satisfaction United States D
Eichinger et al. (1991) 78 NR 78 F NR T: Coping United States A
Embich (2001) 300 40.19 NR 11.18 T: Age, EL, TE; Sc: SSP, WH United States A
Fimian and Blanton 415 NR NR NR St: SN, O; T: Age, gender, stress, O; Sc: WH United States A
(1986)
Freed (1994) 342 20–39 (219), 40–49 (88), >50 (35) 63 M, 279 F 11.04 St: Disability, grade, setting, SN; T: Age, United States D
gender, TE; Sc: Stress, SSP, WH
Garwood et al. (2018) 64 41.16 4 M, 60 F 14.94 T: Self-efficacy; Sc: WH United States A
Goetzinger (2006) 224 NR 12 M, 211 F, 1 NR 13.4–19.9 St: SN; T: TE; Sc: WH United States D
Gong et al. (2013) 256 30 (108), 31–50 (102),51 (46) 72 M, 184 F 0–9 (174), 10–29 (67), 30 (15) T: Age, gender, TE, O; Sc: SSP United States A
Goodall (1986) 212 20–69 66 M, 146 F 1–45 T: Age, EL, gender, TE United States D
Hassan and Mohamed 81 32.38 81 F 1–5 (19), 6–10 (33), 10 (29) T: Stress, TE; Sc: SSP Oman A
(2015)
Hastings and Brown 55 NR 14 M, 41 F 7.34 St: Behavior problem; T: Coping United Kingdom A
(2002)
Hill (2011) 71 34.69 20 M, 51 F 6.05 T: Age, EL, TE; Sc: SSP United States D
Hopman et al. (2018) 141 38.7 55 M, 86 F 5.1 St: Behavior problem, O; T: O; Sc: SSP, WH NL A
LaMonica (1983) 99 NR 39 M, 60 F 12.86 T: TE; Sc: EE, stress United States D
Martin (2010) 105 21–38 (48), 39–55 (39), 56 (18) 12 M, 93 F 1–15 (61), 16–25 (16), 25 (28) T: Self-efficacy, TE United States D
St: Disability, grade, SN; Sc: Stress, SSP, WH
McDow (1993) 426 NR NR NR St: SN United States D
McIntyre (1984) 399 NR NR NR T: Coping United States A
Moss (2014) 72 NR 1 M, 71 F 10.79 Sc: WH United States D
Platsidou (2010) 123 39.6 47 M, 79 F 6.2 T: Age, satisfaction, TE; Sc: EE Greece A
Platsidou and Agaliotis 127 39.6 49 M, 78 F 4 (68), 5–9 (32), 10–23 (27) T: Satisfaction, stress Greece A
(2008)
Ruble and McGrew 47 NR NR NR St: Age, behavior problem, O; T: TE, stress, United States A
(2013) O; Sc: SSP
Ruble et al. (2011) 35 NR 2 M, 33 F 10.6 T: TE; Sc: SSP, WH United States A
Sari (2005) 295 34.67 141 M, 154 F NR T: Age, coping, gender, satisfaction Turkey A
Sariçam and Sakiz 118 33.9 58 M, 60 F NR T: Self-efficacy Turkey A
(2014)
Shea (1984) 215 20–69 215 F 1–35 St: Disability, SN; Sc: WH United States D
(continued)
 Table 1. (continued)

Author (year) N Age years (n) Gender Years of teaching (n) Related variables Country Type

Shyman (2010) 100 21–71 11 M, 89 F NR (3), 1 (20), 1–2 (20), 2–3 (25), T: Self-efficacy; Sc: SSP, WH United States A
3 (32)
Weber and Toffler 68 34.5 17 M, 51 F 4.36 St: Age, SN, SES, O; T: Age, EL, TE; Sc: R, SSPUnited States A
(1989)
Wong et al. (2017) 79 46 3 M, 76 F 11.02 T: Stress, other; St: O; Sc: WH United States A
R. H. Zabel and Zabel 298 NR NR NR Sc: SSP United States A
(2002)
Zarafshan et al. (2013) 93 36.81–40.29 93 F 8.53–19.32 St: Disability; T: Satisfaction, O; Sc: EE, SSP, Iran A
WH

Note. M = male; F = female; T = teacher-related variable; EL = education level; TE = teaching experience; D = dissertation; NR = not reported; A = article; Sc = school-related variable; SSP =
support from school personnel; EE = emotional experience; NSP = number of support professional; WH = work hindrance; R = resource; O = other; SN = student number; St = student-related
variable; SES = socioeconomic status.

7
8 SAGE Open

Figure 2. A forest plot of emotional exhaustion.

Note. CI = confidence interval; SES = socioeconomic status.

Publication Bias personal accomplishment (Fisher’s Z = 0.107, 95% CI =

[0.023, 0.191]).
As effect size calculation was performed by dividing the sub-
categories of emotional exhaustion, depersonalization, and
personal accomplishment, publication bias was also divided Correlation Effects of Teacher-Related Variables
into three. There was no evidence of publication bias in this
The effect of teacher-related variables on burnout is pre-
study in terms of emotional exhaustion and depersonaliza-
sented in Table 3. Certain variables had a small effect on
tion. There were zero studies in the Trim and Fill, and
emotional exhaustion among teacher-related variables. The
adjusted values were identical to the random effect size of
significant effect size was shown in other (Fisher’s Z =
observed values. There was also no evidence of personal
0.068, 95% CI = [0.025, 0.110]), satisfaction (Fisher’s Z =
accomplishment. The difference of effect sizes between
−0.223, 95% CI = [−0.264, −0.182]), self-efficacy (Fisher’s
observed value and an adjusted value was trivial, and only
Z = −.390, 95% CI = [−0.562, −0.218]), and stress (Fisher’s
three values were trimmed.
Z = 0.366, 95% CI = [0.270, 0.461]). Higher levels of satis-
faction and self-efficacy were related to significantly lower
emotional exhaustion, while a higher level of stress was
Correlation Effects of Student-Related Variables
related to significantly higher emotional exhaustion.
The effect size of student-related variables is displayed in Education level, satisfaction, self-efficacy, stress, and
Table 2. There were no variables related to significant levels other displayed significant effect size on depersonalization.
of emotional exhaustion. Regarding depersonalization, only Specific significant effect sizes were: −0.110 of education
student’s age was significant with a small-to-medium effect level (95% CI = [−0.183, −0.037]), −.263 of satisfaction
size, 0.316 (95% CI = [0.127, 0.465]). Higher age was (95% CI = [−0.368, −0.159]), −0.321 of self-efficacy (95%
related to a significantly higher level of depersonalization. CI = [−0.145, −0.227]), 0.240 of stress (95% CI = [0.135,
Other variables showed significant relations with level of 0.345]), and 0.207 of other (95% CI = [0.025, 0.388]).
Park and Shin 9

Figure 3. A forest plot of depersonalization.

Note. CI = confidence interval; SES = socioeconomic status.

Higher levels of education, satisfaction, and self-efficacy (95% CI = [0.184, 0.298]), respectively. A higher level of
were related to a significantly lower level of depersonaliza- support from school personnel was related to a significantly
tion, while a higher level of stress was related to a signifi- lower level of emotional exhaustion (Fisher’s Z = −0.119,
cantly lower level of depersonalization. 95% CI = [−0.172, −0.065]). Regarding depersonalization,
Regarding personal accomplishment, higher level of cop- higher levels of emotional experience and support from
ing, self-efficacy, and long duration of teaching experience school personnel were related to a significantly lower level
were related to significantly higher personal accomplish- of burnout. The effect sizes were −0.107 (95% CI = [−0.175,
ment. Moreover, a higher level of stress was related to a sig- −0.038]) and −0.140 (95% CI = [−0.157, −0.122]), respec-
nificantly lower level of personal accomplishment. Detailed tively. Higher levels of work hindrances were related to a
effect sizes were 0.338 in coping (95% CI = [0.229, 0.448]), significantly higher level of depersonalization (Fisher’s Z =
0.075 in teaching experience (95% CI = [0.044, 0.106]), 0.382, 95% CI = [0.353, 0.412]). Higher levels of emotional
0.633 in self-efficacy (95% CI = [0.474, 0.793]), and −0.110 experience and support from school personnel were related
in stress (95% CI = [−0.191, 0.030]). to a significantly higher level of personal accomplishment.
The effect sizes were 0.245 (95% CI = [0.126, 0.364]) and
0.172 (95% CI = [0.127, 0.217]), respectively.
Correlation Effects by School-Related Variables
The effect of school-related variables on burnout is presented
Discussion
in Table 4. The number of support personnel, support from
school personnel, and work hindrances were significantly This study aimed to verify the effect size of variables related
correlated with emotional exhaustion. A higher number of to special education teachers’ burnout through meta-analysis.
support personnel and level of work hindrances was related To achieve this, quantitative relations between burnout and
to significantly higher levels of emotional experience. The its associated variables reported in previous studies were
effect sizes were 0.036 (95% CI = [0.104, 0.507]) and 0.241 analyzed. Specifically, the relation of student-, teacher-, and
10 SAGE Open

Figure 4. A forest plot of personal accomplishment.

Note. CI = confidence interval; SES = socioeconomic status.

school-related variables with three dimensions of burnout the implications behind disability types affecting teacher
(emotional exhaustion, personal accomplishment, and deper- burnout indicates that the problem behaviors of students,
sonalization) were analyzed. which differ according to disability type, affect burnout.
Aloe et al. (2014) reported effects between misbehavior and
emotional exhaustion (d = 0.44), depersonalization (d =
Correlation Effects by Student-Related Variables 0.36), and personal accomplishment (d = −0.31). Special
There were a few variables such as age, disability type, and education teachers of students with emotional disorders
other that showed significant relations with level of burnout. experienced higher burnout than those of students with intel-
Although grade showed a significant effect size in relation to lectual disabilities (Banks & Necco, 1990). Nichols and
overall burnout, the size was trivial. Furthermore, student Sosnowsky (2002) also argued that the proportion of stu-
age was not significantly associated with the emotional dents with emotional disorders who exhibit challenging
exhaustion or personal accomplishment of special education behaviors was associated with a higher intensity of teacher
teacher; yet, their age showed a significant correlation with burnout. Special educators who teach disability types with a
depersonalization. The results of this study differed from the high frequency of problem behaviors are more likely to have
three studies previously reviewed in this article, where the difficulty controlling the classroom, and these difficulties
effects of student age on burnout were consistent (Brunsting may lead to burnout. This result can be interpreted based on
et al., 2014). Carlson and Thompson (1995) found that teach- previous research that teachers who have less classroom con-
ers of older students experienced higher levels of deperson- trol report more burnout (d = 0.62; Fitchett et al., 2018).
alization and lower levels of personal accomplishment, and
Weber and Toffler (1989) reported that student age was a
Correlation Effects by Teacher-Related Variables
predictor of the increase in the emotional exhaustion.
Student disability type had a significant effect size on The fact that teacher-related variables influence burnout is an
overall burnout and depersonalization. There were no sig- accepted result of previous studies (Brewer & McMahan,
nificant variables related to emotional exhaustion. Examining 2004; Coman et al., 2013; Gong et al., 2013; Sari, 2005).
Park and Shin 11

Table 2. Effect Size of Student-Related Variables.

95% CI

Burnout Related variables K Fisher’s Z SE Q statistics (p) Lower Upper

Emotional exhaustion Age 2 −0.048 0.096 .918 −0.236 0.140
Behavior problem 5 0.087 0.107 .009 −0.124 0.297
Disability 7 0.005 0.083 .000 −0.158 0.169
Grade 2 −0.021 0.048 .119 −0.114 0.073
Student number 21 0.015 0.047 .000 −0.078 0.107
Other 9 −0.106 0.080 .000 −0.264 0.051
SES 1 −0.030 0.124 1.000 −0.273 0.213
Setting 2 −0.010 0.038 .297 −0.085 0.065
Depersonalization Age 2 0.316 0.096 .094 0.127 0.465
Behavior problem 4 0.121 0.074 .161 −0.023 0.259
Disability 5 0.012 0.109 .000 −0.2201 0.225
Grade 1 0.110 0.054 1.000 −0.306 0.491
Student number 8 0.035 0.024 .285 −0.011 0.081
Other 6 −0.075 0.043 .274 −0.159 0.009
SES 1 −0.030 0.124 1.000 −0.273 0.213
Setting 2 0.055 0.038 .240 −0.020 0.130
Personal accomplishment Age 2 −0.104 0.096 .325 −0.292 0.083
Behavior problem 4 −0.080 0.074 .349 −0.224 0.065
Disability 5 0.016 0.083 .000 −0.147 0.178
Grade 2 −0.060 0.048 .125 −0.154 0.033
Student number 9 −0.004 0.023 .381 −0.049 0.041
Other 6 0.107 0.043 .643 0.023 0.191
SES 1 −0.192 0.124 1.000 −0.045 0.051
Setting 2 −0.045 0.038 .696 −0.120 0.030

Note. Bold font indicates statistically significant results, where 95% CI does not include zero. K = number of effect size; CI = confidence interval; SES =
socioeconomic status.

However, these studies do not consistently report the relation as an explained variable of teacher burnout. Teacher self-
of burnout with teacher-related variables. For example, efficacy correlated negatively with both emotional exhaus-
Fisher (2011) concluded that there was a significant differ- tion (.29) and depersonalization (.41) in structural equation
ence in burnout between novice and experienced teachers modeling (Skaalvik & Skaalvik, 2010). The correlation study
among a sample of 400, with novice teachers displaying high results reported a significant, medium, and negative correla-
levels of burnout. Carlson and Thompson (1995) reported tion between teacher self-efficacy and burnout levels of 163
that teaching experience did not significantly affect the secondary school teachers (Savaş et al., 2014). The negative
degree of burnout. These inconsistent results may be attrib- relation between self-efficacy and burnout was also con-
uted to the fact that they were analyzed without considering firmed in special education teachers. Ruble et al. (2011)
the dimensions of burnout. reported an inverse correlation of special education teachers’
In this meta-analysis, the correlation of each teacher- self-efficacy for students with autism spectrum disorder
related variable was different across emotional exhaustion, (ASD) with burnout. Increasing teachers’ self-efficacy (Savaş
personal accomplishment, and depersonalization. Special et al., 2014) is considered important for ensuring schools’
education teachers’ higher satisfaction and self-efficacy indi- effectiveness. Based on the results of this meta-analysis, an
cated significant relations to lower emotional exhaustion and increase in teachers’ self-efficacy is expected to have the
depersonalization, and the direction of stress with emotional maximum influence on enhancing personal accomplishment.
exhaustion was opposite. The level of satisfaction was not These results are consistent with another meta-analysis ana-
related to the level of personal accomplishment. Self-efficacy lyzing associations between job burnout and self-efficacy in
and stress among teacher-related variables demonstrated that various professions (Shoji et al., 2016) and focusing on indi-
the effect size was significant on all three dimensions. Given vidual protective factors (Li et al., 2013). The association
this discrepant correlation effect among the three dimensions between special education teachers’ self-efficacy and burnout
of burnout, each component should be interpreted alone implies that self-efficacy may be a protective factor against
(Maslach et al., 2001). burnout. These findings are in line with that of Jennett et al.
Subsequent discussion focusing on self-efficacy and stress (2003), in which teachers of students with ASD who had
displayed medium-to-large effect size. Self-efficacy is known higher professional self-efficacy with greater commitment to
12 SAGE Open

Table 3. Effect Size of Teacher-Related Variables.

95% CI

Burnout Related variables K Fisher’s Z SE Q statistics (p) Lower Upper

Emotional exhaustion Other 12 0.068 0.022 .000 0.025 0.110
Age 17 −0.034 0.055 .000 −0.142 0.074
Coping 14 0.060 0.063 .000 −0.063 0.183
Education level 9 −0.095 0.078 .009 −0.248 0.058
Gender 5 0.011 0.095 .000 −0.175 0.196
Teaching experience 33 −0.006 0.040 .023 −0.084 0.073
Satisfaction 24 –0.223 0.021 .169 –0.264 –0.182
Self-efficacy 7 –0.390 0.088 .000 –0.562 –0.218
Stress 21 0.366 0.049 .000 0.270 0.461
Depersonalization Other 8 0.207 0.093 .000 0.025 0.388
Age 15 −0.019 0.064 .000 −0.144 0.107
Coping 13 −0.036 0.071 .000 −0.176 0.104
Education level 9 –0.110 0.037 .171 –0.183 –0.037
Gender 5 0.015 0.105 .000 −0.191 0.221
Teaching experience 31 −0.006 0.016 .134 −0.037 0.026
Satisfaction 24 –0.263 0.053 .015 –0.368 –0.159
Self-efficacy 5 –0.321 0.048 .346 –0.415 –0.227
Stress 21 0.240 0.054 .000 0.135 0.345
Personal accomplishment Other 8 0.091 0.070 .000 −0.047 0.229
Age 16 0.007 0.048 .000 −0.087 0.101
Coping 13 0.338 0.056 .000 0.229 0.448
Education level 9 0.040 0.037 .826 −0.033 0.113
Gender 5 −0.055 0.078 .000 −0.209 0.099
Teaching experience 33 0.075 0.016 .257 0.044 0.106
Satisfaction 25 0.027 0.014 .000 −0.053 0.107
Self-efficacy 6 0.633 0.081 .003 0.474 0.793
Stress 21 –0.110 0.041 .000 –0.191 –0.030

Note. Bold font indicates statistically significant results, where 95% CI does not include zero. K = number of effect size; CI = confidence interval.

their theoretical orientation of teaching approach is related to particular, intervention programs such as stress management
lower burnout dimension. Intervention such as global theo- skills were helpful in reducing the emotional exhaustion of
retical education can be a very important buffer against low special education teacher (Cooley & Yovanoff, 1996).
teaching efficacy and high burnouts (Jennett et al., 2003).
Stress itself, a sudden failure of the mediation coping
Correlation Effects by School-Related Variables
mechanism, or an inefficiency of the mediation coping
mechanism over time can generate individual burnout In school-related variables, work hindrances, number of sup-
(Brackett et al., 2010; Guglielmi & Tatrow, 1998). port personnel, and support from school personnel displayed
Montgomery and Rupp (2005) conducted a meta-analysis of significant small-to-medium effect size on emotional exhaus-
the causes and effects of job-related stress on English- tion; various effect sizes were reported about work hindrances
speaking teachers in North America. The relation of external (k = 60) and support from school personnel (k = 66).
stress and psychological burnout experienced by teachers Significant meaningful variables were work hindrances. Work
indicated that active coping with stress and psychological hindrances had a significant small-to-medium effect size on
burnout were of a 0.26 and 0.39 effect size, respectively. emotional exhaustion and depersonalization. Work hindrances
Teachers’ stress and burnout are positively correlated; teach- increased emotional exhaustion, and depersonalization.
ers experiencing high stress levels are more likely to be A higher level of support from school personnel was
exposed to greater rates of burnout. Teachers classified at related to a decrease in burnout level. Although a small effect
risk for stress reported additional burnout symptoms size was shown, the significance of support from school per-
(d = 1.48; Fitchett et al., 2018). Stress is a risk factor for sonnel was shown in all three dimensions. Recent meta-anal-
teacher burnout, and stress reduction through protection such ysis by Iancu et al. (2018) highlighted the effects of social
as self-efficacy improvement is necessary because the pos- support such as peer collaboration groups on individual per-
sibility of exhaustion increases when stress recurs. In sonal accomplishment at schools. Support from school
Park and Shin 13

Table 4. Effect Size of School-Related Variables.

95% CI

Burnout Related variables K Fisher’s Z SE Q statistics (p) Lower Upper

Emotional exhaustion Emotional experience 12 −0.008 0.070 .000 −0.145 0.128
Number of support 2 0.306 0.103 .921 0.104 0.507
personnel
Resources 3 −0.014 0.133 .003 −0.274 0.246
Support from school 66 –0.119 0.027 .000 –0.172 –0.065
personnel
Work hindrances 60 0.241 0.029 .000 0.184 0.298
Depersonalization Emotional experience 11 –0.107 0.035 .000 –0.175 –0.038
Number of support personnel 2 0.112 0.103 .378 −0.090 0.313
Resources 1 0.234 0.124 1.000 −0.009 0.477
Support from school 58 –0.140 0.009 .000 –0.157 –0.122
personnel
Work hindrances 33 0.382 0.015 .000 0.353 0.412
Personal accomplishment Emotional experience 11 0.245 0.061 .000 0.126 0.364
Number of support personnel 2 −0.058 0.152 .005 −0.357 0.240
Resources 1 0.010 0.124 1.000 −0.233 0.253
Support from school 59 0.172 0.023 .000 0.127 0.217
personnel
Work hindrances 38 −0.005 0.031 .000 −0.065 0.055

Note. Bold font indicates statistically significant results, where 95% CI does not include zero. K = number of effect size; CI = confidence interval.

personnel such as the principal and administrators were also keep their jobs. Future research should conduct a follow-up
frequently reported to be related to teachers’ burnout meta-analysis, considering how the correlation effect sizes
(Embich, 2001). A noticeable result with respect to support from longitudinal studies could be mediated by time factors
was the number of support personnel. The effect direction of at multiple time points to further investigate whether certain
number of support personnel was opposite to the support of types of variables aggravate or improve relations with spe-
school personnel in emotional exhaustion. This suggests that cial education teacher burnout rates. Third, we analyzed cor-
it is necessary to pay attention to qualitative aspects rather relation effects by related variables (e.g., teaching
than quantitative aspects of support to reduce the level of experience, students’ disabilities, and support from school
special education teachers’ burnout. personnel) for each student-related, teacher-related, and
school-related variable. However, these correlation effects
Limitations and Recommendations for are possibly mediated by different categorical variables
(e.g., less than 5 years vs. greater than 5 years for teaching
Future Research experience; ASD, learning disabilities, emotional and
There are some limitations to this meta-analysis. First, behavioral disorders for students’ disabilities; and princi-
although we conducted a meta-analysis of special education pals, general education teachers, para-professionals for sup-
teachers’ burnout and other related variables regarding stu- port from school personnel, cultural contexts). Thus, future
dents, teachers, and schools, these results only indicated research should conduct moderator analysis based on these
how significantly these variables correlated with each other. categorical variables. Although the studies completed in
We still could not estimate the causational relation between various countries were included in this meta-analysis, we
burnout and related variables. Thus, in future research, the could not analyze the effect of cultural context including
effects of student-, teacher-, and school-related variables school contexts, school systems, regulations, and practices
should be investigated by analyzing the experimental about special education. Fourth, when examining the 41
designs of intervention studies. Second, since the main included studies, the number of correlation effect size (k) for
interest of present meta-analysis was to assess special edu- each variable was less than 3 in several cases. Considering
cation teacher burnout associated with student-, teacher-, the limited published articles on related variables such as
and school-related variables, we could not provide longitu- teacher resources and school personnel, generalizing the
dinal evidence of why special education teachers refuse to findings should be considered carefully. Thus, in the future,
14 SAGE Open

researchers should conduct studies focusing on burnout 2013), educators should avoid solely calculating total burn-
issues and related variables that are not being currently out that combines these three scores. The three burnout
investigated. results should be considered individually when examining
teachers’ burnout degrees.
Implications for Teacher Education and Training
Conclusion
The findings of this meta-analysis reveal four implications
for teacher education and training. First, for teacher-related The significant variables associated with special education
variables, special education teachers’ self-efficacy was sig- teacher burnout rates were level of self-efficacy in teacher-
nificantly related to all three burnout scales (95% CI = related variables and support from school personnel in
[−0.562, −0.218] for emotional exhaustion, 95% CI = school-related variables. The effects of student-related vari-
[−0.415, −0.227] for depersonalization, 95% CI = [0.474, ables vary depending on the extent of burnout, suggesting
0.793] for personal accomplishment). Thus, when local that the situation of individual teachers, including student
school districts provide teacher training for special education characteristics, the evaluation of teacher burnout, and man-
teachers, the program content should emphasize the improve- agement, needs to be considered. The effect sizes of student-
ment of their self-efficacy level, which is closely related to related variables were different across burnout dimensions.
their burnout degree. Second, student-related variables were Teacher education and training programs that consider the
found to be differently related to the three dimensions of significant variables, especially self-efficacy, might be effec-
burnout. A possible explanation is that student-related vari- tive in reducing special education teachers’ burnout levels.
ables may differ more and more complex interaction accord- Regarding the results of school-related variables, effective
ing to the individual special education teacher’s contest than support might be developed for special education teachers,
other variables. Individual and school-level intervention including support from school personnel.
plans are needed to strengthen the capacity of teachers so
that individual teachers can cope with their situation Declaration of Conflicting Interests
(Argentin et al., 2014; Caputo & Rastelli, 2014). Third, our The author(s) declared no potential conflicts of interest with respect
results regarding school-related variables indicated that spe- to the research, authorship, and/or publication of this article.
cial education teachers’ support from school personnel was
another variable that was significantly related to their burn- Funding
out scale (95% CI = [−0.172, −0.065] for emotional exhaus- The author(s) received no financial support for the research, author-
tion; 95% CI = [−0.157, −0.122] for depersonalization; 95% ship, and/or publication of this article.
CI = [0.127, 0.217] for personnel accomplishment). Lack of
support from school personnel has one of the most frequently ORCID iD
cited relations with teachers’ burnout. Furthermore, the
Mikyung Shin https://orcid.org/0000-0001-7907-9193
potential role of perceived support was also reported to be
associated with reduced emotional exhaustion and improved
References
personal accomplishment and with developing professional
special education teachers (Langher et al., 2017). This can be References marked with an asterisk (*) indicate studies included in
the meta-analysis.
seen by looking at the number of effect sizes analyzed in this
Adera, B. A., & Bullock, L. M. (2010). Job stressors and teacher job
meta-analysis. The number of effect sizes of support from satisfaction in programs serving students with emotional and
school personnel was 66 in emotional exhaustion, 58 in behavioral disorders. Emotional and Behavioural Difficulties,
depersonalization, and 59 in personnel accomplishment. 15, 5–14. https://doi.org/10.1080/13632750903512365
Perceived lack of support may lead to feelings of burnout. Aloe, A. M., Shisler, S. M., Norris, B. D., Nickerson, A. B., & Rinker,
Research has shown that special education teachers feel T. W. (2014). A multivariate meta-analysis of student misbe-
more supported by the school when they have administrators havior and teacher burnout. Educational Research Review, 12,
who value their opinions, provide effective feedback, and 30–44. https://doi.org/10.1016/j.edurev.2014.05.003
engage teachers in policy decisions (Bettini et al., 2015). *Ambrosini, M. (2013). The relationship of role-based, task-based,
When considering support from school personnel, schools boundary-spanning, and conflict-mediating stress experienced
should provide various types of support for special education by New York state special education administrators and emo-
tional exhaustion, depersonalization, and personal accom-
teachers. Finally, the results revealed that in several cases,
plishment [Doctoral dissertation]. ProQuest Dissertations &
the correlation effects displayed various features for each Theses Database. (UMI No. 3570133)
burnout dimension: emotional exhaustion, depersonaliza- Araujo, J., & Born, D. G. (1985). Calculating percentage agreement
tion, and personal accomplishment. Given that these dimen- correctly but writing its formula incorrectly. The Behavior
sions of emotional exhaustion, depersonalization, or personal Analyst, 8, 207–208. https://doi.org/10.1007/BF03393152
accomplishment under the MBI (Maslach & Jackson, 1996) Argentin, G., Pennisi, A., Vidoni, D., Abbiati, G., & Caputo, A.
were constructed distinctively (Byrne, 1994; Denton et al., (2014). Trying to raise (low) math achievement and to promote
Park and Shin 15

(rigorous) policy evaluation in Italy: Evidence from a large- intermediate, and secondary teachers. American Edu­
scale randomized trial. Evaluation Review, 38, 99–132. https:// cational Research Journal, 31, 645–673. https://doi.
doi.org/10.1177/0193841X14529125 org/10.3102/00028312031003645
*Banks, S. R., & Necco, E. G. (1990). The effects of special edu- Caputo, A., & Rastelli, V. (2014). School improvement plans and
cation category and type of training on job burnout in special student achievement: Preliminary evidence from the Quality
education teachers. Teacher Education and Special Education, and Merit Project in Italy. Improving Schools, 17(1), 72–98.
13, 187–191. https://doi.org/10.1177/088840649001300309 https://doi.org/10.1177/1365480213515800
*Bataineh, O. (2009). Sources of social support among special *Caputo, A., & Langher, V. (2015). Validation of the collaboration
education teachers in Jordan and their relationship to burnout. and support for inclusive teaching scale in special education
International Education, 39, 65–78. teachers. Journal of Psychoeducational Assessment, 33, 210–
*Bataineh, O., & Alsagheer, A. (2012). An investigation of social 222. https://doi.org/10.1177/0734282914548335
support and burnout among special education teachers in Carlson, B. C., & Thompson, J. A. (1995). Job burnout and job
the United Arab Emirates. International Journal of Special leaving in public school teachers: Implications for stress man-
Education, 27, 5–13. agement. International Journal of Stress Management, 2, 15–
Bettini, E. A., Cheyney, K., Wang, J., & Leko, C. (2015). Job 29. https://doi.org/10.1007/BF01701948
design: An administrator’s guide to supporting and retaining *Chatlos, C. N. (2016). Special education teachers and autism
special educators. Intervention in School and Clinic, 50, 221– spectrum disorders: The role of knowledge and self-efficacy
225. https://doi.org/10.1177/1053451214532346 on burnout [Doctoral dissertation]. ProQuest Dissertations &
*Biglan, A., Layton, G. L., Jones, L. B., Hankins, M., & Rusby, Theses Database. (UMI No. 10133318)
J. C. (2013). The value of workshops on psychological flex- *Coman, D., Alessandri, M., Gutierrez, A., Novotny, S., Boyd,
ibility for early childhood special education staff. Topics in B., Hume, K., . . . Odom, S. (2013). Commitment to class-
Early Childhood Special Education, 32, 196–210. https://doi. room model philosophy and burnout symptoms among high
org/10.1177/0271121411425191 fidelity teachers implementing preschool programs for chil-
Biostat. (2017). Comprehensive Meta-Analysis [Computer soft- dren with autism spectrum disorders. Journal of Autism
ware] (Version 3). https://www.meta-analysis.com/ and Developmental Disorders, 43, 345–360. https://doi.
*Bloom, J. A. (1992). Effect of student-teacher ratio on student org/10.1007/s10803-012-1573-1
achievement and teacher reported burnout in special educa- Cooley, E., & Yovanoff, P. (1996). Supporting professionals-at-
tion [Doctoral dissertation]. ProQuest Dissertations & Theses risk: Evaluating interventions to reduce burnout and improve
Database. (UMI No. 9307329) retention of special educators. Exceptional Children, 62, 336–
Borenstein, M., Hedges, L. V., Higgins, J., & Rothstein, H. (2009). 355. https://doi.org/10.1177/001440299606200404
Introduction to meta-analysis. John Wiley. Cooper, H. (1998). Synthesizing research: A guide for literature
*Boujut, E., Dean, A., Grouselle, A., & Cappe, E. (2016). reviews (3rd ed.). Sage.
Comparative study of teachers in regular schools and teachers *Crane, S., & Iwanicki, E. (1986). Perceived role conflict, role
in specialized schools in France, working with students with ambiguity, and burnout among special education teach-
an Autism Spectrum Disorder: Stress, social support, coping ers. Remedial and Special Education, 7, 24–31. https://doi.
strategies and burnout. Journal of Autism and Developmental org/10.1177/074193258600700206
Disorders, 46, 2874–2889. https://doi.org/10.1007/s10803- Denton, E., Chaplin, W. F., & Wall, M. (2013). Teacher burnout:
016-2833-2 A comparison of two cultures using confirmatory factor and
Brackett, M., Palomera, R., Mojsa-Kaja, J., Reyes, M. R., & Salovey, item response models. International Journal of Quantitative
P. (2010). Emotion-regulation ability, burnout, and job satis- Research in Education, 1, 147–166. https://doi.org/10.1504/
faction among British secondary-school teachers. Psychology IJQRE.2013.056463
in the Schools, 47, 406–417. https://doi.org/10.1002/pits.20478 *Dickerson, E. G. (2017). A quantitative study on burnout for
Breeman, L. D., van Lier, P. A., Wubbels, T., Verhulst, F. C., van teachers who work with students who have moderate to severe
der Ende, J., Maras, A., . . . Tick, N. T. (2016). Effects of the disabilities [Doctoral dissertation]. ProQuest Dissertations &
Good Behavior Game on the behavioral, emotional, and social Theses Database. (UMI No. 10606610)
problems of children with psychiatric disorders in special edu- Egger, M., Smith, G. D., Schneider, M., & Minder, C. (1997). Bias
cation settings. Journal of Positive Behavior Interventions, 18, in meta-analysis detected by a simple, graphical test. British
156–167. https://doi.org/10.1177/1098300715593466 Medical Journal, 315, 629–634. https://doi.org/10.1136/
Brewer, E. W., & McMahan, J. (2004). Job stress and burnout bmj.315.7109.629
among industrial and technical teacher educators. Journal of *Eichinger, J., Heifetz, L. J., & Ingraham, C. (1991). Situational
Vocational Education Research, 28, 125–140. shifts in sex role orientation: Correlates of work satisfaction
Bronfenbrenner, U. (1977). Toward an experimental ecology of and burnout among women in special education. Sex Roles, 25,
human development. American Psychologist, 32, 513–531. 425–440. https://doi.org/10.1007/BF00292532
https://doi.org/10.1037/0003-066X.32.7.513 *Embich, J. L. (2001). The relationship of secondary special educa-
Brunsting, N. C., Sreckovic, M. A., & Lane, K. L. (2014). Special tion teachers’ roles and factors that lead to professional burn-
education teacher burnout: A synthesis of research from 1979 out. Teacher Education and Special Education, 24, 58–69.
to 2013. Education and Treatment of Children, 37, 681–711. https://doi.org/10.1177/088840640102400109
https://doi.org/10.1353/etc.2014.0032 Emery, D. W., & Vandenberg, B. (2010). Special education teacher
Byrne, B. M. (1994). Burnout: Testing for the validity, replica- burnout and ACT. International Journal of Special Education,
tion, and invariance of causal structure across elementary, 25, 119–131.
16 SAGE Open

Every Student Succeeds Act, 20 U.S.C. §§ 6301 et seq. (2015). analysis. Review of Educational Research, 68, 61–69. https://
Fernet, C., Guay, F., Senécal, C., & Austin, S. (2012). Predicting doi.org/10.3102/00346543068001061
intraindividual changes in teacher burnout: The role of per- Hagaman, J. L., & Casey, K. J. (2018). Teacher attrition in
ceived school environment and motivational factors. Teaching special education: Perspectives from the field. Teacher
and Teacher Education, 28, 514–525. https://doi.org/10.1016/j. Education and Special Education, 41, 277–291. https://doi.
tate.2011.11.013 org/10.1177/0888406417725797
*Fimian, M. J., & Blanton, L. P. (1986). Variables related to stress Hall, N. C. (2013). Emotion, motivation, and self-regulation: A
and burnout in special education teacher trainees and first-year handbook for teachers. Emerald Group Publishing.
teachers. Teacher Education and Special Education, 9, 9–21. *Hassan, A., & Mohamed, H. (2015). Burnout and work stress
https://doi.org/10.1177/088840648600900102 among disability centers staff in Oman. International Journal
Fisher, M. H. (2011). Factors influencing stress, burnout, and reten- of Special Education, 30, 25–36.
tion of secondary teachers. Current Issues in Education, 14, *Hastings, R. P., & Brown, T. (2002). Coping strategies and
1–37. http://cie.asu.edu/ the impact of challenging behaviors on special educa-
Fitchett, P. G., McCarthy, C. J., Lambert, R. G., & Boyle, L. (2018). tors’ burnout. Mental Retardation, 40, 148–156. https://doi.
An examination of U.S. first-year teachers’ risk for occupa- org/10.1352/0047-6765(2002)040<0148:CSATIO>2.0.CO;2
tional stress: Associations with professional preparation and *Hill, K. A. (2011). Burnout experience of teachers serving stu-
occupational health. Teachers and Teaching, 24, 99–118. dents with emotional behavioral disorders in grades PreK-8
https://doi.org/10.1080/13540602.2017.1386648 within non-public special education day schools [Doctoral dis-
Fore, C., III, Martin, C., & Bender, W. N. (2002). Teacher burnout sertation]. ProQuest Dissertations & Theses Database. (UMI
in special education: The causes and the recommended solu- No. 3434701)
tions. High School Journal, 86, 36–44. https://doi.org/10.1353/ *Hopman, J. A., Tick, N. T., van der Ende, J., Wubbels, T., Verhulst,
hsj.2002.0017 F. C., Maras, A., . . . van Lier, P. A. (2018). Special education
Frank, A. R., & McKenzie, R. (1993). The development of teachers’ relationships with students and self-efficacy moder-
burnout among special educators. Teacher Education ate associations between classroom-level disruptive behaviors
and Special Education, 16, 161–170. https://doi. and emotional exhaustion. Teaching and Teacher Education,
org/10.1177/088840649301600208 75, 21–30. https://doi.org/10.1016/j.tate.2018.06.004
*Freed, E. H. (1994). The relationship of social support, occupa- Iancu, A. E., Rusu, A., Măroiu, C., Păcurar, R., & Maricuțoiu, L.
tional stress, burnout, and job satisfaction among special edu- P. (2018). The effectiveness of interventions aimed at reduc-
cation teachers [Doctoral dissertation]. ProQuest Dissertations ing teacher burnout: A meta-analysis. Educational Psychology
& Theses Database. (UMI No. 9422637) Review, 30, 373–396. https://doi.org/10.1007/s10648-017-
*Garwood, J. D., Werts, M. G., Varghese, C., & Gosey, L. 9420-8
(2018). Mixed-methods analysis of rural special educa- Jennett, H. K., Harris, S. L., & Mesibov, G. B. (2003). Commitment to phi-
tors’ role stressors, behavior management, and burnout. losophy, teacher efficacy, and burnout among teachers of children
Rural Special Education Quarterly, 37, 30–43. https://doi. with autism. Journal of Autism and Developmental Disorders, 33,
org/10.1177/8756870517745270 583–593. https://doi.org/10.1023/B:JADD.0000005996.19417.57
Gersten, R., Chard, D. J., Jayanthi, M., Baker, S. K., Morphy, P., Kaff, M. S. (2004). Multitasking is multitaxing: Why special educa-
& Flojo, J. (2009). Mathematics instruction for students with tors are leaving the field. Preventing School Failure, 48, 10–17.
learning disabilities: A meta-analysis of instructional com- Kvande, M. N., Belsky, J., & Wichstrøm, L. (2018). Selection
ponents. Review of Educational Research, 79, 1202–1242. for special education services: The role of gender and socio-
https://doi.org/10.3102/0034654309334431 economic status. European Journal of Special Needs Education,
*Goetzinger, E. K. (2006). Burnout among special educators: Do 33, 510–524. https://doi.org/10.1080/08856257.2017.1373493
experience, certification, caseload, and school size make a *LaMonica, A. (1983). An investigation of factors related to psy-
difference? [Doctoral dissertation]. ProQuest Dissertations & chological stress and burnout in teachers of severely handi-
Theses Database. (UMI No. 3237516) capped children [Doctoral dissertation]. ProQuest Dissertations
*Gong, T., Zimmerli, L., & Hoffer, H. E. (2013). The effects & Theses Database. (UMI No. 8410505)
of transformational leadership and the sense of call- Langher, V., Caputo, A., & Ricci, M. E. (2017). The potential role
ing on job burnout among special education teachers. of perceived support for reduction of special education teach-
Journal of School Leadership, 23, 969–993. https://doi. ers’ burnout. International Journal of Educational Psychology,
org/10.1177/105268461302300603 6, 120–147. https://doi.org/10.17583/ijep.2017.2126
*Goodall, B. J. B. (1986). Intergroup variability of burnout among Lee, J. H. J., & Ok, C. (2012). Reducing burnout and enhancing job
special education teachers of the emotionally disturbed, satisfaction: Critical role of hotel employees’ emotional intelli-
behavioral disordered, severe learning disabled, and the train- gence and emotional labor. International Journal of Hospitality
able mentally handicapped [Doctoral dissertation]. ProQuest Management, 31, 1101–1112. https://doi.org/10.1016/j.
Dissertations & Theses Database. (UMI No. 8622979) ijhm.2012.01.007
Gopalakrishnan, S., & Ganeshkumar, P. (2013). Systematic reviews Lee, Y. L., Patterson, P. P., & Vega, L. A. (2011). Perils to self-
and meta-analysis: Understanding the best evidence in primary efficacy perceptions and teacher-preparation quality among
healthcare. Journal of Family Medicine and Primary Care, 2, special education intern teachers. Teacher Education Quarterly,
9–14. https://doi.org/10.4103/2249-4863.109934 38, 61–76.
Guglielmi, R. S., & Tatrow, K. (1998). Occupational stress, burn- Leko, M. M., & Smith, S. W. (2010). Retaining beginning spe-
out and health in teachers: A methodological and theoretical cial educators: What should administrators know and do?
Park and Shin 17

Intervention in School and Clinic, 45, 321–325. https://doi. Rosenthal, R. (1979). The file drawer problem and tolerance for
org/10.1/1053451209353441 null results. Psychological Bulletin, 86, 638–641. https://doi.
Leung, D. Y., & Lee, W. W. (2006). Predicting intention to quit org/10.1037/0033-2909.86.3.638
among Chinese teachers: Differential predictability of the com- *Ruble, L. A., & McGrew, J. H. (2013). Teacher and child predic-
ponents of burnout. Anxiety, Stress, and Coping, 19, 129–141. tors of achieving IEP goals of children with autism. Journal of
https://doi.org/10.1080/10615800600565476 Autism and Developmental Disorders, 43, 2748–2763. https://
Li, C., Wang, C. K. J., Pyun, D. Y., & Kee, Y. H. (2013). Burnout doi.org/10.1007/s10803-013-1884-x
and its relations with basic psychological needs and motiva- *Ruble, L. A., Usher, E. L., & McGrew, J. H. (2011). Preliminary
tion among athletes: A systematic review and meta-analysis. investigation of the sources of self-efficacy among teach-
Psychology of Sport and Exercise, 14, 692–700. https://doi. ers of students with autism. Focus on Autism and Other
org/10.1016/j.psychsport.2013.04.009 Developmental Disabilities, 26, 67–74. https://doi.
Lipsey, M. W. (2003). Those confounded moderators in meta- org/10.1177/1088357610397345
analysis: Good, bad, and ugly. The ANNALS of the American Ryan, S. V., von der Embse, N. P., Pendergast, L. L., Saeki, E.,
Academy of Political and Social Science, 587, 69–81. https:// Segool, N., & Schwing, S. (2017). Leaving the teaching profes-
doi.org/10.1177/0002716202250791 sion: The role of teacher stress and educational accountability
Lipsey, M. W., & Wilson, D. B. (2001). Practical meta-analysis. policies on turnover intent. Teaching and Teacher Education,
Sage. 66, 1–11. https://doi.org/10.1016/j.tate.2017.03.016
*Martin, A. M. (2010). Predictors of burnout and self-efficacy *Sari, H. (2005). How do principals and teachers in special schools
among special education teachers [Doctoral disserta- in Turkey rate themselves on levels of burnout, job satisfaction,
tion]. ProQuest Dissertations & Theses Database. (UMI No. and locus of control? Alberta Journal of Educational Research,
3400670) 51, 172–192.
Maslach, C. (1982). Burnout: The cost of caring. Prentice Hall. *Sariçam, H., & Sakiz, H. (2014). Burnout and teacher self-efficacy
Maslach, C., & Jackson, S. (1996). Maslach Burnout Inventory among teachers working in special education institutions in
manual (3rd ed.). Consulting Psychologists Press. Turkey. Educational Studies, 40, 423–437. https://doi.org/10.
Maslach, C., Schaufeli, W. B., & Leiter, M. P. (2001). Job burn- 1080/03055698.2014.930340
out. Annual Review of Psychology, 52, 397–422. https://doi. Savaş, A. C., Bozgeyik, Y., & Eser, İ. (2014). A study on the rela-
org/10.1146/annurev.psych.52.1.397 tionship between teacher self-efficacy and burnout. European
*McDow, S. M. S. (1993). A study of special education STRs and Journal of Educational Research, 3, 159–166.
caseloads in Oregon and their impact upon teachers’ percep- Sedgwick, M. L. (1998). Teacher burnout, stressful student mis-
tions of success, job satisfaction and burnout [Unpublished behavior, and the strategies teachers use to influence that
doctoral dissertation]. Oregon State University. misbehavior [Doctoral dissertation]. ProQuest Dissertations &
*McIntyre, T. C. (1984). The relationship between locus of control Theses Database. (UMI No. 9907756)
and teacher burnout. British Journal of Educational Psychology, *Shea, C. A. (1984). An investigation of the emotional exhaus-
54, 235–238. https://doi.org/10.1111/j.2044-8279.1984.tb0 tion aspect of burnout and stressors in resource LD teach-
2585.x ers [Doctoral dissertation]. ProQuest Dissertations & Theses
Montgomery, C., & Rupp, A. A. (2005). A meta-analysis for Database. (UMI No. 8500604)
exploring the diverse causes and effects of stress in teach- Shen, B., McCaughtry, N., Martin, J., Garn, A., Kulik, N., &
ers. Canadian Journal of Education, 28, 458–486. https://doi. Fahlman, M. (2015). The relationship between teacher burn-
org/10.2307/4126479s out and student motivation. British Journal of Educational
*Moss, C. L. (2014). Role conflict and role ambiguity as predic- Psychology, 85, 519–532. https://doi.org/10.1111/bjep.12089
tors of burnout in special and general education co-teachers Shoji, K., Cieslak, R., Smoktunowicz, E., Rogala, A., Benight,
[Doctoral dissertation]. ProQuest Dissertations & Theses C. C., & Luszczynska, A. (2016). Associations between job
Database. (UMI No. 3646995) burnout and self-efficacy: A meta-analysis. Anxiety, Stress, &
Nichols, A. S., & Sosnowsky, F. L. (2002). Burnout among special Coping, 29, 367–386. https://doi.org/10.1080/10615806.2015.
education teachers in self-contained cross-categorical class- 1058369
rooms. Teacher Education and Special Education, 25, 71–86. *Shyman, E. (2010). Identifying predictors of emotional exhaustion
https://doi.org/10.1177/088840640202500108 among special education paraeducators: A preliminary inves-
*Platsidou, M. (2010). Trait emotional intelligence of Greek spe- tigation. Psychology in the Schools, 47, 828–841. https://doi.
cial education teachers in relation to burnout and job satisfac- org/10.1002/pits.20507
tion. School Psychology International, 31, 60–76. https://doi. Skaalvik, E. M., & Skaalvik, S. (2010). Teacher self-efficacy
org/10.1177/0143034309360436 and teacher burnout: A study of relations. Teaching and
*Platsidou, M., & Agaliotis, I. (2008). Burnout, job satisfac- Teacher Education, 26, 1059–1069. https://doi.org/10.1016/j.
tion and instructional assignment-related sources of stress in tate.2009.11.001
Greek special education teachers. International Journal of Special Education Act for Individuals with Disabilities and Others.
Disability, Development and Education, 55, 61–76. https://doi. (2008). Act No. 8852. https://elaw.klri.re.kr/kor_service/
org/10.1080/10349120701654613 lawView.do?hseq=16226&lang=ENG
Rosenberg, M. S. (2005). The file-drawer problem revisited: U.S. Department of Education, Office of Postsecondary Education.
A general weighted method for calculating fail-safe num- (2017). Teacher Shortage Areas (TSA) nationwide listing for
bers in meta-analysis. Evolution, 59, 464–468. https://doi. 1990-1991 through 2017-2018. https://www2.ed.gov/about/
org/10.1111/j.0014-3820.2005.tb01004.x offices/list/ope/pol/tsa.html#list
18 SAGE Open

*Weber, D. B., & Toffler, J. D. (1989). Burnout among teachers IEP outcomes. Exceptional Children, 83, 412–427. https://doi.
of students with moderate, severe, or profound mental retarda- org/10.1177/0014402917690729
tion. Teacher Education and Special Education, 12, 117–125. Zabel, M. K., Dettmer, P. A., & Zabel, R. H. (1984). Factors of
https://doi.org/10.1177/088840648901200306 emotional exhaustion, depersonalization, and sense of accom-
Williams, J., & Dikes, C. (2015). The implications of demographic plishment among teachers of the gifted. Gifted Child Quarterly,
variables as related to burnout among a sample of special edu- 28, 65–69. https://doi.org/10.1177/001698628402800204
cation teachers. Education, 135, 337–345. *Zabel, R. H., & Zabel, M. K. (2002). Burnout among special edu-
Wisniewski, L., & Gargiulo, R. M. (1997). Occupational stress cation teachers and perceptions of support. Journal of Special
and burnout among special educators: A review of the litera- Education Leadership, 15, 67–73.
ture. Journal of Special Education, 31, 325–346. https://doi. *Zarafshan, H., Mohammadi, M. R., Ahmadi, F., & Arsalani, A.
org/10.1177/002246699703100303 (2013). Job burnout among Iranian elementary school teachers
*Wong, V. W., Ruble, L. A., Yu, Y., & McGrew, J. H. (2017). Too of students with autism: A comparative study. Iranian Journal
stressed to teach? Teaching quality, student engagement, and of Psychiatry, 8, 20–27.