Psychology
2013. Vol.4, No.1, 73-82
Published Online January 2013 in SciRes (http://www.scirp.org/journal/psych) http://dx.doi.org/10.4236/psych.2013.41010
Validity and Reliability of the Socio-Contextual Teacher
Burnout Inventory (STBI)
Janne Pietarinen1*, Kirsi Pyhältö2, Tiina Soini3, Katariina Salmela-Aro4
1School of Educational Sciences and Psychology, University of Eastern Finland, Joensuu, Finland
2University of Helsinki Centre for Research and Development of Higher Education,
Institute of Behavioural Sciences, University of Helsinki, Helsinki, Finland
3School of Education, University of Tampere, Tampere, Finland
4Helsinki Collegium for Advanced Studies, University of Helsinki, Helsinki, Finland
Received October 1st, 2012; revised November 7th, 2012; accepted December 7th, 2012
Recent research on teacher burnout has advanced our understanding of its dimensions and contributing
factors. However, the complexity and dynamics of the social working environments in schools has often
been neglected in teacher burnout studies, and hence a valid and reliable context-sensitive instrument for
studying teacher burnout in terms of social interaction in schools is needed. This study examined the de-
velopment of the Socio-Contextual Teacher Burnout Inventory (STBI), its validity as well as reliability,
among Finnish teachers (n = 2310). The validity and reliability of the items composing the STBI were
determined based on the confirmatory factor analysis. The results showed that the correlated three-factor
solution and second-order-factor solution fitted the data. More specifically, teacher exhaustion, cynicism
towards the teacher community and inadequacy in the pupil-teacher relationship were found to be closely
related but separate constructs. The results also supported the main hypothesis that teacher burnout can be
examined in terms of interpersonal problems in an individual’s transactions with others in the workplace.
Therefore the sources of teacher burnout may vary not only between schools but also between the social
working contexts within a single school. The instrument introduced in this study is a potentially useful
tool for exploring interpersonal teacher burnout.
Keywords: Teacher Burnout Inventory; Occupational Well-Being; School; Social Working Contexts;
Confirmatory Factor Analysis; Construct Validity
Introduction
Teacher burnout has been recognised as a serious occupa-
tional problem (e.g. Borg & Riding, 1991; Loonstra, Browers,
& Tomic, 2009; Kinnunen, Parkatti, & Rasku, 1994; Rudow,
1999). In comparison with other academic client-related pro-
fessions, teachers have been found to surpass the average levels
of stress (Travers & Cooper, 1993). In particular, they have
been found to experience high levels of exhaustion and cyni-
cism, both of which constitute the core dimensions of burnout
(Kalimo & Hakanen, 2000; Schaufeli & Enzmann, 1998). Re-
cent research on teacher burnout has enhanced our understand-
ing of its levels and dimensions and the factors that contribute
to it (Dorman, 2003; Skaalvik & Skaalvik, 2009). However, the
complexity and dynamics of the social working environments
in schools has often been neglected in studies and measure-
ments of burnout among teachers (Devos, Dupriez, & Paquay,
2012; Leiter & Maslach, 1988), and hence a valid and reliable
context-sensitive instrument for examining teacher burnout in
terms of social interactions within schools is needed. School as
a complex working environment that includes multiple social
contexts impose at least partly unique demands on teachers’
occupational well-being (Skaalvik & Skaalvik, 2010; Pyhältö et
al., 2011). This study reports on the development of an instru-
ment for measuring teacher burnout within the social contexts
of a teacher’s working environment. The developed instrument
builds on one of the most commonly used Maslach and Jack-
son’s (1981) burnout inventory.
Teacher Burnout
Burnout develops gradually as a consequence of prolonged
and extensive work-related stress (Freudenberg, 1974; Holland,
1982; Peeters & Rutte, 2005). The syndrome has three distinct
symptoms, exhaustion (Maslach, Schaufeli, & Leiter, 2001),
cynicism (Bakker, Schaufeli, Leiter, & Taris, 2008; Maslach &
Leiter, 2005; Maslach & Leiter, 1999; Schaufeli & Buunk,
2003) and professional inadequacy (Brouwers & Tomic, 2000;
Hakanen, Bakker, & Schaufeli, 2006; meta-analysis by Mont-
gomery & Rupp, 2005).
Exhaustion means a lack of emotional energy, and feelings of
being strained and tired at work (Maslach et al., 2001), whereas
cynicism means indifference or aloofness towards work in gen-
eral, and also a disaffected or acerbic attitude towards students,
parents or colleagues, as well as low organisational commit-
ment (Schaufeli & Buunk, 2003). Professional inadequacy,
referring to feelings of insufficient competence, encompasses
both social and non-social aspects of occupational accom-
plishments (Brouwers & Tomic, 2000; Hakanen et al., 2006;
meta-analysis by Montgomery & Rupp, 2005). Burnout as a
social problem in many service professions has been the impe-
tus for the much earlier research (Maslach, 2003). However, the
*Corresponding author.

J. PIETARINEN ET AL.
role played by social behaviour has been largely neglected,
even though burnout was identified as a social problem by
both workers and social commentators (Maslach, 2003) long
before it became a focus of systematic empirical study. In fact,
since the concept emerged, burnout has been presented not as
an individual stress response, but rather in terms of interper-
sonal problems in an individual’s transactions with others in
the workplace, and theories of social comparison (Buunk,
Ybema, Gibbons, & Ipenburg, 2001; Festinger, 1954) and so-
cial exchange (Adams, 1965) provided an early framework
(Buunk & Schaufeli, 1993; Geurts, Schaufeli, & De Jonge,
1998; Schaufeli, Van Dierendonck, & van Gorp, 1996),
whereas more recent studies have focused on social support as a
buffering resource (Xanthopoulou, Bakker, Demerouti, &
Schaufeli, 2007).
Social interrelations are very frequent in the working envi-
ronment of teachers within the school community. For example,
during a school day teachers often work with different pupil
groups and various members of the professional community.
The destructive friction and other problems in the working
environments of schools, such as lack of social support, per-
ceived inequity and a poor sense of community (Santavirta,
Solovieva, & Theorell, 2007; Milfont, Denny, Ameratunga,
Robinson, & Merry, 2008; Sharplin, O’Neill, & Chapman,
2011) have been suggested to be central sources of exhaustion
and stress for teachers. As well, unresolved confrontations,
complex social interactions have been suggested to contribute
to teachers feeling burdened (Cano-Garcia, Padilla-Munoz, &
Carrasco-Ortiz, 2005; Dorman, 2003; Leung & Lee, 2006;
Schaufeli & Bakker, 2004). This indicates that the sources of
teacher burnout may vary not only between schools but also
between the social working contexts within a single school (see
also Kokkinos, 2007; Skaalvik & Skaalvik, 2009; Fernet, Guay,
Senégal, & Austin, 2012). However, thus far the role of differ-
ent social contexts in teacher burnout has been neglected in the
research.
The authors’ previous study on teacher burnout showed that
it is situated primarily in social interactions within the school
community. More specifically, the professional community and
teacher-pupil interaction are the primary arenas of teacher
burnout, particularly in terms of perceived inadequacy and
cynicism (Pyhältö, Pietarinen, & Salmela-Aro, 2011). However,
an inventory is lacking on teacher burnout in its social context.
Hence, this study analyses the teacher burnout experience em-
bedded in teacher-pupil and teacher-community interactions,
and aims to develop such an inventory.
Work Environment and Teacher Burnout
Previous studies have shown that gender, school size and
academic level have an influence on teacher burnout. Female
teachers, for instance, have been found to have a higher risk for
exhaustion in comparison to male colleagues (Tatar & Horenc-
zyk, 2003; Antoniou, Polychroni, & Vlachakis, 2006; Fernet et
al., 2012). Moreover, evidence shows that teachers who work in
large school tend to receive less social support from the pro-
fessional community compared to those working in smaller
schools. School size has been reported to play a role in teacher
burnout as manifested in decreased job satisfaction, lower ac-
complishment and sense of depersonalization (Skaalvik &
Skaalvik, 2010).
Further, some indicators show that the school level at which
a teacher works, i.e. primary or secondary school, may also be
associated with teacher exhaustion and lower job satisfaction
(Jepson & Forrest, 2006). In this regard, pupil age is signifi-
cant (Klassen, 2010; Tatar & Horenczyk, 2003; Wolters &
Daugherty, 2007). Accordingly, the working environment at
different school levels is likely to place at least partly unique
demands on teacher groups, including class, subject and special
education teachers, working at those particular levels (Byrne,
1993; Antoniou et al., 2006; Jepson & Forrest, 2006; Embich,
2007).
The Present Study
This study examined the development of the Socio-Contextual
Teacher Burnout Inventory (STBI), its validity as well as reli-
ability, among class, subject and special education teachers in
Finland. The aim was first to determine the construct validity of
the developed instrument. In the light of previous research on
teacher burnout (Byrne, 1993; Maslach & Leiter, 1999; Lau,
Yuen, & Chan, 2005; Pyhältö et al., 2011) the structure of the
STBI was tested by analysing the goodness-of-fit of the
three-factor model of teacher burnout. It was expected that a
model consisting of three correlated factors measuring exhaust-
tion, cynicism towards the teacher community and sense of
inadequacy in teacher-pupil interaction would describe the
phenomenon of teacher burnout embedded in key social con-
texts of teachers’ work (Pyhältö et al., 2011; Salmela-Aro,
Rantanen, Hyvönen, Tilleman, & Feldt, 2011). Moreover, a
second-order-factor model existing behind the three-factor
model was also expected. These two models were expected to
describe the phenomenon of socio-contextual teacher burnout
better than a one-factor model representing overall burnout. In
addition, various aspects of reliability of the STBI, i.e. item
reliability and scale reliability, were investigated. Finally, the
socio-contextual sensitivity of the STBI was investigated by
examining the observed differences in class, subject and special
education teachers’ perceived exhaustion, cynicism towards the
teacher community and sense of inadequacy in teacher-pupil
interaction.
The first hypothesis was that teachers’ experienced exhaust-
tion, cynicism towards the teacher community and inadequacy
in the pupil-teacher relationship were closely related but sepa-
rate constructs (e.g. Maslach & Leiter, 1999; Lee & Ashforth,
1996). Hence, the three-factor model (M1) was selected as a
primary model (see Figure 1). Previous studies suggest that
these socio-contextual components of teacher burnout should
also indicate teachers’ overall risk for burnout (Byrne, 1993;
Pyhältö et al., 2011; Salmela-Aro et al., 2011). For this reason,
the second-order model (M2) was also expected to fit the data
(hypothesis 2). Moreover, it was expected that gender, educa-
tional background, school size and partly specified pedagogical
tasks at different academic levels would influence teachers’
experienced exhaustion, cynicism towards the teacher commu-
nity and inadequacy in teacher-pupil interaction (hypothesis 3).
Methods
Participants
Altogether 2310 Finnish comprehensive school teachers, in-
cluding primary (n = 815; 35%), subject (n = 729; 32%), and
special education teachers (n = 761; 33%) completed the survey.
A probability sampling method (N = 6000) was used. The total
74

J. PIETARINEN ET AL.
Figure 1.
Three-factor (model M1), second-order (model M2) and one-factor (model 3) factorial structures of the Socio-Contextual Teacher
Burnout Inventory (STBI). EXH exhaustion, CYN cynicism towards professional community, INAD inadequacy in teacher-pupil in-
teraction and OB overall socio-contextual burnout.
Table 1.
response rate was 39%. All respondents had MA degrees, and
all were in various phases of their careers. Hence the response
rates of class, subject, and special education teachers were not
biased. Moreover, the schools in which the participants worked
varied both in size and in terms of grades that were taught. The
specified non-response analysis is presented in Table 1 .
The non-response analysis.
Sample statistics
*National statistics
n = 2310 teachers N = 36,890 teachers
Gender
Female 81% 73%
Male 19% 27%
Teachers’ age
<40 years 38% 32%
40 - 49 years 30% 33%
≥50 years 32% 35%
The analysis showed that the representativeness of the sam-
ple was plausible (see Table 1). The mean age of the respon-
dents was 45.3 years (SD = 9.84; Min/Max: 25/68 years).
Hence in terms of age the sample was well-representative of the
Finnish teacher population (see also National Board of Educa-
tion, 2010). The majority of the respondents were women (n =
1878) and the minority men (n = 429). Accordingly, female
teachers were slightly over-represented in the sample (see Ta-
ble 1).
Measures Note: *(National Board of Education, 2010).
The Socio-contextual Teacher Burnout Inventory (STBI) was
developed by the authors. It draws on both Maslach and Jack-
son’s (1981) burnout scale and Elo, Leppänen and Jahkola’s
(2003) single-item stress scale in terms of measuring teachers’
perceived exhaustion. The STBI was constructed by specifying
the social contexts of experienced cynicism and inadequacy
(Soini, Pyhältö, & Pietarinen, 2010; Pyhältö et al., 2011). The
STBI included the 9 items that had the highest reliability scores
and that best suited the social working environments within the
school1. The inventory consists of nine items measuring three
factors of socio-contextual teacher burnout: a) exhaustion (3
items), b) cynicism towards the teacher community (3 items),
and c) inadequacy in teacher-pupil interaction (3 items). All
items were rated on a 7-point Likert-type scale ranging from 1
(completely disagree) to 7 (completely agree) (excluding the
stress item that was rated on a 10-point scale). The final ver-
sion of the STBI items is shown in Table 2.
Control Variables for Analysing the Finnish
Teachers’ Soci o- Con textual Burnout Symptoms
The teachers’ educational background was measured by
asking participants to identify their professional qualifications,
i.e. whether they had class teacher, subject teacher, class and
subject teacher or special education teacher qualifications. In
Finland, all comprehensive school teachers hold a Master’s
degree either in educational sciences or some other domain
such as mathematics or biology, with compulsory additional
studies (35 credits) in educational science. Class teachers who
typically work (primary school) in grades (0) 1-6 have an MA
degree in educational science, with the main subject being ap-
plied educational sciences or e ucational psychology, while
1Due to the preliminary reliability analyses two exhaustion items with insuf-
ficient internal consistencies-one, cynicism towards the professional com-
munity, and two, inadequacy in the teacher-pupil interaction-were excluded
from the inventory. d

J. PIETARINEN ET AL.
Table 2.
The final version of the socio-contextual teacher burnout inventory (STBI) (translated from Finnish).
Please choose the alternative that best describes your current situation at work.
Completely disagree 1 2 3 4 5 6 Completely agree 7
1. I’m disappointed in our teacher community’s ways of handling our shared affairs. (CYN)
2. Dealing with problem situations considering my pupils often upsets me. (INAD)
3. I feel burnt out. (EXH)
4. I often feel like an outsider in my work community. (CYN)
5. The challenging pupils make me question my abilities as a teacher. (INAD)
6. With this work pace I don’t think I’ll make it to the retiring age. (EXH)
7. In spite of several efforts to develop the working habits of our teacher community they haven’t really changed. (CYN)
8. I often feel I have failed in my work with pupils. (INAD)
9. Stress means a situation in which a person feels tense, restless, nervous or anxious or is unable to sleep at
night because his/her mind is troubled all the time. Do you feel this kind of work-related stress? (EXH)*
Note: *For single stress item the scale is from one to ten: not at all 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 very much.
subject teachers who typically teach grades 7-9 (secondary
school) usually have an MA in some subject with an additional
compulsory one year of study in educational science. Special
education teachers who teach in both primary and secondary
schools in grades (0) 1-9 have an MA in educational science,
with the main subject being special education.
The socio-contextual factors of the teacher’s working envi-
ronments were measured with two items. The school size was
identified by asking about the number of pupils in the school
where the teacher currently worked. The classification of the
school size item included five categories: less than 100 pupils,
101 - 300 pupils, 301 - 450 pupils, 451 - 600 pupils and over
600 pupils. The academic level was measured by identifying the
grades that were taught in the school. Three academic levels
were categorised for further analysis: primary school (grades 0-
1-6), secondary school (grades 7-9) and combined school
(grades 0-1-9). Hence by using these control variables the
socio-contextual value of the developed teacher burnout invent-
tory was more specifically tested.
Analytic Strategy
In the first phase of analysis the structure and construct va-
lidity of the STBI were determined. Three alternative theoretic-
cal models were estimated separately (M1-M3). A three-factor
model (M1) assumed that three correlated latent factors, namely
exhaustion, cynicism towards the teacher community and in-
adequacy in teacher-pupil interaction, underlie the STBI items.
The M1 was extended to a second-order-factor model (M2) that
assumed one overall latent factor behind the three latent factors
identified in the M1 model. However, to have sufficient validity
and reliability, the scale of the second-order latent factor meas-
uring overall socio-contextual teacher burnout required rela-
tively high correlations between first-order factors (Sal-
mela-Aro et al., 2011). Finally, a one-factor model (M3) that
assumed there is one latent factor behind all nine items was also
tested. All three models (M1-M3) are data-equivalent and in-
clude the same number of estimated parameters (e.g. Feldt,
Leskinen, Kinnunen, & Ruoppila, 2003; Salmela-Aro et al.,
2011). All theoretical models are presented in Figure 1.
The validity and reliability of the items composing the STBI
were determined based on the confirmatory factor analysis. The
analyses were performed using an Mplus statistical package
(version 6.1; Muthén & Muthén, 1998-2010) and a miss-
ing-data method. This method uses all data that are available in
order to estimate the model without inputting data. Because the
socio-contextual teacher burnout sub-scales were skewed, the
parameters of the models were estimated using an MLR proce-
dure, which produces maximum likelihood estimates with
standard errors and 2 test statistics that are robust to non-nor-
mality (Muthén & Muthén, 1998-2010; Salmela-Aro, Kiuru,
Leskinen, & Nurmi, 2009). The goodness-of-fit of the esti-
mated standardised model was evaluated by an Chi-Square test,
the Comparative fit index (CFI), Tucker-Lewin Index (TLI),
Root Mean Square Error of Approximation (RMSEA), and
Standardised Root Mean Square Error of Approximation
(SRMR). A non-significant Chi-Square value, CFI and TLI
values above .95, a RMSEA value below .06 and SRMR value
below .08 indicated a good fit with the data (Muthén & Muthén,
1998-2010, Salmela-Aro et al., 2009). Finally, the model fit
was estimated with a relative goodness-of-fit Normed Fit Index
(NFI ≥ .95) that takes into account large sample sizes (Bentler
& Bonnet, 1980; Hu & Bentler, 1999).
Item reliability was measured by estimating the reliability
coefficients (i.e., the squared correlation between the item and
the factor, see Bollen, 1989; Liukkonen & Leskinen, 1999).
The structural validity of the items was measured by estimating
the standardised validity coefficients (i.e., standardised factor
loadings), which indicate direct structural relations between the
factor and the item (Bollen, 1989; Salmela-Aro et al., 2011).
The internal consistency of the inventory was examined by
estimating the factor score scale reliabilities (squared correla-
tions between the factor score scale and the latent factor) and
Cronbach’s alphas. Finally, on the basis of the three-factor
model the sum scores for each factor were constructed and
analysed in terms of teacher gender, educational background,
school size and academic level variables. The observed mean
differences between the groups were measured by t-test and
one-way analysis of variance.
Results
The main aim of the study was to determine the construct va-
76

J. PIETARINEN ET AL.
lidity of the developed STBI instrument. The means, variances
and the correlations are presented in Table 3. The validity and
reliability of the instrument were estimated with three exhaust-
tion items, three cynicism items and three inadequacy items.
As shown in Table 4, the three-factor primary model (M1)
fits the data. The three latent factors were correlated as ex-
pected [r (exhaustion with cynicism) = .45, p < .001: r (ex-
haustion with inadequacy) = .61, p < .001 and r (cynicism with
inadequacy) = .34, p < .001]. The goodness-of-fit criteria sug-
gested that the nine items, three-factor model described the
whole data, i.e. the model fit the class, subject and special edu-
cation teachers’ groups (see also Byrne, 1993; Hakanen et al.,
2006). The Normed Fit Index was exploiting to compensate for
the sensitivity of the Chi-Square test to a large sample size. In
addition, the NFI suggested that the model fit the data.
The second-order factor model (M2) fit the data with identi-
cal goodness-of-fit statistics compared with the M1 model.
Although all models (M1-M3) were data-equivalent, the one-
factor model (M3) did not sufficiently fit the data (see Table 4).
The results confirmed that teachers’ sense of work-related ex-
haustion was related to cynicism towards the teacher commu-
nity and inadequacy in the pupil-teacher relationship. Moreover,
teachers’ overall socio-contextual burnout could be measured
with three latent factors and nine items. On the basis of the
goodness-of-fit criteria, further reliability analyses were con-
ducted for the three-factor (M1) and second-order factor models
(M2).
The item reliability, standardised validity coefficients (i.e.
factor loadings) and internal consistency (i.e. Factor determina-
cies and Cronbach’s alphas) values are shown in Table 5. The
results confirmed that the items included in the final three-
factor model (M1) were valid indicators for latent factors, as
the factor loadings were ≥.60 and Cronbach’s alphas ≥.70
(Salmela-Aro et al., 2011). Moreover, the item reliability, stan-
dardised validity coefficients and internal consistency were also
sufficient for the second-order-factor model (M2), i.e. overall
socio-contextual teacher burnout.
The construct validity and reliability analyses confirmed that
the three-factor model (M1) was the most acceptable construct
for exploiting the developed instrument. The compact con-
text-sensitive teacher burnout inventory seemed to adequately
measure two essential social contexts in which the teachers’
work-related well-being is challenged. To analyse the socio-
contextual sensitivity of the developed inventory, the relations
to the teachers’ background information and work environment
variables, i.e. teachers’ qualification, gender, school size and
academic level, were analysed.
Finally, teachers’ gender, educational background, school
size and academic level were used as control variables for ana-
lysing the context sensitivity of the STBI. As shown in Table 6,
the STBI provides a potential instrument for analysing the
socio-contextual sources of teacher burnout, not only between
schools but also between the social working contexts within a
single school (Skaalvik & Skaalvik, 2010; Pyhältö et al., 2011).
The results suggest that teachers’ perceived exhaustion was
de-contextual rather than associated with the context-sensitive
burnout indicator (see Table 6). In terms of perceived exhaust-
tion, the differences between class, subject and special educa-
tion teachers or between academic levels were not statistically
significant. However, teachers working in large schools (over
600 pupils) were more likely to experience exhaustion than
those working in small schools (less than 100 pupils). These
results suggested that the structural characteristics of the school
are also risk factors for experiencing socio-contextual teacher
burnout symptoms.
The STBI was especially developed for identifying the socio-
contextual nature of teachers’ experienced cynicism and inade-
quacy at work. The results show that class teachers experienced
less cynicism towards the professional community than subject
and special education teachers did. Moreover, teachers who
worked in small school communities (less than 100 pupils) or
primary schools (grades (0-1-6) did not experience cynicism
towards professional communities to a similar extent as their
colleagues working in bigger schools and in more diverse pro-
fessional environments.
Teachers’ sense of inadequacy in teacher-pupil interaction
differed between the teacher groups and academic levels (see
Table 6). All teacher groups differed statistically significantly
from each other. Subject teachers had the highest level of ex-
perienced inadequacy in teacher-pupil interaction. In turn, spe-
cial education teachers reported the lowest levels of experi-
enced inadequacy. Further investigation showed that secondary
school teachers (grades 7-9) differed from both primary (gra-
des 1-6) and combined school teachers (grades 1-9) in terms of
experienced levels of inadequacy in teacher-pupil interaction.
School size was not found to be a significant factor in teachers’
perceived inadequacy concerning their work with pupils.
Teachers overall risk for burnout in terms of gender, educa-
tional background, school size and academic level was also
analysed. The results showed that there were not statistically
significant (p level ≤ .01) gender differences in teachers ex-
perienced overall burnout. However, the differences between
teacher groups remained. Subject teachers’ risk for overall
burnout was statistically significantly (p level ≤ .01) higher than
the class and special education teachers. Moreover, the differ-
ences in experienced burnout symptoms in terms of school size
and between academic levels increased (p level ≤ .01). Teach-
ers working in small schools (less than 100 pupils) experienced
less burnout symptoms than teachers operating in the medium
size (301 - 450 pupils) or big schools (over 600 pupils). The
academic level, i.e. working environment was also statistically
significant determinant ((p level ≤ .01) for teachers experienced
overall burnout symptoms. Teachers working in the secondary
school experienced more overall burnout symptoms than teach-
ers in other working environments.
Discussion
Limitations of the Study
The construct validity of the developed Socio-Contextual
Teacher Burnout Inventory was found to be sufficient. How-
ever, thus far the instrument has not been validated in other coun-
tries, school systems or teachers’ work environments. Further-
more, the data used in this study was cross-sectional. Therefore
additional construct validation of the STBI is needed. This re-
quires longitudinal research designs employed in different con-
texts, educational systems and cultures (Elo et al., 2003; Sal-
mela-Aro et al., 2011). Moreover, in further studies the concur-
rent validity (Salmela-Aro et al., 2009) of the identified three-
factor model should be tested with covariates other than the
work environment variables used in this study (e.g. gender and
educational background). For instance, the relation between indi-
vidual teacher’s perceived strategies for regulating their well-
being in different social contexts, (e.g. experienced self-efficacy

J. PIETARINEN ET AL.
78
Table 3.
Correlation matrix of socio-contextual teacher burnout inventory items, item means, and variances.
1. 2. 3. 4. 5. 6. 7. 8. 9.
1) EXH11 1.000
2) EXH12 .619 1.000
3) EXH13 .609 .658 1.000
4) CYN21 .237 .302 .129 1.000
5) CYN22 .253 .290 .147 .680 1.000
6) CYN23 .297 .275 .187 .468 .465 1.000
7) INAD31 .347 .343 .381 .164 .196 .213 1.000
8) INAD32 .322 .350 .281 .183 .195 .260 .537 1.000
9) INAD33 .377 .310 .444 .147 .170 .201 .477 .442 1.000
Means 4.69 2.28 3.00 3.22 3.36 2.34 3.46 2.10 2.90
Variances 6.06 2.52 4.07 3.18 2.87 2.67 3.37 1.56 2.43
Note: All the correlations were significant at p level < .001.
Table 4.
Goodness-of-fit criteria for the estimated models.
Chi-Square df p RMSEA SRMR CFI TLI NFI
M1: Three-factor model 156.37 24 .000 .05 .04 .98 .97 .97
M2: Second-order-factor model 156.37 24 .000 .05 .04 .98 .97 .97
M3: One-factor model 1903.17 27 .000 .17 .11 .67 .56 .67
Note: The modification indices (MI) were not used to improve the model fit.
Table 5.
Estimated item reliability, standardised validity coeffients, factor determinancies, and Cronbach’s alphas for the nine-item, three-factor, and corre-
sponding second-order factor for the Socio-Contextual Teacher Burnout Inventory.
Indicators Second-order-factor model
Three-factor model Overall burnout
Exhaustion Cynicism Inadequacy
1) Exhaustion11 .58 (.76) - - -
2) Exhaustion12 .66 (.82) - - -
3) Exhaustion13 .65 (.80) - - -
4) Cynicism21 - .67 (.82) - -
5) Cynicism22 - .67 (.82) - -
6) Cynicism23 - .34 (.58) - -
7) Inadequacy31 - - .54 (.73) -
8) Inadequacy32 - - .52 (.72) -
9) Inadequacy33 - - .41 (.64) -
EXHAUSTION - - - .81 (.90)
CYNICISM - - - .25 (.50)
INADEQUACY - - - .46 (.68)
Factor determinacy .92 .91 .88 .86
Cronbach’s α .81 .77 .74 .82
Note: A count without parenthesis refers to the estimated item reliability of an indicator, and a count in parenthesis to standardised validity coeffients, that is, the factor
loading of each indicator.

J. PIETARINEN ET AL.
Table 6.
Teacher gender, educational background, school size and academic level as control variables for analysing socio-contextual teacher burnout.
Control variables Exhaustion Cynicism Inadequacy
M SD p M SD p M SD p
Gender (t-test)
a) Female (n = 1878) 1.04 5.27 .01 8.91 4.25 ns 8.52 3.79 ns
b) Male (n = 429) 9.26 5.18 - 8.89 4.19 - 8.07 3.69 -
Educational background (one-way anova)
a) Class teacher (n = 815) 9.72 5.18 ns 8.16 4.30 .00bc 8.53 3.76 .01bc
b) Subject teacher (n = 729) 1.14 5.30 ns 9.30 4.04 .00a 9.09 3.85 .01ac
c) Special education teacher (n = 761) 9.80 5.29 ns 9.39 4.27 .00a 7.73 3.58 .00ab
School size (one-way anova)
a) Less than 100 pupils (n = 273) 9.22 5.10 ns 7.58 4.43 .00bcde 8.16 3.80 ns
b) 101 - 300 pupils (n = 836) 9.83 5.26 ns 8.82 4.26 ns 8.49 3.65 ns
c) 301 - 450 pupils (n = 659) 1.07 5.18 ns 9.30 4.20 ns 8.65 3.88 ns
d) 451 - 600 pupils (n = 289) 9.69 5.30 ns 9.15 4.12 ns 8.18 3.78 ns
e) over 600 pupils (n = 206) 1.83 5.48 .01a 9.47 3.98 ns 8.54 3.86 ns
Academic level (one-way anova)
a) Primary school (0-1-6), (n = 1014) 9.72 5.14 ns 8.45 4.35 .00bc 8.33 3.65 .01b
b) Secondary school (7-9), (n = 445) 1.24 540 ns 9.74 4.04 .01a 9.00 3.94 .01ac
c) Combined school (0-1-9), (n = 638) 9.86 5.32 ns 9.09 4.20 .01a 8.24 3.83 .01b
Notes: EXH (range 3 - 24), CYN and INAD (range 3 - 21) are sum variables constructed on the basis of the three-factor model. Mean differences were significant at p level
≤ .01.
in receiving and giving social support) and teacher’s experi-
enced socio-contextual burnout symptoms need to be re-
searched more deeply (Soini et al., 2010; Skaalvik & Skaalvik,
2010).
The study’s response rate was moderate. However, the rep-
resentativeness of the sample was plausible. Previous studies
(Krosnick, 1999; Cook, Heath, & Thompson, 2000) have
shown that the representativeness of samples is a much more
important criterion for evaluating the validity of a study than
the response rate, particularly if a probability sampling method
is used. In this study the sample was not biased in terms of the
teachers’ perceived socio-contextual burnout. More specifically,
teachers who experienced high levels exhaustion, cynicism
towards the teacher community and inadequacy in the teacher-
pupil interaction also responded to the questionnaire.
Conclusion
The study aimed at developing a context-sensitive instrument
to study teacher burnout by testing the validity and reliability of
the Socio-Contextual Teacher Burnout Inventory with a sample
of Finnish teachers. Moreover, the study took a social and in-
terpersonal approach to teacher burnout (Masclach, 2003;
Buunk & Schaufeli, 1993; Geurts et al., 1998; Schaufeli et al.,
1996; Xanthopoulou et al., 2007). Our main hypothesis was
that teacher burnout is situated primarily in social interactions
within the school community (Pyhältö et al., 2011). Therefore
the sources of teacher burnout may vary, and not only between
schools but also between the social working contexts within a
single school (Skaalvik & Skaalvik, 2009; Fernet et al., 2012).
The results showed that the correlated three-factor solution
and second-order-factor solution fitted the data. Hence the re-
sults supported hypotheses 1 and 2 by showing that teachers’
experienced exhaustion, cynicism towards the teacher commu-
nity and inadequacy in the pupil-teacher relationship were
closely related but separate constructs. The findings indicate
that these components are ingredients of teacher’s overall
socio-contextual burnout. The findings are also in accordance
with those of previous work-related teacher burnout studies
(Byrne, 1993; Maslach & Leiter, 1999). Moreover, the results
supported the main hypothesis that teacher burnout can be ex-
amined in terms of interpersonal problems in an individual’s
relations with others in the workplace (Buunk et al., 2001; Fer-
net et al., 2012).
This study focused on developing measures for exploring
teacher burnout in terms of primary social working environ-
ments i.e. teacher-pupil and professional community interact-
tions within the school (Soini et al., 2010). In further studies the
socio-contextual components of the teacher burnout invent-
tory could be expanded to multiple social contexts which may

J. PIETARINEN ET AL.
also play a role in teacher burnout including collaboration with
parents or other stakeholders in the educational system (e.g.
Pillay, Goddard, & Wilss, 2005; Milfont et al., 2008). More
detailed research evidence concerning the significance and
emphasis of different social contexts with respect to teacher
exhaustion, cynicism and inadequacy is also needed (Pyhältö et
al., 2011).
In line with previous studies, the results showed that female
teachers’ were more likely to experience exhaustion than their
male colleagues (Tatar & Horenczyk, 2003; Antoniou et al.,
2006). However, no gender differences were noted in terms of
perceived inadequacy in teacher-pupil interaction or cynicism
towards the professional community. Further, class, subject and
special education teachers’ experienced inadequacy in teacher-
pupil interaction and cynicism towards the professional com-
munity differed from each other. The academic level, moreover,
at which teachers work was found to affect the perceived in-
adequacy and cynicism. These results supported hypothesis 3
by indicating that partly specified pedagogical tasks and educa-
tional expertise at different academic levels (i.e. class, subject
and special education qualification) contribute to teacher burn-
out in the social settings of a school.
The developed inventory can be utilised in the field of school
development and teacher learning research. Previous studies
have suggested that perceived work-related well-being is a pre-
condition for teachers’ learning and sense of agency in terms of
educational innovations (Vermunt & Endedijk, 2011; Soini et
al., 2010; Pyhältö, Pietarinen, & Soini, 2012). Hence the STBI
could be used as a measure to explore the relationship between
socio-contextual burnout and teachers’ learning. Moreover, for
teacher communities the STBI provides a tool for identifying
and analysing the significance of varying social contexts to
teachers’ perceived work-related well-being (see Table 2).
Finally, the complexity of gradually developing teacher
burnout in multiple social contexts has not been studied exten-
sively. The instrument introduced in this study has the potential
to contribute to interpersonal teacher burnout research in the
future (see also Fernet et al., 2012). Moreover, it seems that
reducing socio-contextual burnout symptoms, i.e. teacher cyni-
cism in terms of professional community and inadequacy in the
teacher-pupil interaction, requires context-sensitive and adap-
tive strategies adopted by teachers that enable not only the ac-
cess to the resources of the community (Schaufeli & Bakker,
2004) but also have potential to contribute to further devel-
opment of the environment.
Acknowledgements
This research is funded by the Ministry of Education and
Culture and Academy of Finland (research project: 259489).
REFERENCES
Adams, J. (1965). Inequity in social exchange. In L. Berkowitz (Ed.),
Advances in experimental social psychology (Vol. 2, pp. 267-299).
New York: Academic press.
Antoniou, A.-S., Polychroni, F., & Vlachakis, A.-N. (2006). Gender
and age differences in occupational stress and professional burnout
between primary and high-school teachers in Greece. Journal of
Managerial Psychology, 21, 682-669.
doi:10.1108/02683940610690213
Bakker, A. B., Demerouti, E., & Schaufeli, W. B. (2002). Validation of
the maslach burnout inventory—General survey: An internet study.
Anxiety, Stress & Coping, 15, 245-260.
doi:10.1080/1061580021000020716
Bakker, A. B., Schaufeli, W. B., Leiter, M. P., & Taris, T. W. (2008).
Work engagement: An emerging concept in occupational health psy-
chology. Work & Stress, 22, 187-200.
doi:10.1080/02678370802393649
Bentler, P. M., & Bonett, D. G. (1980). Significance tests and goodness
of fit in the analysis of covariance structures. Psychological Bulletin,
88, 588-606. doi:10.1037/0033-2909.88.3.588
Bollen, K. A. (1989). Structural equations with latent variables. New
York: Wiley.
Borg, M. G., & Riding, R. J. (1991). Towards a model for the determi-
nants of occupational stress among schoolteachers. European Jour-
nal of Psychology of Education , 6, 355-373.
doi:10.1007/BF03172771
Brouwers, A., & Tomic, W. (2000). A longitudinal study of teacher
burnout and perceived self-efficacy in classroom management.
Teaching and Teacher Education, 16, 239-253.
doi:10.1016/S0742-051X(99)00057-8
Buunk, B., Ybema, J., Gibbons, F., & Ipenburg, M. (2001). The affec-
tive consequences of social comparison as related to professional
burnout and social comparison orientation. European Journal of So-
cial Psychology, 31, 337-351. doi:10.1002/ejsp.41
Buunk, B. P., & Schaufeli, W. B. (1993). Professional burnout: A per-
spective from social comparison theory. In W. B. Schaufeli, C.
Maslach, & T. Mareks (Eds.) Professional burnout: Recent devel-
opments in theory and research (pp. 53-69). Washington DC: Taylor
& Francis.
Byrne, B. (1993). The Maslach burnout inventory: Testing for factorial
validity and invariance across elementary, intermediate and secon-
dary teachers. Journal of Occupational and Organizational Psy-
chology, 66, 197-212. doi:10.1111/j.2044-8325.1993.tb00532.x
Cano-Garcia, F. J., Padilla-Muñoz, E. M., & Carrasc o-Ortiz, M. A.
(2005). Personality and contextual variables in teacher burnout. Per-
sonality and Individual Differences, 38, 929-940.
doi:10.1016/j.paid.2004.06.018
Cook, C., Heath, F., & Thompson, R. L. (2000). A meta-analysis of
response rates in web- or internet-based surveys. Educational and
Psychological Measurement, 6 0 , 821-836.
doi:10.1177/00131640021970934
Devos, C., Dupriez, V., & Paquay, L. (2012). Does the social working
environment predict beginning teachers’ self-efficacy and feelings of
depression? Teaching and Teacher Educat ion, 28, 206-217.
doi:10.1016/j.tate.2011.09.008
Dorman, J. P. (2003). Relationship between school and classroom en-
vironment and teacher burnout: A LISREL analysis. Social Psychol-
ogy of Education, 6, 107-127. doi:10.1023/A:1023296126723
Elo, A.-L., Leppänen, A., & Jahkola, A. (2003). Validity of a single-
item measure of stress symptoms. Scandinavian Journal of Work,
Environment & Health, 29, 444-451. doi:10.5271/sjweh.752
Embich, J. L. (2001). The relationship of secondary special education
teachers’ roles and factors that lead to professional burnout. Teacher
Education and Special E duc ation, 24, 58-69.
doi:10.1177/088840640102400109
Feldt, T., Leskinen, E., Kinnunen, U., & Ruoppila, I. (2003). The sta-
bility of sense of coherence: comparing two age groups over a 5-year
follow-up study. Personality and Individual Differences, 35, 1151-
1165. doi:10.1016/S0191-8869(02)00325-2
Fernet, C., Guay, F., Senécal, C., & Austin, S. (2012). Predicting intra-
individual changes in teacher burnout: The role of perceived school
environment and motivational factors. Teaching and Teacher Educa-
tion, 28, 514-525. doi:10.1016/j.tate.2011.11.013
Festinger, L. (1954). A theory of social comparison processes. Human
Relations, 7, 117-140. doi:10.1177/001872675400700202
Freudenberger, H. J. (1974). Staff burn-out. Journal of Social Issues, 30,
159-165. doi:10.1111/j.1540-4560.1974.tb00706.x
Geurts, S., Schaufeli, W., & De Jonge, J. (1998). Burnout and intention
to leave among mental health-care professionals: A social psycho-
logical approach. Journal of Social and Clinical Psychology, 17,
341-362. doi:10.1521/jscp.1998.17.3.341
80

J. PIETARINEN ET AL.
Hakanen, J., Bakker, A., & Schaufeli, W. (2006). Burnout and engage-
ment among teachers. Journal of School Psychology, 43, 495-513.
doi:10.1016/j.jsp.2005.11.001
Holland, R. P. (1982). Special educator burnout. Educational Horizons,
60, 58-64.
Hu, L., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in co-
variance structure analysis: Conventional criteria versus new alterna-
tives. Structural Equation Modeling, 6, 1-55.
doi:10.1080/10705519909540118
Jepson, E., & Forrest, S. (2006). Individual contributory factors in
teacher stress: The role of achievement striving and occupational
commitment. British Journal of Educational Psychology, 76, 183-
197. doi:10.1348/000709905X37299
Kalimo, R., & Hakanen, J. (2000). Työuupumus. In T. Kauppinen, P.
Heikkilä, S. Lehtinen, K. Lindström, S. Näyhä, A. Seppälä, J. Toik-
kanen, & A. Tossavainen (Eds.), Työ ja terveys Suomessa v.200. (pp.
119-126). Helsinki: Työterveyslaitos.
Kinnunen, U., Parkatti, T., & Rasku, A. (1994). Occupational well-
being among aging teachers in Finland. Scandinavian Journal of
Educational Research, 38, 315-332. doi:10.1080/0031383940380312
Klassen, R. M. (2010). Teacher stress: The mediating role of collective
efficacy beliefs. The Journal of Educational Research, 103, 342-350.
doi:10.1080/00220670903383069
Kokkinos, C. M. (2007). Job stressors, personality and burnout in pri-
mary school teachers. British Journal of Educational Psychology, 77,
229-243. doi:10.1348/000709905X90344
Krosnick, J. A. (1999). Survey research. Annual Review of Psychology,
50, 537-567. doi:10.1146/annurev.psych.50.1.537
Lau, P. S. Y., Yuen, M. T., & Chan, R. M. C. (2005). Do demographic
characteristics make a difference to burnout among Hong Kong sec-
ondary school teachers? Social Indicators Research, 71, 491-516.
doi:10.1007/1-4020-3602-7_17
Lee, R., & Ashforth, B. (1996). A meta-analytic examination of the
correlates of the three dimensions of job burnout. Journal of Applied
Psychology, 81, 123-133. doi:10.1037/0021-9010.81.2.123
Leiter, M., & Maslach, C. (1988). The impact of interpersonal environ-
ment on burnout and organizational commitment. Journal of Organ-
izational Behavior, 9, 297-308. doi:10.1002/job.4030090402
Leung, D. Y. P., & Lee, W. W. S. (2006). Predicting intention to quit
among Chinese teachers: Differential predictability of the component
of burnout. Anxiety, Stress, & Coping, 19, 129-141.
doi:10.1080/10615800600565476
Liukkonen, J., & Leskinen, E. (1999). The reliability and validity of
scores from the children’s version of the perception of success ques-
tionnaire. Educational and Psychological Measurement, 59, 651-664.
doi:10.1177/00131649921970080
Loonstra, B., Brouwers, A., & Tomic, W. (2009). Feelings of existen-
tial fulfillment and burnout among secondary school teachers. Teach-
ing and Teacher Education, 25, 752-757.
doi:10.1016/j.tate.2009.01.002
Maslach, C. (2003). Job burnout: New directions in research and inter-
ventions. Current Di r e ctions in Psychological Science, 12, 189-192.
doi:10.1111/1467-8721.01258
Maslach, C., & Jackson, S. E. (1981). The measurement of experienced
burnout. Journa l of Occupational Behavior, 2, 99-113.
doi:10.1002/job.4030020205
Maslach, C., & Leiter, M. P. (2005). Banishing burnout: Six strategies
for improving your relationship with work. San Francisco: Jossey-
Bass.
Maslach, C., Schaufeli, W., & Leiter, P. (2001). Job burnout: New
directions in research and intervention. Current Directions in Psy-
chological Science, 12, 189-192. doi:10.1111/1467-8721.01258
Maslach, C., & Leiter, M. (1999). Teacher burnout: A research agenda.
In R. Vandenberg & A. Huberman (Eds.), Understanding and pre-
venting teacher burnout (pp. 295-303). Cambridge: Cambridge Uni-
versity Press.
Milfont, T. L. Denny, S., Ameratunga, S., Robinson, E., & Merry, S.
(2008). Burnout and wellbeing: Testing the Copenhagen burnout in-
ventory in New Zealand teachers, Social Indicators Research, 89,
169-177. doi:10.1007/s11205-007-9229-9
Montgomery, C., & Rupp, A. A. (2005). A meta-analysis for exploring
the diverse causes and effects of stress in teachers. Canadian Journal
of Education, 28, 458-486. doi:10.2307/4126479
Muthén, L., & Muthén, B. O. (1998-2010). Mplus users guide (6th ed.).
Los Angeles, CA: Muthén & Muthén.
National Board of Education (2010). Opettajat Suomessa [Teachers in
Finland]. URL (last checked 3 April 2012).
http://www.oph.fi/julkaisut/2011/opettajat_suomessa_201.
Peeters, M., & Rutte, C. G. (2005). Time management behavior as a
moderator for the job demand-control. Interaction Journal of Occu-
pational Health Psychology, 10, 64-75.
doi:10.1037/1076-8998.10.1.64
Pillay, H., Goddard, R., & Wilss, L. (2005). Well-being, burnout and
competence: Implications for teachers. Australian Journal of Tea-
cher Education, 30, 22-33.
Pyhältö, K., Pietarinen, J., & Salmela-Aro, K. (2011). Teacher-working
environment fit as a framework for burnout experienced by Finnish
teachers. Teaching and Teacher Education, 27, 1101-1111.
doi:10.1016/j.tate.2011.05.006
Pyhältö, K., Pietarinen, J., & Soini, T. (2012). Do comprehensive
school teachers perceive themselves as active agents in school re-
forms? Journal of Educational Change, 13, 95-116.
doi:10.1007/s10833-011-9171-0
Rudow, B. (1999). Stress and burnout in the teaching profession: Euro-
pean studies, issues, and research perspectives. In R. Vanderbergue,
& M. A. Huberman (Eds.), Understanding and preventing teacher
burnout: A source book of international practice and research (pp.
38-58). Cambridge: Cambridge University Press.
doi:10.1017/CBO9780511527784.004
Salmela-Aro, K., Kiuru, N., Leskinen, E., & Nurmi, J. (2009). School
burnout inventory: Reliability and validity. European Journal of
Psychological Assessment, 25, 48-57.
doi:10.1027/1015-5759.25.1.48
almela-Aro, K., Rantanen, J., Hyvönen, K., Tilleman, K., & Feldt, T.
(2011). Bergen burnout inventory: Reliability and validity among
Finnish and Estonian managers. International A
tional and Environmental He al th ,
S
rchives of Occupa-
84, 635-645.
doi:10.1007/s00420-010-0594-3
antavirta, N., Solovieva, S., & Theorell, T. (2007). The association
between job strain and emotional exhaustion in a cohort of 1028 Fin-
nish teachers. British Journal of Educational P
S
sychology, 77, 213-
228. doi:10.1348/000709905X92045
chaufeli, W. B., & Bakker, A. B. (2004). Job demands, job resources,
and their relationship with burnout and engagement: A mu
study. Journal of Org
S
lti-sample
anizational Behavior, 25, 293-437.
doi:10.1002/job.248
chaufeli, W. B., & Buunk, B. P. (2003). Burnout: An overview of 25
years of research in theorizing. In M. J. Winnubst, & C. L. Cooper
(Eds.), The handboo
S
k of work and health psychology (pp. 383-425).
S study
S
odel. Work &
Chichester: Wiley.
chaufeli, W., & Enzmann, D. (1998). The burnout companion to
and practice: A critical analy s i s. London: Taylor and Francis.
chaufeli, W., van Dierendonck, D., & van Gorp, K. (1996). Burnout
and reciprocity: Towards a dual-level social exhange m
Stress, 10, 225-237. doi:10.1080/02678379608256802
harplin, E., O’Neill, M., & Chapman, A. (2011). Coping strategies for
adaptation to new teacher appointments: Interve
Teaching and Teacher Educatio
S
ntion for retention.
n, 27, 136-146.
doi:10.1016/j.tate.2010.07.010
kaalvikS
nd job satisfaction. Teaching and
Sd Teacher Education,
, E. M., & Skaalvik, S. (2009). Does school context matter?
Relations with teacher burnout a
Teacher Education, 25, 518-524.
kaalvik, E. M., & Skaalvik, S. (2010). Teacher self-efficacy and tea-
cher burnout: A study of relations. Teaching an
26, 1059-1069. doi:10.1016/j.tate.2008.12.006
oini,T., Pyhältö, K., & Pietarinen, J. (2010). Pedagogical well-being—
Reflecting learning and well-being in teache
S
rs’ work. Teaching and
Tut among
ion, 19, 397-408.
teachers: Theory and practic e , 16, 765-782.
atar, M., & Horenczyk, G. (2003). Diversity-related burno
teachers. Teaching and Teacher E ducat
doi:10.1016/S0742-051X(03)00024-6
ravers, C. J., & Cooper, C. L. (1993). Mental health, job satisfaction T

J. PIETARINEN ET AL.
82
ers. Work and Stress, 7, 203- and occupational stress among UK teach
219. doi:10.1080/02678379308257062
ermunt, J. D., & Endedijk, M. D. (2011). Patterns in teacher learning
in different phases of the professional
Vdivid-career. Learning and In
ual Differences, 21, 294-302. doi:10.1016/j.lindif.2010.11.019
olters, C. A., & Daugherty, S. G. (2007). Goal structures and teach-
ers’ sense of efficacy: Their relation and association to teaching ex-
perience and academic level. Journal of Ed
W
ucational Psychology, 99,
181-193. doi:10.1037/0022-0663.99.1.181
anthopoulou, D., Bakker, A. B., Demerouti, E., & Schaufeli, W. B.
(2007). The role of personal resources in the job demands-resource
model. International Journal of St
X
s
ress Management, 14, 121-141.
doi:10.1037/1072-5245.14.2.121