Running PDCA Cycles with Stress Check Year-over-Year Data — How to Read Trends and Drive Workplace Improvement
Introduction
"We've been running the stress check every year, but we can't really tell what's changed from last year." This is one of the most common complaints HR teams raise about their annual stress check programs.
Japan's Industrial Safety and Health Act requires companies of 50 or more employees to conduct stress checks annually. But the legal minimum — completing the check and reviewing the results — leaves out the part that actually drives change: comparing this year's data to previous years and using those trends to run continuous improvement cycles.
This guide explains how to read multi-year group analysis data, link it to specific interventions, and build a PDCA loop that makes your stress check program a genuine workplace improvement tool rather than a compliance obligation.
1. Why Year-over-Year Comparison Is the Key Step Most Organizations Skip
A Single Check Gives a Snapshot, Not a Direction
Group analysis data from one stress check cycle tells you the current state of a team — high-stress ratio, scale scores for job demand, control, and support. But a single data point gives you no direction: Is 25% high-stress rate getting better? Worse? Is it normal for your industry?
| Comparison type | What it reveals |
|---|---|
| Current year only | Absolute level (no directional context) |
| Vs. industry benchmark | Relative position (useful, but external) |
| Vs. prior years (internal) | Change over time — did interventions work? |
The internal year-over-year comparison is the most actionable for operations teams because it directly answers the question: "Did what we tried last year make a difference?"
Year-over-Year Data as Evidence for Legal Compliance
The Industrial Safety and Health Act encourages employers to document workplace improvement measures taken in response to group analysis findings. Year-over-year data showing "we identified the problem, we took action, the scores improved" creates an evidence trail that's useful for occupational physicians, senior leadership, and any external review.
2. Three Metrics to Track Across Years
Metric 1: High-Stress Ratio by Department
The most direct measure. Track the percentage of employees scoring in the high-stress range by department and overall.
Interpretation guide:
| Year-over-year change | Interpretation | Recommended action |
|---|---|---|
| -5 points or more (improvement) | Clear effect of intervention | Identify success factors; replicate across departments |
| Within ±3 points | Flat | Review whether interventions were substantive |
| +5 points or more (deterioration) | Warning signal | Increase 1on1 frequency, investigate supervisor support |
Metric 2: Scale Score Movement by Department
High-stress ratio is an outcome. Scale scores are the leading indicators — they tell you why stress is high before it fully shows up in the ratio.
The standard 57-item stress check (厚生労働省 standard version) measures scales including:
| Scale group | Key scales | What movement means |
|---|---|---|
| Job demands | Workload, work pace | Rising score = overload risk |
| Job control | Autonomy, skill use | Declining = engagement/autonomy erosion |
| Workplace support | Supervisor support, peer support | Consistently low = structural management issue |
| Stress response | Vitality, fatigue, anxiety | Direct health indicator |
Cross-year patterns to watch: - Supervisor support declining 2+ consecutive years → Time for manager training, 1on1 practice overhaul - Workload high for 3 consecutive years despite interventions → Structural staffing or process change needed - Scale improves the year after specific intervention → Causal evidence that the intervention worked
Metric 3: Intervention Timeline Mapping
The most important management discipline is recording what interventions were implemented, when, and for which departments — and keeping that record alongside your stress check data year over year.
Example timeline:
| Year | Stress check result | Interventions taken |
|---|---|---|
| Year 1 | Supervisor support low; high-stress rate 28% | None |
| Year 2 | Supervisor support slight improvement; high-stress rate 25% | Manager training (all); monthly 1on1 required |
| Year 3 | Supervisor support strong improvement; high-stress rate 19% | 1on1 logs introduced; pulse survey launched monthly |
Without this record, an improving score tells you something worked — but not what or why.
3. The PDCA Loop in Four Steps
Step 1: Plan — Define Priority Themes from Last Year's Results
After reviewing the previous cycle's group analysis, identify: - The department with the highest high-stress rate - The scale with the lowest score (potential cause) - Whether any prior interventions produced visible change
Example:
Last year: Sales department had the highest high-stress rate (31%) and lowest "supervisor support" score Theme for this year: Improve supervisor support and workload balance in sales
Step 2: Do — Run Interventions and Monitor Mid-Year
Annual stress checks create a 12-month blind spot. Running pulse surveys on a monthly or bi-weekly basis fills that gap and lets you catch early signals that an intervention isn't working — before the next stress check cycle.
Recommended annual schedule:
April: Review previous group analysis; finalize intervention plan
May–July: Launch interventions (1on1 training, workload redistribution, etc.)
August: Pulse survey mid-year check — is the target department's trend moving?
September–November: Continue or adjust interventions based on pulse data
December: Annual stress check conducted
January–March: Compare results to prior year; evaluate PDCA outcomes
Step 3: Check — Year-over-Year Comparison as Evaluation
When results arrive, compare against the prior year on three levels:
- Priority department/scale: Did the targeted issue improve?
- Intervention correlation: Does the improvement timing align with when interventions were introduced?
- Unexpected deterioration: Which departments or scales moved negatively despite no major changes?
The third point is often overlooked. An organization focused on one problem area may miss a new one developing elsewhere.
Step 4: Act — Replicate Successes; Elevate New Problems
| Result | Action |
|---|---|
| Priority area improved | Document success factors; replicate in adjacent departments |
| Priority area flat | Deepen intervention or switch approach |
| Unexpected deterioration in another area | Add to next year's priority list |
| Organization-wide deterioration | Escalate to leadership; structural-level response |
4. Three Traps to Avoid in Year-over-Year Comparison
Trap 1: Department Restructuring Breaks Comparability
When departments are merged, split, or renamed between cycles, comparing this year's "Department A" to last year's "Department A" may be meaningless if the populations changed. Document any structural changes that affect group composition at the time they occur, so future comparisons can be annotated appropriately.
Trap 2: Score Improvement Obscures Survey Participation Issues
If a department's response rate drops significantly, the remaining respondents may not represent the full range of stress levels. Track response rate alongside stress scores every year.
| Response rate change | Potential interpretation |
|---|---|
| Drop of 10+ points | High-stress individuals may have opted out |
| Rise of 10+ points | Improved trust in the process (positive signal) |
Trap 3: Improvement Without an Explanation
If scores improve but you have no record of what changed, you cannot reproduce the improvement or confirm it isn't regression to the mean. Keep intervention records even for small changes — a meeting format adjustment, a policy update, a manager change.
5. Designing the Physician and Counselor Referral Threshold Using Multi-Year Data
One powerful use of year-over-year trends is setting evidence-based thresholds for when to escalate to occupational physician consultation or external counseling services.
Rather than treating each year's high-stress rate as a standalone event, use the trajectory:
| Multi-year trend | Response level |
|---|---|
| High-stress rate ≤20% and improving | Continue current interventions; pulse survey observation |
| High-stress rate 20–30%, flat | Strengthen 1on1s; promote EAP or external counseling availability |
| High-stress rate >30% or deteriorating 2+ years | Proactive occupational physician engagement; workload review |
| A specific scale at the bottom for 3 consecutive years | Structural response: management change, process redesign |
This approach prevents the "panic mode" that can occur when a single year's results look alarming, while also preventing complacency when year-to-year variation masks a multi-year decline.
6. How COCKPITOS Period Comparison Works
COCKPITOS's stress check module supports multi-cycle data management with a period comparison dashboard designed for ongoing PDCA operations.
Key features
- Cycle selector: Choose any prior stress check cycles to display side-by-side
- Scale trend visualization: Line charts showing each scale score over time by department
- Department change heatmap: Color-coded view of which departments improved, held, or deteriorated
- Intervention timeline notes: Attach notes to specific time periods to record what interventions were in effect
- Pulse survey overlay: View monthly pulse data on the same timeline as annual stress check results
What this replaces
The traditional workflow: download this year's Excel report, open last year's report, manually compute differences, rebuild charts. This is time-consuming and prone to errors, especially when departments have changed.
With COCKPITOS, the period comparison is two clicks: select the comparison periods, and all differences, charts, and departmental breakdowns render automatically.
Summary
| PDCA step | Stress check application |
|---|---|
| Plan | Set priority themes from last year's group analysis |
| Do | Launch interventions + pulse survey mid-year monitoring |
| Check | Year-over-year comparison of high-stress rate and scale scores |
| Act | Replicate successes; escalate newly emerging problem areas |
Turning the annual stress check into a genuine PDCA tool doesn't require new technology or complicated statistical methods. It requires two habits: comparing this year to prior years and recording what you did between checks. Both of those become significantly easier when your data lives in a system designed for multi-cycle management.
Close the PDCA Loop with COCKPITOS
COCKPITOS combines stress check management, pulse surveys, and 1on1 records in a single platform — built for organizations that want to run continuous improvement cycles rather than annual compliance exercises. If you'd like to see the period comparison dashboard in action, we'd be happy to show you how it works for your scale of organization.