Oldenburg Burnout Inventory (OLBI)

Structural validity Moderatecontestedevidence form: mixed

indirect German, Greek, US, Chinese, Polish and Brazil/Portugal samples, no dedicated UK working-adult structural study located

The intended two-factor structure (exhaustion, disengagement) has been supported by confirmatory factor analysis in the English validation across two US samples (Halbesleben J. 2005) and in the original Greek convergent-validity study (Demerouti E. 2003), and across German employees and students with partial invariance (Reis D. 2015). However, the separability of the two factors is contested: a Chinese nurse validation found a very high inter-factor relationship and reported strong criterion overlap with the MBI (Xu H. 2021), a Brazil/Portugal study reduced the instrument to 15 items and noted a high disengagement-exhaustion correlation (Sinval J. 2019), and a within/between-person multilevel study of daily diary data concluded a unidimensional model was more defensible because of the strong exhaustion-disengagement correlation and low within-person reliability of disengagement (Gruszczynska E. 2021). The mixed item wording generates method effects: the Polish validation found each of the two components split into positively and negatively worded sub-clusters, distorting the intended two-factor solution (Baka Ł. 2016). This wording-method controversy is the central structural issue for the OLBI.

Sub-grades (evidence differs by subgroup):

• {"subgroup": "Two-factor solution", "grade": "Moderate", "note": "supported in several samples but with high inter-factor correlation"}
• {"subgroup": "Wording-method effects", "grade": "Low", "note": "positively vs negatively worded items form method sub-factors in some samples"}

Convergent and discriminant validity Moderatewell-establishedevidence form: mixed

indirect Greek, US, Chinese, Polish and Brazil/Portugal samples

OLBI exhaustion and disengagement correlate strongly with the corresponding MBI-GS scales: multitrait-multimethod analyses in the original Greek study supported convergent and discriminant validity against the MBI-GS (Demerouti E. 2003), the English validation positioned the OLBI as a viable alternative to the MBI-GS on MTMM grounds (Halbesleben J. 2005), and a Chinese validation reported a very high correlation with the MBI (r = 0.87 for the total scale) (Xu H. 2021). The scales relate positively to perceived stress and negatively to work engagement (Baka Ł. 2016; Sinval J. 2019). The very high MBI correlations support convergent validity but also raise the question of what the OLBI adds beyond existing measures.

Criterion validity: reference standard Absent (a finding about the literature)untestedevidence form: canonical

indirect no reference-standard study exists and no diagnostic standard is available

There is no diagnostic gold standard for burnout, and no validation of the OLBI against a clinical or health reference standard was located; the COSMIN-based systematic review found the OLBI far less studied than the MBI and did not identify robust criterion-validity evidence against a health standard (Shoman Y. 2021). A recent critique argues the OLBI, like the MBI, does not measure the construct it claims to (Bianchi R. 2024). This property is untested for the OLBI.

Criterion validity: organisational Absent (a finding about the literature)thinevidence form: canonical

indirect no OLBI-specific organisational-outcome study located, and the general evidence rests on other instruments

Direct prospective links between OLBI scores and registered organisational outcomes (sickness absence, turnover, performance) were not located in in the latest review pass's retrieval; the systematic review did not surface such evidence for the OLBI (Shoman Y. 2021). The instrument is validated mainly through internal structure and cross-instrument correlations rather than against work outcomes. The general burnout-outcome literature (for example burnout predicting turnover intention and sickness absence) is largely built on the MBI and CBI rather than the OLBI (Zheng J. 2024; Borritz M. 2006), so it cannot be read across to the OLBI without direct evidence. This is an important gap: organisational criterion validity for the OLBI is effectively absent in the retrieved literature.

Internal consistency Moderatewell-establishedevidence form: mixed

indirect Chinese, Greek and Polish samples

Internal consistency is generally acceptable to good. The Chinese OLBI gave Cronbach's alpha of 0.905 (total), 0.933 (exhaustion) and 0.876 (disengagement) with split-half 0.883 (Xu H. 2021); a Greek short OLBI gave exhaustion alpha 0.810 (omega 0.823) and disengagement alpha 0.742 (omega 0.756) (Gkontelos A. 2023); the Polish validation supported reliability by alpha and test-retest across three samples (Baka Ł. 2016). Disengagement is typically the weaker scale, and its low within-person reliability was one reason a daily-diary study favoured a unidimensional model (Gruszczynska E. 2021).

Sub-grades (evidence differs by subgroup):

• {"subgroup": "Exhaustion", "grade": "High", "note": "alpha typically 0.80 to 0.93"}
• {"subgroup": "Disengagement", "grade": "Moderate", "note": "alpha often 0.74 to 0.88; weakest at within-person level"}

Test-retest reliability Lowthin

indirect the English validation reports test-retest as acceptable without an extractable coefficient in open text, and the Polish study used a 6-week interval but its numeric retest value was not in the retrieved abstract

Temporal stability of the OLBI is under-reported. The canonical English validation states test-retest reliability was acceptable but the coefficient is not in the openly retrievable text (the full paper sits behind a paywall and could not be read in the latest review pass), and the Polish validation used a 6-week retest interval across three samples but did not surface a numeric coefficient in the retrieved abstract. No clean subscale-level ICC or r with interval and sample was extractable, so this is graded Low and thin rather than well-established, and the specific coefficients should be treated as not yet verified.

Measurement invariance Lowcontestedevidence form: mixed

indirect German, Greek, Brazil/Portugal and teacher samples; much of it on reduced-item forms

Reis et al. found the two-factor OLBI structure held within German employee and student samples and supported partial invariance of job and academic burnout within the German samples and of academic burnout across Greek and German students (Reis D. 2015). The Brazil/Portugal study examined invariance across countries and sexes on a reduced 15-item form (Sinval J. 2019), and a Greek short-form study reported measurement invariance across teacher demographic characteristics (Gkontelos A. 2023). A daily-diary multilevel study found cross-level invariance was not supported, indicating that person-level and within-person burnout are not the same latent variable (Gruszczynska E. 2021). Invariance evidence is partial and often on reduced or short forms rather than the canonical 16-item instrument.

Sub-grades (evidence differs by subgroup):

• {"subgroup": "Across sex / country (students, teachers)", "grade": "Low", "note": "partial invariance on reduced forms"}
• {"subgroup": "Across measurement levels (within vs between person)", "grade": "Very low", "note": "cross-level invariance not supported"}

Responsiveness and MIC Absent (a finding about the literature)untestedevidence form: canonical

indirect no responsiveness or MIC study located

No responsiveness statistic or minimal important change (MIC) value for the OLBI was located, and the systematic review did not identify responsiveness evidence for the instrument (Shoman Y. 2021). The OLBI's balanced wording is argued to make it suitable for tracking change toward engagement, but this is a design rationale rather than demonstrated responsiveness.