Why Generic Wellbeing Questions Produce Useless Data
"On a scale of 1–10, how happy are you at work?" is not a wellbeing survey. It is an engagement sentiment check with no diagnostic value.
The difference matters because the WHS Act 2011 (Cth) — and its harmonised equivalents in every state except Victoria, which runs its own OHS Act 2004 — requires that you identify specific psychosocial hazards, assess the risk they pose, and implement controls. A single happiness score tells you nothing about which hazards are present, which teams are most exposed, or what controls might work.
Validated instruments map survey items to specific constructs. When your COPSOQ score for "quantitative demands" is 72 out of 100 in your logistics division, you know where to look next. When your "trust in management" subscale drops 12 points year-on-year in one business unit, you have a signal worth investigating before it becomes a WorkCover claim.
Generic questions produce numbers you can put in a slide deck. Validated, construct- based questions produce actionable findings.
The Six Psychosocial Constructs Every Survey Should Cover
ISO 45003:2021 — the international standard for psychological health and safety at work, increasingly referenced by Australian regulators — identifies work design, work environment, organisational culture, and relationships as the four domains of psychosocial risk. Translating those domains into survey constructs gives you the following six areas:
| Construct | What it measures | Example survey item |
|---|---|---|
| Job demands | Workload, cognitive load, emotional demands, time pressure | How often do you have too much work to do everything well? (Never – Always) |
| Job control | Autonomy over tasks, pace, and methods | To what extent can you influence the amount of work assigned to you? |
| Social support | Practical and emotional support from supervisors and colleagues | My immediate supervisor helps me solve problems that arise in my work. (Agree scale) |
| Role clarity | Clear, consistent expectations and responsibilities | I know exactly what my responsibilities are. (Agree scale) |
| Interpersonal relationships | Conflict, bullying, trust, belonging | In the past 12 months, have you been subjected to unwanted behaviour that you would describe as bullying or harassment? |
| Work–life balance | Interference between work demands and personal life | How often does your work interfere with your family life or personal commitments? |
You do not need to measure all six with equal depth in every survey cycle. Rotate depth based on your current priority hazards and known exposure profiles.
Specific Questions That Surface Real Risk
Below are 20 questions drawn from validated Australian and international instruments. These are not all one survey — they are a menu. Select 15–20 items appropriate to your risk profile.
- 1.How often do you not have enough time to complete your work tasks?
- 2.How often does your work require you to hide how you are feeling?
- 3.How often are you required to make difficult decisions as part of your job?
- 1.Can you influence the order in which you carry out your work tasks?
- 2.Do you have a say in choosing who you work with?
- 3.Are you able to take a break from your work when you need one?
- 1.My manager gives me the information I need to do my job well.
- 2.My manager addresses conflicts and disagreements within the team fairly.
- 3.I feel comfortable raising a work-related health concern with my manager.
- 1.In this team, it is safe to speak up about a problem without fear of negative consequences.
- 2.When I make a mistake at work, it is held against me.
- 3.Members of this team are able to raise difficult issues and problems.
- 1.Over the past 4 weeks, how often have you felt calm and relaxed? (WHO-5)
- 2.Over the past 4 weeks, how often has your daily life been filled with things that interest you? (WHO-5)
- 3.Overall, how would you rate your work ability today compared with your lifetime best? (0–10 scale)
- 1.In the past 12 months, have you experienced or witnessed bullying, harassment, or discrimination at work?
- 2.I feel I belong in this organisation.
- 3.My colleagues treat me with respect.
The Legal Context: What Australian Employers Are Actually Required to Do
The obligation does not come from a single regulation — it is layered.
The Work Health and Safety Act 2011 (Cth) places a positive duty on PCBUs to ensure the health — including psychological health — of workers so far as is reasonably practicable. Section 47 specifies that this duty includes consulting workers on health and safety matters that affect them. A structured survey is the most scalable way to fulfil that consultation requirement across a large or dispersed workforce.
Safe Work Australia's Model Code of Practice: Managing Psychosocial Hazards at Work (2022)— adopted in most jurisdictions — explicitly lists workforce surveys as a recommended method for identifying psychosocial hazards. It also recommends that survey results be analysed by workgroup and integrated with other data sources (workers' compensation claims, absenteeism records, incident reports) to form a complete picture.
From a privacy standpoint, survey data is personal information under the Privacy Act 1988 (Cth). Health-related survey responses (distress scores, injury disclosures) may constitute sensitive information under the Act, requiring explicit consent and heightened security obligations. Some state health privacy laws — notably the Health Records Act 2001 (Vic) — impose additional requirements.
Where survey items ask about discrimination, reasonable adjustments, or disability- related barriers, the Disability Discrimination Act 1992 (Cth) and relevant state anti-discrimination legislation become relevant to how you act on the responses.
Five Design Decisions That Determine Whether Your Survey Produces Action
Good questions are only half the problem. The other half is survey architecture.
- 1
Anonymity vs. confidentiality — know the difference
Anonymous surveys cannot be linked back to an individual; confidential surveys can but are protected. Most enterprise platforms offer confidentiality, not anonymity. Be honest with staff about which you are running. Misrepresenting this destroys trust and depresses future response rates.
- 2
Set a reporting threshold of n=5 minimum
Never report data for groups smaller than five respondents. Below this threshold, individuals can often be identified by inference, particularly in specialist or small-team roles. Build this threshold into your reporting platform, not just your data analyst's good intentions.
- 3
Use 5-point Likert scales consistently
Mixing 4-point and 6-point scales within a single survey makes cross-construct comparison unreliable. Stick to one scale type. The COPSOQ III uses 5-point frequency and degree scales — adopt those unless your validated instrument specifies otherwise.
- 4
Close the loop visibly, within 30 days
The most common reason wellbeing survey response rates fall below 40% by the second cycle: staff never saw what happened with the first one. Publish a summary of findings and at least three specific actions you are taking, within 30 days of closing the survey. Even if the actions are small.
- 5
Triangulate with administrative data
Survey responses describing high demands and low support should be cross-referenced with your workers' compensation claims data, absence records, and any EAP utilisation figures. When survey signals and administrative signals align, your risk rating goes up and your business case for intervention becomes much harder to argue against.
Turning Survey Results Into a Psychosocial Risk Register
A survey is the starting point, not the end product. Your goal is a risk register that feeds a control plan.
Once you have analysed results by workgroup, flag any subscale score that falls in the "adverse" or "high risk" range on your chosen instrument's normative benchmarks. COPSOQ III publishes Danish population norms; if you are using it in an Australian context, treat scores in the top quartile of adverse responses as requiring investigation rather than waiting for a benchmark comparison.
For each flagged construct and workgroup, convene a short (45-minute) facilitated group session to explore the "why". Survey data tells you the what. These conversations — sometimes called hazard conversations or structured debrief sessions — tell you the causes. Without the causes, your controls will be generic and ineffective.
Document findings against the ISO 45003:2021 hazard taxonomy, assign a risk rating (using your existing WHS risk matrix), nominate an accountable manager, and set a review date. That is a psychosocial risk register — built directly from employee survey data.