Eurofound EWCS/EQLS wellbeing and working-conditions item sets

Structural validity Moderatecontestedevidence form: mixed

indirect evidence is European cross-national general-working-population data; index structure is debated between competing constructions.

Structural evidence is uneven and index-specific. Munoz de Bustillo et al. 2011 provide a detailed critical appraisal of the dimensional structure of job-quality indicators and argue for theory-led multidimensional indices rather than a single composite (10.1093/ser/mwr005). Green et al. 2013 operationalise EWCS data into distinct job-quality dimensions and examine their distributional structure across Europe (10.1177/001979391306600402). The ETUI European Job Quality Index is built as a set of sub-indices rather than one latent factor (leschke2012, piasna2018). The WHO-5 embedded in EQLS is well established as a unidimensional scale in the wider literature (topp2015). No single agreed latent structure spans the whole EWCS/EQLS item set, consistent with its treatment as an item-set.

Convergent and discriminant validity Moderatewell-establishedevidence form: mixed

indirect convergent evidence is European general-population and, for WHO-5, clinical/general-population, not UK-workplace-specific.

Job-quality indices show expected convergent relationships with wellbeing, health and job-satisfaction outcomes in the EWCS-based literature, and Munoz de Bustillo et al. 2011 discuss the risk that composite indices blur discriminant boundaries between conceptually distinct dimensions (10.1093/ser/mwr005). Green et al. 2013 relate the dimensions to inequality and outcome gradients across Europe (10.1177/001979391306600402). The embedded WHO-5 has extensive convergent evidence with depression and wellbeing measures in the systematic review by Topp et al. 2015 (10.1159/000376585), though that evidence is largely clinical and general-population rather than workplace-specific.

Criterion validity: reference standard Moderatewell-establishedevidence form: parent

indirect WHO-5 reference-standard evidence is clinical/general-population (parent-form evidence for the embedded items), not workplace.

For the job-quality item sets there is no diagnostic reference standard and none is claimed. For the embedded WHO-5, criterion validity against clinical depression reference standards is well established in the systematic review by Topp et al. 2015, which reports good screening performance for depression across many studies (10.1159/000376585). That evidence is a property of WHO-5 as fielded, but was earned in clinical and general-population samples, not in EU working-population deployment.

Criterion validity: organisational Lowthinevidence form: canonical

indirect associations are ecological/cross-sectional at European population level; no individual-level predictive validation of work outcomes.

EWCS job-quality indices are widely used to study associations with work outcomes such as absence, health and job satisfaction at the population level, but these are cross-sectional descriptive and secondary-analysis associations rather than validated predictive criterion studies with verified organisational outcomes (10.1093/ser/mwr005; 10.1177/001979391306600402). Because EWCS/EQLS are anonymous cross-national repeated cross-sections, they cannot link individual scores to that individual's later absence, turnover or performance, so individual-level organisational criterion validity is structurally unavailable and effectively Absent as a validated property.

Internal consistency Moderatewell-establishedevidence form: mixed

indirect WHO-5 alpha is clinical/general-population parent-form evidence; job-quality sub-index reliability is from European technical literature.

Internal consistency is meaningful only for the multi-item sub-indices, not for the item set as a whole. Reliability of the ETUI and Eurofound job-quality sub-indices is discussed in the index-construction literature, with the multidimensional design chosen partly because a single composite would conflate distinct dimensions (leschke2012; 10.1093/ser/mwr005). The embedded WHO-5 has consistently high internal consistency (Cronbach alpha typically around 0.8 or above) in the systematic review by Topp et al. 2015 (10.1159/000376585). No single overall alpha is appropriate for the heterogeneous item set.

Sub-grades (evidence differs by subgroup):

• {"subgroup": "embedded WHO-5", "grade": "High", "note": "high internal consistency, alpha ~0.8+, but parent-form clinical/general-population evidence (topp2015)."}
• {"subgroup": "job-quality sub-indices", "grade": "Low", "note": "multi-item indices; reliability discussed in grey/technical literature, not consistently reported as coefficients."}

Test-retest reliability Absent (a finding about the literature)untested

indirect EWCS/EQLS are repeated cross-sections of different respondents; no panel retest of these item sets located.

No test-retest study of the EWCS or EQLS job-quality/wellbeing item sets was located in the latest review pass. The surveys are repeated cross-sections without a re-interview panel design, so stability coefficients are not estimable from the standard data and none were found. The embedded WHO-5 has retest evidence in its own clinical/general-population literature, but that is parent-form evidence not earned in this deployment; it is therefore not entered as an EWCS/EQLS finding. Absence of retest evidence for these item sets is the finding.

Measurement invariance Lowcontestedevidence form: mixed

indirect invariance is the key open question for cross-national comparison; formal scalar invariance not comprehensively demonstrated for job-quality indices.

Cross-country comparability is the central methodological concern for EWCS/EQLS, and it is actively worked on but not fully resolved. Munoz de Bustillo et al. 2011 highlight cross-national comparability and equivalence as core challenges for job-quality indices (10.1093/ser/mwr005), and Green et al. 2013 conduct cross-country comparisons that presuppose comparability while noting its limits (10.1177/001979391306600402). Eurofound invests heavily in translation and harmonisation, but formal measurement-invariance testing (configural/metric/scalar) across all countries and waves is not comprehensively established for the job-quality item sets in the located sources. The embedded WHO-5 has broader invariance evidence in its own literature (topp2015).

Sub-grades (evidence differs by subgroup):

• {"subgroup": "cross-country (job-quality indices)", "grade": "Low", "note": "comparability assumed and worked on, not comprehensively tested to scalar level."}
• {"subgroup": "embedded WHO-5 cross-group", "grade": "Moderate", "note": "broader invariance evidence exists but as parent-form (topp2015)."}

Responsiveness and MIC Very lowthinevidence form: mixed

indirect population-level trend tracking without a validated individual change threshold.

No minimal-important-change or formal responsiveness evidence for the EWCS/EQLS job-quality item sets was located in the latest review pass. The indices are used to track change over time at population level (leschke2012; piasna2018), but no anchor-based minimal-important-change threshold has been established for them. WHO-5 responsiveness is documented in its own clinical literature (topp2015) but is parent-form evidence for the embedded items.