Assessment
05 Safety Culture Baseline
Document FSSMS-ASM-05Stage 1Revision 5Version 0.3.3
Topic pages open on your jurisdiction, and requirements tables show its line. You can change this on any page.
To measure, before anything else changes, how safe people feel to speak up, how much they believe their officers and their chief mean what the policy says, and how well the department actually learns from events; then to measure the same things again at 12 and 24 months so the chief can see the number move. Culture is not a chapter at the end of this package. It is the thing the controls depend on, and it is the one thing a department cannot audit from documents.
This note gives three instruments, a protocol for running them, a complete self-assessment worksheet, and the rules for feeding the results into the gap assessment and the management review.
- Safety climate: the shared perception, at a point in time, of how much safety is valued by management and supervisors (Zohar, 2000)DOI.
- Safety culture: the deeper, slower-changing shared assumptions about what matters here, which climate reflects (USFA & IAFC, 2015).
- Psychological safety: the shared belief within a team that it is safe to speak up, ask, admit error and challenge (Edmondson, 1999)DOI.
- Psychological health and safety: the member’s wellbeing under the workplace factors of the national standard (CSA Z1003-13); distinct from psychological safety.
- Learning system: how an event becomes a report, a review, understanding, action and sustained change (The Human Factor, 2026c).
- Maturity band: one of five levels describing how an organization handles safety information and events, from Reacting to Generative (Westrum, 2004DOI; Hudson, 2003DOI).
- Learning firewall: the commitment that information gathered for learning is not used for discipline (The Human Factor, 2026c).
- Normalized deviance: a shortcut repeated without consequence until it becomes the norm (Vaughan, 1996).
Three different things get called “safety culture.” Keep them apart:
| Term | What it means | How this note measures it |
|---|---|---|
| Safety culture | The shared, mostly unspoken assumptions about what matters here; slow to change; what people do when nobody is watching | Inferred from climate, behaviour and the learning-system assessment over time; not measured directly |
| Safety climate | The shared perception, at a point in time, of how much safety is valued by management and supervisors; “the room temperature”; changes within months (Zohar, 2000DOI; USFA & IAFC, 2015) | Survey (section 2) |
| Learning system | How an event becomes a report, a review, understanding, action and sustained change (The Human Factor, 2026c) | Self-assessment on six constructs (section 4) |
The national culture-change initiative put it well: to change long-term culture you have to understand short-term climate (USFA & IAFC, 2015). The package therefore measures climate and the learning system, because both move within the life of an implementation, and reads culture from their trend.
First choice: a fire-service-specific validated instrument. The Fire Service Organizational Culture of Safety survey (FOCUS), developed at a US university research centre with federal funding, measures management commitment to safety, supervisor support for safety, safety behaviour, and outcomes including injury and job satisfaction (Drexel University, n.d.). It has been used by several hundred departments and tens of thousands of firefighters, gives a department a benchmarked report, and is the source of the most useful fire-service climate findings: management commitment runs about ten points higher in career than in volunteer departments; supervisor support is consistently rated higher than management commitment; and higher scores on both are associated with lower injury odds (Geczik et al., 2024)DOI. Its validation evidence comes mainly from the developers and from convenience samples (Geczik et al., 2024)DOI. Access for a Canadian department would need an agreement with the research centre.
Fallback: a free generic instrument. The Nordic Safety Climate Questionnaire (NOSACQ-50) is free, validated across industries, available in French, and 50 items long, which is short enough for a volunteer station meeting (Kines et al., 2011)DOI. It measures seven dimensions: management safety priority and ability; management safety empowerment; management safety justice; workers’ safety commitment; workers’ safety priority and risk non-acceptance; safety communication, learning and trust in co-workers’ competence; and trust in the efficacy of safety systems (Kines et al., 2011)DOI. Its fire-specific validation is limited.
Supplement: a short psychological-safety scale. Seven items on whether it is safe to take a risk, raise a problem, admit a mistake or ask for help in this crew (Edmondson, 1999)DOI, scored at the crew level. This is the lever most within a company officer’s control (Geczik et al., 2024)DOI, and the one the package treats as the unit of intervention.
A five-level ladder, after Westrum (2004)DOI and Hudson (2003)DOI, used two ways: as a qualitative companion to the survey that a crew can locate itself on in ten minutes, and as the implementation road map (“where are we, what is the next rung”), which is how the New Zealand research proposed it for volunteer brigades (NZFS, 2016b). The same five bands score the learning-system self-assessment in section 4.
| Band | How information is handled | How events are handled | What safety is here |
|---|---|---|---|
| Reacting (pathological) | Messengers are shot; failure is hidden; responsibility is shirked | Nothing happens unless the regulator or a death forces it | Who cares, as long as we are not caught |
| Complying (bureaucratic) | Messengers are tolerated; information may be ignored; responsibility is compartmentalized | Reports are filed; the form is the outcome; fixes are local patches | A paperwork exercise we do because we have to |
| Building (calculative) | Information is collected on a system; some of it is used | Investigations are done by trained people; recommendations exist; some are tracked | A set of rules and numbers managed by the safety office |
| Learning (proactive) | Messengers are trained; information is sought; responsibility is shared | Events are triaged, reviewed and acted on; near misses rise; the loop closes visibly | Something the line owns and works on before harm |
| Generative | Messengers are rewarded; information flows across levels; failure leads to inquiry | Normal work is studied as well as failure; lessons are shared outside; the system itself is reviewed | How we do business |
A one-page exercise for a station or a crew, drawn from 05 Design Principles and Lessons Learned section 6.5 and The Human Factor’s conference teaching on normalization of deviance (The Human Factor, 2025b; Vaughan, 1996). The crew is shown the list of practices the fire service commonly normalizes (no SCBA on overhaul, no backer when reversing, seat belts off in the cab, no 360 on automatic alarms, PASS alarms tolerated, working tired, no eye protection on medical calls, crews of three, continuing a drill past the low-air warning, intersections not cleared, SCBA not checked at the start of shift, gear not cleaned after a fire) and asked, anonymously, “which of these are normal here?” Count the hands. The result is a leading-indicator set for that station and a conversation starter that no survey item can match. Repeat it annually.
| Step | When | What |
|---|---|---|
| Baseline | Within the first 60 days of implementation, before any visible intervention | Climate survey; learning-system self-assessment; normalized-deviance self-check at a sample of stations; leading-indicator baseline from 04 Baseline Metrics and Indicator Set |
| Feedback | Within 30 days of the survey closing | Results to every station, by station and overall, with what will be done |
| Re-measure 1 | 12 months after baseline | Same instruments, same method |
| Re-measure 2 | 24 months after baseline | Same instruments; add the trend; input to the external review |
| Ongoing | Every 24 months thereafter, and after any serious event or restructuring | Same |
- Census, not sample. Every member, career and volunteer, including officers, chiefs, dispatch if part of the department, and civilian staff in operational support. A culture survey that reaches the career core and misses the volunteer stations measures half the department; the New Zealand service segmented its indicator by workforce segment for that reason (NZFS, 2016a).
- Target response rate: at least 60 percent overall and at least 50 percent at every station. A station below 50 percent is reported as “insufficient response,” not as a score. In the first year accept the gap and fix the method; do not chase individuals.
- Segmentation. Report by station or platoon, by career and volunteer, by rank band (firefighter, company officer, chief officer), and by years of service band. Never report a cell smaller than seven responses; merge it with the nearest cell and say so.
- Volunteer inclusion. Run the survey at a station meeting or training night with 20 minutes set aside, on paper or by phone, with a QR code, and keep it open for six weeks. Send the invitation from the fire chief and the volunteer association together, saying why and what will happen with the results; the volunteer research is clear that people take part when the process is simple and they know what happens as a result (NZFS, 2016b).
- Anonymous, not merely confidential. No names, employee numbers or dates of birth. Demographics limited to the segmentation fields above, each with a “prefer not to say” option.
- Administered by someone outside the chain of command: the JOHSC worker co-chair with the health and safety officer, a corporate survey unit, or the external reviewer. Paper forms go into a sealed box opened by two people, one from each side of the JOHSC.
- Raw data is held by the administrator, not the department, and is handled under the applicable access-to-information and privacy law (J10 Privacy and Health Information). The department receives aggregated results only.
- The learning firewall applies: nothing written on a survey is used in any accountability process.
- Say all of this on the front of the form and in the chief’s message. People believe anonymity when they can see how it works.
Within 30 days of closing, each station receives a one-page sheet: its scores on each dimension, the department average, the career and volunteer averages, three things that scored well, three that did not, and what the department will do about the three that did not, with an owner and a date. The same sheet, summarized, goes to the JOHSC, the senior team and the Program Improvement Teams. The finding from volunteer research is blunt: people report when it is simple and when they know what happens as a result (NZFS, 2016b). A survey whose results vanish is a withdrawal from the trust account, and the next one will score worse.
- A rise in near-miss reports between baseline and the first re-measure is the expected early success signal, not a deterioration (NZFS, 2016b; FENZ, 2023). Say so in advance.
- Supervisor support is the dimension to watch first. It moves within a year when company officers are trained and held to the supervisor competency in 01 Roles and Responsibilities section 7; management commitment moves later and follows visible resourcing decisions (Geczik et al., 2024)DOI.
- A station whose score falls while the department’s rises has usually had a change of officer or an event. Look there first.
The six constructs follow the path an event travels: it is reported; it is triaged; it is analyzed; the people in it are treated justly; the recommendations are acted on; the leaders and the labour partners keep the system honest. The constructs are drawn from the After the Call learning-system maturity model (The Human Factor, 2026c) and from the literature anchors recorded in 05 Design Principles and Lessons Learned.
| Part | Construct | In one line | Anchors |
|---|---|---|---|
| 1 | Reporting and near misses | People can and do report events, including near misses and good catches, through a trusted route, and get acknowledgement and feedback | (Reason, 1997; Westrum, 2004DOI; Edmondson, 1999DOI; Conklin, 2019; CPSE, 2019, criterion 11A.6) |
| 2 | Triage and tiering | A written, prompt, representative process sorts events into review levels using defined triggers, records the decision, and has independence arranged in advance | (NFPA 1550, 2024, s. 5.7; Vaughan, 1996) |
| 3 | Analysis quality | Reviews are led by trained people with a structured method, reconstruct how the situation looked to those in it, look across the system, interview with respect, and finish on time with a quality check | (Dekker, 2014; Woods & Cook, 1999; Fischhoff, 1975DOI; CSA Z1005:21) |
| 4 | Learning firewall and just culture | A written, counsel-reviewed commitment that learning information is not used for discipline; a separate process for conduct concerns; leaders ask what made sense before who; support for people is kept apart from the review | (Reason, 1997; Marx, 2001; Dekker, 2016; Edmondson, 1999DOI) |
| 5 | Recommendations and follow-through | Stronger, system-level actions are favoured; every recommendation has an owner, a date and a status; leaders record their rationale; effectiveness is checked; lessons are shared | (Hollnagel, 2014; Weick & Sutcliffe, 2015) |
| 6 | Leadership and labour partnership | Senior leaders receive regular reports and take a visible part; the JOHSC co-designs and reviews; labour and management agreed the policy together; the learning system itself is reviewed | (Westrum, 2004DOI; Hudson, 2003DOI; Weick & Sutcliffe, 2015; Vaughan, 1996) |
- Each item is scored 1 to 5 by a group, not an individual: both JOHSC co-chairs, the health and safety officer, one Program Improvement Team lead, one company officer and one firefighter, at least one of them from a volunteer station where the department has them. Score on evidence the group can point to, not on intent.
- Scale. 1 = not true here; 2 = true in a few places or on paper only; 3 = true in most places, with gaps the group can name; 4 = true everywhere the group can see, with evidence; 5 = true, evidenced, and reviewed in the last 12 months.
- Part score is the mean of its six items, to one decimal.
- Band cut points (design choice): below 2.0 Reacting; 2.0 to 2.7 Complying; 2.8 to 3.4 Building; 3.5 to 4.2 Learning; 4.3 and above Generative.
- Weakest-part rule (design choice): the department’s overall band is the band of its weakest part. A learning system is a chain; an event that is reported but never analyzed, or analyzed but never acted on, has not been learned from. Report all six part bands alongside the overall band so the weakest part is visible.
- Any item scored 1 or 2 generates a line in the 02 Gap Assessment Tool and, where it is a required element at the department’s stage, a corrective action.
Score each item 1 to 5 using the scale in 4.3. Record the evidence the group relied on in the last column; “we think so” is not evidence.
Part 1: Reporting and near misses
| # | Item | Score | Evidence |
|---|---|---|---|
| 1.1 | A member can report a near miss or good catch in under five minutes, by phone, paper or QR code, without logging into anything | ||
| 1.2 | The department’s definition of a near miss, with fire-service examples, is posted in every station and taught to every recruit | ||
| 1.3 | Every report is acknowledged to the person who made it within five working days, and that person hears what happened to it within 30 days | ||
| 1.4 | Near-miss reports per 100 calls have been counted for at least 12 months, and the number is rising or stable at a level the JOHSC accepts as credible | ||
| 1.5 | Members at volunteer and outlying stations report at a rate comparable to the career core, and the department has checked | ||
| 1.6 | Members can name a change that came from a report in the last year |
Part 2: Triage and tiering
| # | Item | Score | Evidence |
|---|---|---|---|
| 2.1 | A written policy sets out who triages events, within what time, and against what triggers | ||
| 2.2 | The triage group includes a JOHSC worker representative and meets, or confers, within 24 hours or the next business day of a report | ||
| 2.3 | Every triage decision is recorded with the trigger that set the tier and the names of the roles that decided | ||
| 2.4 | The triggers for the three tiers are the same ones in the policy and have been applied consistently in the last 12 months (the group has checked a sample) | ||
| 2.5 | Any member can ask for a higher tier, and de-escalation needs the whole group’s agreement with reasons recorded | ||
| 2.6 | Independent review capacity for the highest tier is arranged in advance, not found after the event |
Part 3: Analysis quality
| # | Item | Score | Evidence |
|---|---|---|---|
| 3.1 | At least two people in the department have completed incident investigation training to a recognized standard in the last three years | ||
| 3.2 | Reviews follow a written method: terms of reference, timeline, learning interviews, systems analysis, contributing factors, recommendations | ||
| 3.3 | Reports describe how the situation looked to the people in it at the time, before saying what they should have done | ||
| 3.4 | Contributing factors are classified against work-system elements (person, task, tools, organization, environment, external), not only against “human error” | ||
| 3.5 | Second-tier reviews finish within 30 days (60 at most) and the sponsor checks the report against a quality standard before release | ||
| 3.6 | Reviews also record what went well and why, not only what went wrong |
Part 4: Learning firewall and just culture
| # | Item | Score | Evidence |
|---|---|---|---|
| 4.1 | A written, counsel-reviewed policy clause states that learning information is not used for discipline, and the bargaining agents and volunteer association agreed to it | ||
| 4.2 | Conduct concerns (intentional harm, impairment, reckless disregard, criminal acts) go to a separate process run by different people, and that route is written down | ||
| 4.3 | Participants in a review are told in writing, before they speak, what the firewall covers and what it does not | ||
| 4.4 | In the last 12 months no member has been disciplined on the basis of a learning review, a near-miss report or a learning-team write-up (the JOHSC has checked) | ||
| 4.5 | Officers are trained to ask “what made sense at the time” before “who,” and members say they do | ||
| 4.6 | Peer support and psychological first aid are offered after significant events independently of whether the member takes part in a review |
Part 5: Recommendations and follow-through
| # | Item | Score | Evidence |
|---|---|---|---|
| 5.1 | Every recommendation from a review is logged in the Corrective Action Register with an owner, a due date and a status | ||
| 5.2 | Recommendations are rated for strength (stronger, moderate, weaker) and the register shows that stronger actions are favoured | ||
| 5.3 | Leaders decide to accept, reject or defer each recommendation within 30 days and record the rationale, with interim controls for anything deferred | ||
| 5.4 | Closed actions have an effectiveness check recorded, not only a completion date | ||
| 5.5 | Lessons are shared across the department through change notices and learning bulletins within the quarter, and externally where appropriate | ||
| 5.6 | The JOHSC’s written recommendations are answered within 21 days, or sooner where the statute requires, and the responses are filed with the recommendations |
Part 6: Leadership and labour partnership
| # | Item | Score | Evidence |
|---|---|---|---|
| 6.1 | The fire chief receives a learning report at least quarterly and has acted visibly on one in the last year | ||
| 6.2 | The JOHSC reviews learning-program results at every meeting and has a standing consultant on every Program Improvement Team | ||
| 6.3 | The reporting and investigation policy was agreed with the bargaining agents and the volunteer association, and the agreement is in writing | ||
| 6.4 | Senior leaders’ own decisions are inside the scope of reviews, and a review in the last three years has examined one | ||
| 6.5 | The learning program is evaluated annually, and the evaluation changed something | ||
| 6.6 | Members at every rank can describe how the learning system works in their own words |
Score summary
| Part | Mean score | Band |
|---|---|---|
| 1 Reporting and near misses | ||
| 2 Triage and tiering | ||
| 3 Analysis quality | ||
| 4 Learning firewall and just culture | ||
| 5 Recommendations and follow-through | ||
| 6 Leadership and labour partnership | ||
| Overall (weakest part) |
Date of assessment: ____ Assessors (roles): ____ Next assessment due: ____
| Result | Where it goes | What happens |
|---|---|---|
| Climate survey by station and segment | 02 Gap Assessment Tool (culture dimension); station feedback sheets; Program Improvement Teams | Stations in the lowest quartile on supervisor support get company-officer coaching first; findings become the baseline for 04 Baseline Metrics and Indicator Set |
| Psychological-safety scale by crew | Company officer and district chief; aggregated to the JOHSC | Crew-level conversation led by the officer; not published below the station level |
| Learning-system bands | 02 Gap Assessment Tool; Incident Investigation and Learning Program annual evaluation | Items at 1 or 2 become gap lines; required-at-stage items become corrective actions with owners and dates |
| Normalized-deviance self-check | Station leading indicators; Emergency Operations Safety Program and Decontamination and Contamination Control Program teams | The practices a station names become its observation targets for the year |
| All of the above, with trend | Management Review | A standing input under the management review clause of the management-system standard (ISO 45001:2018, cl. 9.3): scores, trend, response rates by segment, actions taken since last review and their effect; the chief’s decisions recorded |
| Annual summary | 03 Accountability and Reporting annual health and safety report to council | Aggregated; segmented by career and volunteer; no station-level data |
- Stage 1 (S1): baseline climate survey with the free instrument, the learning-system self-assessment, and the self-check at every station; feedback sheets. Cost: administrator time (about three days), 20 minutes per member, printing.
- Stage 2 (S2): re-measure at 12 months; psychological-safety scale by crew; results in the management review; station-level actions tracked.
- Stage 3 (S3): re-measure at 24 months with a benchmarked fire-service instrument if available; trend presented to the external review; learning-system self-assessment repeated with an external facilitator.
| Element | Who | How |
|---|---|---|
| Owner | Fire chief | Commissions the baseline, signs the message to members, receives the results at management review and decides the response |
| Doer | Health and safety officer with the JOHSC worker co-chair as joint administrators | Run the survey, the self-assessment and the self-checks; produce the station sheets |
| Verifier | JOHSC (method and anonymity, each run); internal audit (that feedback reached every station within 30 days); external review (Stage 3) | Audit item in Checklist - Incident Reporting, Investigation and Learning |
| Feedback | Station sheets within 30 days; JOHSC at the next meeting; senior team and Program Improvement Teams in the quarter; council annually. A concern raised by a member about the survey or its results gets an answer back to that member through the joint administrators within ten working days, recorded in the survey log | Sheets and log filed in {{Records system}} |
| Closure | Corrective Action Register | Every “what we will do” line on a station sheet is a register item with an owner, a priority on the package’s one scale, P1 (Imminent) to P4 (Improvement), and the scale’s dates; overdue items escalate on the priority’s clock (a P3 (Moderate) item to the deputy chief at 30 days overdue and to the fire chief at 60) |
Loop test. A volunteer station scores lowest in the department on “I can raise a safety concern without it costing me.” Within 30 days the station has its sheet. Within the quarter the district chief and the JOHSC worker co-chair have held a learning-team session at that station, the station officer has completed the supervisor competency module on speaking up, and a register item records what changed. At the 12-month re-measure the item is compared and the result goes to the chief.
This note measures psychological safety, the organizational condition in which it is safe to speak up, ask, admit error and challenge (Edmondson, 1999)DOI. It is created by leaders’ behaviour, by the norms of crew meetings and debriefs, and by the response to reports (Conklin, 2019). It is distinct from psychological health, the member’s wellbeing under the workplace factors of the national standard (CSA Z1003-13), which belongs to Psychological Health and Safety and Peer Support Program. The two are linked: a crew that cannot speak hides distress. But a department with a good peer support team can still have crews in which nobody will say the roof looks wrong, and this instrument exists to find those crews. Survey questions never ask for personal health information.
This chapter relies on the following legal requirements, which differ by province and territory. Check your jurisdiction’s section on each topic page before adopting it.
| Topic | What to check in your jurisdiction | Page |
|---|---|---|
| Joint Committee and Health and Safety Representative | Whether a committee is required and how it is composed, and its role in developing the OHS program this baseline belongs to; the worker co-chair is a joint administrator of the survey | J05 Joint Committee and Health and Safety Representative |
| Privacy and Health Information | The access-to-information and privacy statute that governs survey records holding personal information, and what it requires of their custody and disclosure | J10 Privacy and Health Information |
| Volunteer Firefighters | Whether volunteers are workers under the OHS statute, and so are surveyed as members and may sit on the committee; confirm the department’s volunteer employment status with counsel | J11 Volunteer Firefighters |
Center for Public Safety Excellence [CPSE]. (2019). Fire and emergency service self-assessment manual (10th ed.). Commission on Fire Accreditation International. https://cpse.org/wp-content/uploads/2019/08/10th-Edition-FESSAM-Strikethrough-Underline-v.4.pdf
Conklin, T. (2019). The 5 principles of human performance. Pre-Accident Media.
CSA Group. (2013). CAN/CSA-Z1003-13/BNQ 9700-803/2013 (R2022), Psychological health and safety in the workplace: Prevention, promotion, and guidance to staged implementation (free to read from CSA Group and the Mental Health Commission of Canada). https://www.csagroup.org/store/product/CAN-CSA-Z1003-13-BNQ%209700-803-2013/
CSA Group. (2021). CSA Z1005:21 (R2026), Workplace incident investigation. https://www.csagroup.org/store/product/CSA%20Z1005:21/
Dekker, S. (2014). The field guide to understanding “human error” (3rd ed.). Ashgate.
Dekker, S. (2016). Just culture: Restoring trust and accountability in your organization (3rd ed.). CRC Press.
Drexel University. (n.d.). FOCUS: Fire service organizational culture of safety [Web page]. FIRST Center, Dornsife School of Public Health. https://drexel.edu/dornsife/research/centers-programs-projects/FIRST/our-projects/focus
Edmondson, A. C. (1999). Psychological safety and learning behavior in work teams. Administrative Science Quarterly, 44(2), 350–383. doi:10.2307/2666999
Fire and Emergency New Zealand [FENZ]. (2023). Report 210: Safety, health and wellbeing evidence brief 2023. https://fireandemergency.nz/assets/Documents/Research-and-reports/Report-210-Safety-Health-and-Wellbeing-Evidence-Brief-2023.pdf
Fischhoff, B. (1975). Hindsight is not equal to foresight: The effect of outcome knowledge on judgment under uncertainty. Journal of Experimental Psychology: Human Perception and Performance, 1(3), 288–299. doi:10.1037/0096-1523.1.3.288
Geczik, A. M., Lee, J., Allen, J. A., Raposa, M., Robinson, L., Quistberg, D. A., Davis, A. L., & Taylor, J. A. (2024). An updated analysis of safety climate and downstream outcomes in two convenience samples of U.S. fire departments (FOCUS 1.0 and 2.0 survey waves). Injury Epidemiology, 11. doi:10.1186/s40621-024-00502-8
Hollnagel, E. (2014). Safety-I and Safety-II: The past and future of safety management. Ashgate.
Hudson, P. (2003). Applying the lessons of high risk industries to health care. Quality and Safety in Health Care, 12(Suppl. 1), i7–i12. doi:10.1136/qhc.12.suppl_1.i7
International Organization for Standardization. (2018). ISO 45001:2018, Occupational health and safety management systems: Requirements with guidance for use. https://www.iso.org/standard/63787.html
Kines, P., Lappalainen, J., Mikkelsen, K. L., Olsen, E., Pousette, A., Tharaldsen, J., Tómasson, K., & Törner, M. (2011). Nordic Safety Climate Questionnaire (NOSACQ-50): A new tool for diagnosing occupational safety climate. International Journal of Industrial Ergonomics, 41(6), 634–646. doi:10.1016/j.ergon.2011.08.004
Marx, D. (2001). Patient safety and the “just culture”: A primer for health care executives. Columbia University.
National Fire Protection Association. (2024). NFPA 1550, Standard for emergency responder health and safety (consolidates NFPA 1500, 1521 and 1561). https://www.nfpa.org/codes-and-standards/nfpa-1550-standard-development/1550
New Zealand Fire Service [NZFS]. (2016a). Report 153: International good practice in safety, health and wellbeing in fire and emergency services (prepared by Litmus; case studies). Hosted by Fire and Emergency New Zealand. https://www.fireandemergency.nz/assets/Documents/Files/Report-153-international-good-practice-in-safety-health-and-wellbeing-in-fire-and-emergency-services.pdf
New Zealand Fire Service [NZFS]. (2016b). Report 158: Improving the safety culture and practices of rural fire personnel in relation to near-miss reporting (prepared by Heathrose Research). Hosted by Fire and Emergency New Zealand. https://www.fireandemergency.nz/assets/Documents/Research-and-reports/Report-158-Improving-the-safety-culture-and-practices-of-Rural-Fire-Personnel-in-relation-to-near-miss-reporting.pdf
Reason, J. (1997). Managing the risks of organizational accidents. Ashgate.
The Human Factor. (2025b). Normalization of deviance in the fire service, parts 1 and 2 [Conference presentation, CPSE Excellence Conference; generalized for this package].
The Human Factor. (2026c). After the Call: The learning-system maturity model (construct definitions, bands and limits) [Unpublished manuscript].
United States Fire Administration [USFA], & International Association of Fire Chiefs [IAFC]. (2015). National Safety Culture Change Initiative: Study of behavioral motivation on reduction of risk-taking behaviors in the fire and emergency service (FA-342). https://apps.usfa.fema.gov/publications/display?id=994
Vaughan, D. (1996). The Challenger launch decision: Risky technology, culture, and deviance at NASA. University of Chicago Press.
Weick, K. E., & Sutcliffe, K. M. (2015). Managing the unexpected: Sustained performance in a complex world (3rd ed.). Wiley.
Westrum, R. (2004). A typology of organisational cultures. Quality and Safety in Health Care, 13(Suppl. 2), ii22–ii27. doi:10.1136/qshc.2003.009522
Woods, D. D., & Cook, R. I. (1999). Perspectives on human error: Hindsight biases and local rationality. In F. T. Durso (Ed.), Handbook of applied cognition (pp. 141–171). Wiley.
Zohar, D. (2000). A group-level model of safety climate: Testing the effect of group climate on microaccidents in manufacturing jobs. Journal of Applied Psychology, 85(4), 587–596. doi:10.1037/0021-9010.85.4.587