Professional Documents
Culture Documents
Reliability and Validity of The Student Version Of.4
Reliability and Validity of The Student Version Of.4
Allison Smith, PT, DPT, PhD, Jennifer Ellison, PT, PhD, Jennifer Bogardus, PT, PhD, and Peggy Gleeson, PT, PhD
WnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8K2+Ya6H515kE= on 10/13/2023
Copyright © 2022 Academy of Physical Therapy Education, APTA Unauthorized reproduction of this article is prohibited.
Table 1. Demographic Information to use and includes both negatively and posi-
tively worded items. Although the OLBI-S has
Total (N = 45 [100%]) been used with university students in Greece
and Germany, it has not yet been validated in PT
Gender, n (%)
students in the United States.
Female 39 (86.7) e purpose of this study was to assess the
Male 6 (13.3) test–retest reliability of the OLBI-S, internal
consistency of the OLBI-S, and convergent
Age group, n (%)
Downloaded from http://journals.lww.com/jopte by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1A
Asian 6 (13.3)
Black or African American 1 (2.2) SUBJECTS
White 37 (82.2)
An a priori power analysis for intraclass corre-
lation coefficients (ICCs) with power = .8, a level
Other 1 (2.2) = .05, and ICC value = .7 revealed that a sample
Ethnicity, n (%) size of 10 participants was needed.31 Participants
included a convenience sample of DPT students
Hispanic or Latino or Spanish Origin 7 (15.6)
attending Texas Woman's University in Hous-
Not Hispanic or Latino or Spanish Origin 36 (84.4) ton during the fall semester of 2020.
the literature on burnout in physical thera- outcome measure created in 1981 that was METHODS
pists (PTs) is far more limited. Existing studies built on the concept that burnout is a mul- Institutional Review Board (IRB) approval was
have been conducted with PTs in Poland,8,12-15 tidimensional construct that involves 3 obtained from Texas Woman’s University be-
Italy,16 Spain,17 Australia,18 and the United distinct but related aspects: emotional ex- fore the initiation of this study (IRB-FY2020-
States.19-24 ough most of the studies con- haustion, depersonalization, and reduced 390). Students from all 3 cohorts were
ducted in the United States were published personal accomplishment.27 Recently, the recruited via email. e recruitment email in-
in the 1980s and 1990s, a 2015 platform pre- MBI has been criticized due to issues with cluded information about the study as well as
sentation at the World Confederation for the measure including its 3 factor structure, contact information for the principal in-
Physical erapy Congress contained results of the unidirectional wording of questions, the vestigator and the Texas Woman’s University
a study conducted using a stratified sample of inconsistencies with cutoff scores, and var- IRB if they had any study-related questions.
6,500 PT members of the APTA.24 From the iable burnout definitions.2,28 e MBI is Each recruitment email included a statement
sample, 1,366 PTs responded, and 29% were also protected by copyright and distributed that indicated that the completion of the sur-
found to have self-reported high emotional by a commercial publisher at a cost, whereas veys constituted the participant’s informed
exhaustion, 15% had high perceived stress, and other burnout measures are free to use. consent to act as a participant in this research.
13% were considered to have burnout (defined as Several outcome measures have been de- e outcome measures used were the
high emotion exhaustion, high depersonalization, veloped in response to the criticisms and OLBI-S and MBI-GSS. e OLBI-S has 2
and low personal accomplishment). Only 2 psychometric limitations of the MBI, in- subscales: exhaustion and disengagement.
studies have been identified that included PT cluding the Oldenburg Burnout Inventory Each eight-item subscale has 4 positively and
students, both of which found an increase in as- (OLBI).1 e OLBI measures feelings of ex- 4 negatively worded questions, and after re-
pects of burnout over the course of a semes- haustion and disengagement from work and verse scoring negative items, scores for the 8
ter.25,26 Physical therapist student burnout is includes both negatively and positively wor- items for each subscale are averaged together.
difficult to address when the prevalence is un- ded items for each dimension.29 e OLBI e OLBI-S is scored on a 5-point Likert-type
known and few outcome measures exist to study was translated into English in 2005 by Hal- scale from strongly agree (1) to strongly dis-
burnout in students. besleben and Demerouti,1 who also estab- agree (5), with higher scores indicating a
Most authors studying burnout in all lished construct validity of the English higher level of burnout. Both the exhaustion
populations have used some version of the version. A student version of the OLBI (OLBI-S) (Cronbach’s a = .87) and the disengagement
Maslach Burnout Inventory (MBI), an was developed by Reis et al.30 e OLBI-S is free (Cronbach’s a = .81) subscales were found to
Table 2. Mean Values and SDs of MBI-GSS and OLBI-S Subscale Scores
Abbreviations: MLBI-GSS = Maslach Burnout Inventory General Survey for Students; OLBI-S = Oldenburg Burnout Inventory for Students.
Copyright © 2022 Academy of Physical Therapy Education, APTA Unauthorized reproduction of this article is prohibited.
Table 3. Correlations Between MBI-GSS and OLBI-S Subscales
Abbreviations: MLBI-GSS = Maslach Burnout Inventory General Survey for Students; OLBI-S = Oldenburg Burnout Inventory for Students.
a
Correlation is significant at the .01 level (2-tailed).
WnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8K2+Ya6H515kE= on 10/13/2023
be reliable.30 e MBI-GSS is a 16-item mea- disengagement).5 For this reason, the pro- confidence interval from .843 to .955 (F(41) =
sure with 3 subscales: exhaustion (5 items), fessional efficacy subscale of the MBI-GSS was 11.638, P < .001). e ICC for the OLBI-S
cynicism (5 items), and professional efficacy (6 not used in this study. disengagement subscale was .955 with a 95%
items).32 It is graded on a 7-point Likert-type confidence interval from .916 to .976 (F(41) =
scale from never (0) to every day (6). e MBI is 21.669, P < .001). Bland–Altman plots were
considered the standard tool for burnout re- RESULTS constructed for both the OLBI-S exhaustion
search and the student version has been vali- Forty-five students fully completed the first subscale and the OLBI-S disengagement
dated in many populations.33-36 survey. Most students who completed the subscale (Figures 1 and 2). e plots did not
Students were asked to complete the OLBI- survey were White (82.2%), not Hispanic or appear to show proportional bias because
S and MBI-GSS via a Psychdata survey. Stu- Latino or Spanish origin (84.4%), female most of the points fell between the 2 confi-
dents who completed the first study were asked (86.7%), and less than 25 years old (66.7%). dence intervals and did not appear to follow a
to complete the OLBI-S a second time 1 week Demographic data are summarized in Table 1. trend. is was confirmed with a follow-up
later. Intraclass correlation coefficients (two- Mean values and SDs for the OLBI-S and linear regression analysis using the difference
way random effects model, absolute agree- MBI-GSS can be found in Table 2 (average and mean scores for each subscale. e linear
ment) were calculated to examine test–retest scores are reported). Cronbach’s a was cal- regression analysis was not significant for the
reliability, and Bland–Altman plots were con- culated for each OLBI-S subscale, and the OLBI-S exhaustion subscale (b = .007, P =
structed to assess level of agreement by plotting internal consistency was found to be good .940) or the OLBI-S disengagement subscale
individual differences against individual mean for both the exhaustion subscale (Cronbach’s (b = .078, P = .249).
scores. e significance level was set at .05. a = .833) and the disengagement subscale
Cronbach’s a was calculated for each OLBI-S (Cronbach’s a = .784). Convergent validity
subscale to assess internal consistency. Con- was assessed using Pearson’s correlation co- DISCUSSION
vergent validity was assessed by calculating efficient and was found to be good between
Pearson’s correlations comparing the exhaus- the exhaustion subscales for the OLBI-S and e results of this study indicate that the
tion subscales for the OLBI-S and MBI-GSS MBI-GSS (r = .741, P < .001) as well as the OLBI-S is a reliable and valid outcome mea-
and comparing the disengagement subscale of disengagement subscale of the OLBI-S and sure to assess burnout in PT students. Pre-
the OLBI-S and cynicism subscale of the MBI- cynicism subscale of the MBI-GSS (r = .766, vious studies have validated the English and
GSS. Qiao and Schaufeli give empirical, theo- P < .001) (Table 3). student versions of the OLBI using factor
retical, clinical, and psychometric evidence that Forty-two of 45 students completed the analysis and multitrait, multimethod matrix
the personal accomplishment/professional ef- second survey. Test–retest reliability was analysis.1,30,37 Demerouti et al.37 reported that
ficacy factor should not be included as part of found to be good for both the OLBI-S ex- the correlation between the MBI General
the burnout construct and agree with previous haustion subscale and the OLBI-S disen- Survey (MBI-GS) exhaustion subscale and the
authors that the core of burnout includes a mix gagement subscale. e ICC for the OLBI-S OLBI exhaustion subscale in Greek em-
of 2 factors (exhaustion and cynicism/ exhaustion subscale was .916 with a 95% ployees was r = .60, whereas we found a larger
value of the correlation between the MBI-GSS
Figure 1. Bland–Altman plot for the Oldenburg Burnout Inventory for Students exhaustion subscale and OLBI-S exhaustion
exhaustion subscale subscale of r = .74. e same study reported
that the correlation between the MBI-GS
cynicism subscale and OLBI disengagement
subscale was r = .60, whereas we found a
larger value of the correlation between the
MBI-GSS cynicism subscale and OLBI-S dis-
engagement subscale of r = .77. Halbesleben
et al.1 reported test–retest reliability of the
OLBI for employees as r = .51 for the ex-
haustion subscale and r = .32 for the disen-
gagement subscale, with 4 months separating
test 1 from test 2. Our study used a shorter
time frame between test 1 and test 2 (1 week)
and found stronger test–retest reliability us-
ing ICC, with an ICC = .916 for the exhaustion
Copyright © 2022 Academy of Physical Therapy Education, APTA Unauthorized reproduction of this article is prohibited.
Figure 2. Bland–Altman plot for the Oldenburg Burnout Inventory for Students healthcare students. Can Med Educ J. 2017;8:
disengagement subscale 90-108.
11. Dyrbye LN, West CP, Satele D, Boone S, Sloan
J, Shanafelt TD. A national study of medical
students’ attitudes toward self-prescribing and
responsibility to report impaired colleagues.
Acad Med. 2015;90:485-493.
12. Pustułka-piwnik U, Ryn ZJ. Burnout syn-
drome in physical therapists: Demographic
Downloaded from http://journals.lww.com/jopte by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1A
Copyright © 2022 Academy of Physical Therapy Education, APTA Unauthorized reproduction of this article is prohibited.
26. Williams P, Mueller K, Carroll H, Cornwall M, Oldenburg Burnout Inventory (OLBI): Fac- Factorial validity, invariance, and latent profiles
Denney L, Kroneberger L. Patterns of aca- torial invariance across samples and countries. of the Italian Version of the Maslach Burnout
demic burnout, emotional distress, and coping Burn Res. 2015;2:8-18. Inventory Student Survey (MBI-SS). Front
in physical therapy students. Int J Heal Well- 31. Bujang MA, Baharum N. A simplified guide to Psychol. 2018;9:1-9.
ness Soc. 2018;8:31-46. determination of sample size requirements for 35. Rostami Z, Abedi M, Schaufeli W, Ahmadi S,
27. Maslach C, Jackson SE. e measurement of estimating the value of intraclass correlation Sadeghi A. e psychometric characteristics of
experienced burnout. J Occup Behav. 1981;2: coefficient: A review. Arch Orofac Sci. 2017;12: Maslach Burnout Inventory Student Survey
99-113. 1-11. among students of Isfahan University. Zahe-
28. Rotenstein LS, Torre M, Ramos MA, et al. 32. Maslach C, Jackson SE, Leiter MP, Schaufeli dan J Res Med Sci. 2013;15:29-33p.
Downloaded from http://journals.lww.com/jopte by BhDMf5ePHKav1zEoum1tQfN4a+kJLhEZgbsIHo4XMi0hCywCX1A
Prevalence of burnout among physicians: A WB. MBI: General Survey for Students. Menlo 36. Schaufeli WB, Martı́nez IM, Pinto AM,
systematic review. J Am Med Assoc. 2018;320: Park, CA: Mind Garden; 2019. https://www. Salanova M, Barker AB. Burnout and en-
1131-1150. mindgarden.com/313-mbi-general-survey- gagement in university students a cross-
29. Demerouti E, Bakker A, Nachreiner F, for-students. Accessed July 17, 2019. national study. J Cross Cult Psychol. 2002;33:
WnYQp/IlQrHD3i3D0OdRyi7TvSFl4Cf3VC4/OAVpDDa8K2+Ya6H515kE= on 10/13/2023
Ebbinghaus M. From mental strain to burn- 33. Hu Q, Schaufeli WB. e factorial validity of 464-481.
out. Eur J Work Organ Psychol. 2002;11: the Maslach Burnout Inventory-Student Sur- 37. Demerouti E, Bakker AB, Vardakou I, Kantas
423-441. vey in China. Psychol Rep. 2009;105:394-408. A. e convergent validity of two burnout
30. Reis D, Xanthopoulou D, Tsaousis I. Measur- 34. Portoghese I, Leiter MP, Maslach C, et al. instruments: A multitrait-multimethod anal-
ing job and academic burnout with the Measuring burnout among university students: ysis. Eur J Psychol Assess. 2003;19:12-23.
Copyright © 2022 Academy of Physical Therapy Education, APTA Unauthorized reproduction of this article is prohibited.