OWHS
OverviewHow to read this registryCorrections and verificationsDownloads
multi-item-scale · reviewed 2026-07-12 contested: Structural validity, Convergent and discriminant validity, Measurement invariance

HSE Management Standards Indicator Tool

Licence verified: 2026-07-12 · record reviewed: 2026-07-12 (pass two)

Identity

Version: 35-item revised Indicator Tool, seven subscales; the same instrument is described across studies as the standard form. A shorter 25-item form is sometimes referenced in practice but no psychometric validation of an official 25-item short form was located in this pass (see record_notes).

Structure: 35 items across seven subscales: Demands (8), Control (6), Managerial Support (5), Peer Support (4), Relationships (4), Role (5), Change (3), per Edwards 2008.

Original citation: Cousins R, MacKay CJ, Clarke SD, Kelly C, Kelly PJ, McCaig RH (2004). 'Management Standards' work-related stress in the UK: practical development. Work & Stress, 18(2), 113-136. doi:10.1080/02678370410001734322; companion policy/science paper MacKay et al. (2004), doi:10.1080/02678370410001727474.

Steward / publisher: UK Health and Safety Executive (HSE), the Great Britain national regulator for workplace health and safety. The tool sits within the HSE Management Standards approach to work-related stress (Cousins 2004; MacKay 2004).

Licence status (verified 2026-07-12): CONFIRMED and now verified. The HSE Management Standards Indicator Tool pages and download pages carry the site-wide statement 'All content is available under the Open Government Licence v3.0, except where otherwise stated'. The tool is thus Crown copyright released under OGL v3.0, free to use with attribution.
Source: https://www.hse.gov.uk/stress/standards/notesindicatortool.htm and https://www.hse.gov.uk/stress/standards/downloads.htm (OGL v3.0 footer, Crown copyright)

Constructs claimed

The Indicator Tool is an exposure measure of psychosocial working conditions, not a wellbeing or health outcome measure. It captures employee perceptions of seven work-stressor dimensions that map onto the HSE Management Standards: Demands (workload, work patterns, environment), Control (autonomy over how work is done), Managerial Support (encouragement and resources from line management), Peer Support (support from colleagues), Relationships (conflict and unacceptable behaviour, e.g. bullying), Role (understanding of role and avoidance of conflicting roles) and Change (how organisational change is managed and communicated). Higher subscale scores denote more favourable (lower-risk) conditions. The framework derives from the demand-control-support tradition and UK epidemiological work on psychosocial hazards (Cousins 2004; MacKay 2004).

Evidence

Deployment context caveat. HSE MSIT measures exposure to psychosocial working conditions, not a wellbeing outcome; its property grades should be read in that framing. (applies to every property below)

Structural validity Moderatecontestedevidence form: canonical

indirectness see findings

The seven-factor structure is well supported in UK data but the Managerial Support and Change dimensions are unstable in several non-UK adaptations. The instrument was developed from an item pool reduced by exploratory factor analysis to 35 items across seven subscales (Cousins 2004). The first confirmatory test on organisational-level UK data (39 organisations, N=26,382) found the original 35-item seven-factor first-order model gave an acceptable fit, and a second-order model was also acceptable, suggesting a possible higher-order single work-related stress dimension (Edwards 2008). Cross-nationally, a seven-factor solution replicated and was equivalent across large UK (N=7,589) and Italian (N=1,298) private-sector samples in multiple-group CFA (Toderi 2013). However, several adaptations do not recover seven distinct factors: the Italian revised tool collapsed Managerial Support and Change into a single factor (termed 'elasticity'), retaining five to seven scales (Magnavita 2012), the Irish ROI-MSIT (N=7,377) likewise merged Managerial Support and Change (Boyd 2016), and an Argentine study retained only 24 items in six factors, discarding Change entirely (Vaamonde 2023). A North Italian healthcare-worker study confirmed a seven-component structure but identified an additional factor relating to participation in work organisation (Veronesi 2022). A heavily revised Iranian version returned a nine-factor solution with new items, and so is best treated as a distinct instrument (Zeinolabedini 2025).

Confidence note: Moderate: the seven-factor model has strong, consistent support in large UK samples and one large cross-national test, but the Managerial Support/Change distinction is not reproduced in several international adaptations, indicating the structure is not fully robust outside the UK.

Convergent and discriminant validity Moderatecontestedevidence form: canonical

indirectness see findings

Convergent validity against other psychosocial work measures is generally adequate, with some weak spots in discriminant terms for specific subscales. In an Italian municipality sample (N=760), Indicator Tool scales showed moderate to strong correlations with the corresponding Job Content Questionnaire scales, and each scale added specific predictive contribution to self-reported stress, job satisfaction and job motivation (Marcatto 2014). Subscales correlated in the expected directions with stress-related outcomes in the UK and Italian cross-cultural study (Toderi 2013). In the Argentine adaptation, discriminant validity between dimensions was adequate but convergent validity was a concern for Control, Role clarity and Relationships, where average variance extracted was at or below 0.50 (Vaamonde 2023). Subscales also related meaningfully but with small effect sizes to burnout dimensions on the Maslach Burnout Inventory, with Demands and Role linked to emotional exhaustion (Carpi 2021).

Confidence note: Moderate: several independent samples show expected-direction correlations with established measures (JCQ, MBI), but detailed convergent/discriminant metrics (e.g. AVE) are reported mainly in non-UK adaptations and are not uniformly strong.

Criterion validity: reference standard Moderatethinevidence form: canonical

indirectness see findings

Associations with health and job-attitude outcomes are consistently reported, but the evidence is overwhelmingly cross-sectional and against self-reported outcomes rather than objective organisational endpoints such as verified sickness absence or turnover. In a UK Health and Social Services Trust (N=707, 29% response), more favourable Management Standards scores were positively associated with job satisfaction and negatively with job-related anxiety, depression and witnessed errors/near misses (Kerr 2009). In a UK call centre (N=304), only Demands (Spearman rho=-0.211) and Relationships (rho=-0.134) correlated significantly with GHQ-12 distress, while other dimensions did not, a mixed result that qualifies claims of uniform criterion validity (Kazi 2013). In an Italian bank the tool related to GHQ-12 distress and Work Ability Index scores (Guidi 2012), and in a UK prison-service sample (N=1,038) odds ratios linked poor psychosocial conditions to impaired psychological wellbeing, though the authors noted exposure scores alone were insufficient to set intervention priorities without an outcome measure (Bevan 2010). A systematic review concluded there was a clear relationship between Indicator Tool scores and alternative wellbeing and stress measures (Brookes 2013). Direct validation against objective sickness absence is weak: an early four-organisation study of the earlier filter-question version found the screening filters insensitive with low positive predictive value, and warned that using work absence as the measure of stress cost may substantially underestimate the true burden (Main 2005).

Confidence note: Moderate: multiple studies link the tool to self-reported distress and job attitudes in expected directions, but findings are cross-sectional, at least one UK study found most subscales unrelated to GHQ-12, and criterion evidence against objective outcomes (verified absence, turnover, diagnosed conditions) is sparse and, for absence specifically, problematic.

Criterion validity: organisational Very lowthinevidence form: canonical

indirectness see findings

Organisational criterion evidence (sickness absence, turnover, performance, diagnosed conditions in a work context): see the criterion findings; graded from the pass-one record.

Confidence note: Moderate: multiple studies link the tool to self-reported distress and job attitudes in expected directions, but findings are cross-sectional, at least one UK study found most subscales unrelated to GHQ-12, and criterion evidence against objective outcomes (verified absence, turnover, diagnosed conditions) is sparse and, for absence specifically, problematic.

Internal consistency Highwell-establishedevidence form: canonical

indirectness see findings

Internal consistency is generally good to excellent across subscales and languages. In the anchor UK analysis, Cronbach's alpha was Demands 0.87, Control 0.82, Managerial Support 0.88, Peer Support 0.82, Relationships 0.78, Role 0.83 and Change 0.80, with the original development study (Cousins 2004) reporting a comparable range of about 0.78 to 0.89 (Edwards 2008). The Italian revised tool reported alphas of 0.75 to 0.86 (Magnavita 2012), the Irish ROI-MSIT 0.75 to 0.91 (Boyd 2016), and the Argentine adaptation composite reliability of 0.70 to 0.82 (Vaamonde 2023). The revised Iranian version reported an overall alpha of 0.949 and McDonald's omega of 0.739 to 0.894, though for a restructured nine-factor instrument (Zeinolabedini 2025).

Confidence note: High: multiple good-quality studies across several countries and large samples consistently report subscale alphas at or above roughly 0.75, meeting conventional adequacy thresholds.

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

indirectness see summary

No test-retest (temporal stability) evidence for the standard UK 35-item Indicator Tool was located in this pass, and this is a genuine gap in the psychometric record. The reliability studies retrieved report internal consistency (alpha, omega) rather than repeat administration over time (Edwards 2008; Magnavita 2012; Boyd 2016). The only intraclass correlation coefficient located (ICC=0.92) comes from the revised nine-factor Iranian version, where the abstract does not clearly establish it as a test-retest coefficient over a defined interval, and in any case applies to a modified instrument rather than the original tool (Zeinolabedini 2025).

Confidence note: Absent: no test-retest reliability evidence for the standard tool was located in the latest review pass; the absence is flagged explicitly.

Measurement invariance Lowcontestedevidence form: canonical

indirectness see findings

Invariance evidence exists across sector and across the UK/Italy language boundary, reaching at least metric level, but scalar invariance is not clearly demonstrated. Multiple-group CFA across large UK (N=7,589) and Italian (N=1,298) private-sector samples found the seven-factor solution equivalent, supporting metric equivalence together with factor variance and factor covariance equivalence (Toderi 2013). A dedicated UK study reported measurement invariance of the Indicator Tool across public and private sector organisations (Edwards 2012); the detailed invariance level (configural/metric/scalar) from that paper was not retrievable from the abstract in this pass. No formal invariance testing across sex, age or occupational group was located in the latest review pass.

Confidence note: Low: cross-national evidence supports at least metric invariance and a UK study supports public/private invariance, but scalar (intercept) invariance is not clearly established and invariance by sex, age and occupation is untested in the retrieved literature.

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

indirectness see findings

No formal responsiveness or minimal important change (MIC) evidence was located in this pass. The tool has been used in pre/post and longitudinal designs, for example a longitudinal healthcare-worker study spanning the onset of the SARS-CoV-2 pandemic (Veronesi 2022) and a pre/post stress-management evaluation among Sierra Leone healthcare workers that observed changes in domain scores (Jones 2020), but neither established responsiveness statistics or a minimal important change threshold. HSE positions the tool for organisational monitoring and comparison rather than individual change detection.

Confidence note: Absent: no responsiveness or MIC estimates for the tool were located in the latest review pass; longitudinal use exists but without formal responsiveness/MIC analysis.

Populations, languages and norms

The tool has been validated in the UK and adapted into several languages and settings, with UK national benchmark norms published at organisational level. UK evidence spans large multi-organisation datasets (Edwards 2008; Cousins 2004), the NHS and health and social care (Kerr 2009), call centres (Kazi 2013), the prison service (Bevan 2010) and Ministry of Defence personnel, where content was judged too narrow without an added work-life balance scale (Bridger 2016). Validated or adapted non-UK versions include Italian (Magnavita 2012; Guidi 2012; Toderi 2013; Veronesi 2022), Irish (Boyd 2016), Argentine (Vaamonde 2023) and a revised Iranian version (Zeinolabedini 2025); the Argentine study also notes prior validation in Italy, Iran and Malta. UK normative percentile benchmark tables derived from the organisational dataset are provided to let employers compare their organisational averages against national reference values (Edwards 2008), consistent with the HSE benchmarking approach.

Criticisms and controversies

Several substantive criticisms recur. First, the seven-factor structure is not fully reproducible outside the UK: the Managerial Support and Change dimensions repeatedly collapse into one factor in Italian and Irish adaptations, and Change is sometimes dropped altogether, questioning the universality of the seven-domain model (Magnavita 2012; Boyd 2016; Vaamonde 2023). Second, criterion evidence is largely cross-sectional and against self-reported outcomes, with at least one UK study finding most subscales unrelated to GHQ-12 distress (Kazi 2013), and an early evaluation showing the screening filters were insensitive with poor positive predictive value and that sickness absence understates the true cost of psychosocial hazards (Main 2005). Third, the tool measures exposure (perceived working conditions) rather than health or wellbeing outcomes, so exposure data alone are insufficient for prioritising interventions and are best paired with an outcome measure (Bevan 2010). Fourth, content validity can be too narrow for specific contexts, for example the military, where a work-life balance dimension was needed (Bridger 2016). Finally, test-retest reliability and formal responsiveness/MIC remain unestablished for the standard tool (this pass), and much criterion evidence shares common-method variance because exposure and outcome are both self-reported.

References (20)

  1. Cousins R, MacKay CJ, Clarke SD, Kelly C, Kelly PJ, McCaig RH (2004). 'Management Standards' work-related stress in the UK: practical development https://doi.org/10.1080/02678370410001734322
  2. MacKay CJ, Cousins R, Kelly PJ, Lee S, McCaig RH (2004). 'Management Standards' and work-related stress in the UK: policy background and science https://doi.org/10.1080/02678370410001727474
  3. Edwards JA, Webster S, Van Laar D, Easton S (2008). Psychometric analysis of the UK Health and Safety Executive's Management Standards work-related stress Indicator Tool https://doi.org/10.1080/02678370802166599
  4. Edwards JA, Webster S (2012). Psychosocial risk assessment: measurement invariance of the UK Health and Safety Executive's Management Standards Indicator Tool across public and private sector organizations https://doi.org/10.1080/02678373.2012.688554
  5. Toderi S, Balducci C, Edwards JA, Sarchielli G, Broccoli M, Mancini G (2013). Psychometric properties of the UK and Italian versions of the HSE Stress Indicator Tool: a cross-cultural investigation https://doi.org/10.1027/1015-5759/a000122
  6. Kerr R, McHugh M, McCrory M (2009). HSE Management Standards and stress-related work outcomes https://doi.org/10.1093/occmed/kqp146
  7. Marcatto F, Colautti L, Larese Filon F, Luis O, Ferrante D (2014). The HSE Management Standards Indicator Tool: concurrent and construct validity https://doi.org/10.1093/occmed/kqu038
  8. Bevan A, Houdmont J, Menear N (2010). The Management Standards Indicator Tool and the estimation of risk https://doi.org/10.1093/occmed/kqq109
  9. Guidi S, Bagnara S, Fichera GP (2012). The HSE indicator tool, psychological distress and work ability https://doi.org/10.1093/occmed/kqs021
  10. Magnavita N (2012). Validation of the Italian version of the HSE Indicator Tool https://doi.org/10.1093/occmed/kqs025
  11. Brookes K, Limbert C, Deacy C, O'Reilly A, Scott S, Thirlaway K (2013). Systematic review: work-related stress and the HSE Management Standards https://doi.org/10.1093/occmed/kqt078
  12. Kazi A, Haslam C (2013). Stress management standards: a warning indicator for employee health https://doi.org/10.1093/occmed/kqt052
  13. Boyd S, Kerr R, Murray P (2016). Psychometric properties of the Irish Management Standards Indicator Tool https://doi.org/10.1093/occmed/kqw163
  14. Carpi M, Bruschini M, Burla F (2021). HSE Management Standards and burnout dimensions among rehabilitation professionals https://doi.org/10.1093/occmed/kqab055
  15. Vaamonde JD, Giacobino AE (2023). Psychometric properties of the HSE Indicator Tool: evidence from Argentina https://doi.org/10.1093/occmed/kqad010
  16. Bridger RS, Dobson K, Davison H (2016). Using the HSE stress indicator tool in a military context https://doi.org/10.1080/00140139.2015.1057544
  17. Veronesi G, Giusti EM, D'Amato A, et al. (2022). The North Italian Longitudinal Study Assessing the Mental Health Effects of SARS-CoV-2 Pandemic on Health Care Workers, Part I: study design and psychometric structural validity of the HSE Indicator Tool and Work Satisfaction Scale https://doi.org/10.3390/ijerph19159514
  18. Jones S, White S, Ormrod J, et al. (2020). Work-based risk factors and quality of life in health care workers providing maternal and newborn care during the Sierra Leone Ebola epidemic https://doi.org/10.1136/bmjopen-2019-032929
  19. Zeinolabedini M, Motlagh ME, Heidarnia A, et al. (2025). Psychometric properties of the revised version of the Health and Safety Executive Management Standards Indicator Tool https://doi.org/10.1038/s41598-025-30714-x
  20. Main CJ, Glozier N, Wright IA (2005). Validity of the HSE stress tool: an investigation within four organizations by the Corporate Health and Performance Group https://doi.org/10.1093/occmed/kqi044

Record notes

[Upgraded from v0.1 to v0.2 structure in pass two; criterion field split, licence re-verified 2026-07-12.] Framing: the Indicator Tool is a work-stressor EXPOSURE measure of perceived psychosocial working conditions, not a wellbeing outcome; the schema's outcome-oriented property fields (criterion validity, responsiveness/MIC) were completed with that in mind and criterion evidence is reported as associations with separate outcome measures. Version ambiguity: the literature retrieved in the latest review pass consistently describes and validates the 35-item seven-subscale form; The maintainers did not locate a psychometric validation of an official 25-item HSE short form, so the item_count and version fields report the 35-item form and flag the 25-item form as referenced but unverified in this pass. Two properties are genuinely ABSENT in the retrieved evidence and should not be read as oversights: test-retest reliability (no temporal-stability study for the standard tool was found; the only ICC comes from a restructured Iranian version and its interpretation as test-retest is unclear) and responsiveness/MIC (longitudinal use exists but no formal statistics). Measurement invariance is graded Low because cross-national and public/private evidence reaches at most metric level and scalar invariance is not clearly demonstrated; the Edwards 2012 abstract was not retrievable in full, so the exact invariance level from that paper is stated conservatively. Licence terms are inferred from the tool's HSE Crown-copyright provenance rather than a cited licence document. Coefficients: all alpha/omega/ICC/correlation values quoted come from abstracts or (for Edwards 2008) the full text retrieved in the latest review pass. Overall confidence in the record is moderate: internal consistency is well established (High), structural, convergent and criterion validity are Moderate with real conflicts reported, while invariance is Low and test-retest and responsiveness are Absent.