Calculator Inputs
Example Data Table
Use this sample to understand typical inputs.
| Period | Unsafe Acts | Observations | Rate (%) |
|---|---|---|---|
| 2026-01-01 | 7 | 120 | 5.833 |
| 2026-01-08 | 6 | 135 | 4.444 |
| 2026-01-15 | 5 | 140 | 3.571 |
| 2026-01-22 | 4 | 150 | 2.667 |
Formula Used
- Unsafe Act Rate (%): Rate = (Unsafe Acts ÷ Total Observations) × 100
- 3-Period Moving Average: average of the latest three rates to smooth noise.
- Trend Slope: linear regression slope of rate versus time, expressed per selected period.
- Percent Change: ((Last Rate − First Rate) ÷ First Rate) × 100
Rates make periods comparable even when observation counts differ.
How to Use This Calculator
- Select a period type that matches your inspections.
- Enter unsafe acts and total observations for each period.
- Click Calculate Trend to view summary metrics.
- Review the moving average to confirm sustained change.
- Set a target rate to compare against the latest period.
- Download CSV or PDF for toolbox talks and reports.
Use the interpretation card to decide if coaching or controls need escalation.
Unsafe Act Trend Article
What the calculator measures
Unsafe act trending turns field observations into a measurable leading indicator. Each period combines the count of unsafe acts with the total observations, so you can compare performance even when inspection effort changes. The output highlights the latest rate, average, and range to show typical exposure.
Unsafe act rate and normalization
The calculator converts raw counts into a rate: (unsafe ÷ observations) × 100. Normalization is critical on construction sites because observation totals vary with crew size, shift coverage, and inspection frequency. A stable rate suggests consistent behaviors; a rising rate signals drift. Use consistent dates for each inspection cycle sitewide.
Trend line and slope interpretation
To quantify direction, a linear regression trend line is fit to the period-by-period rates. The slope is reported as percent change per selected period (daily, weekly, or monthly). A positive slope means the unsafe act rate is increasing over time; a negative slope indicates improvement.
Moving average for noise reduction
Behavior data can be noisy. A 3‑period moving average smooths short spikes caused by one unusual inspection or a brief surge in activity. When both the trend slope and moving average climb together, treat it as a stronger signal than a single high period.
Targets, bands, and decision thresholds
Set a target rate to compare your latest period against an agreed threshold. The calculator also labels the latest rate using practical exposure bands (low, moderate, high, critical). Use bands to standardize escalation: coaching at moderate, focused controls at high, and leadership intervention at critical.
Data quality and sampling discipline
Trend quality depends on consistent observation rules. Define what constitutes an unsafe act, train observers, and document categories. Keep sampling steady across trades and locations, and avoid “inspection bursts” only after incidents. Better sampling produces more trustworthy trend signals.
Turning results into field actions
Use the highest-rate periods to guide short, specific interventions: pre-task plan refreshers, spot checks on recurring behaviors, and peer-to-peer coaching. Pair results with two corrective actions and an owner, then review the next period to confirm whether the rate responds.
Reporting and continuous improvement
Export CSV or PDF reports for toolbox talks, subcontractor meetings, and monthly reviews. Track the same crew or area across time to see whether controls are sustaining change. Over several periods, combine this leading indicator with incidents and audits for a complete safety picture. Share results promptly to keep crews engaged and accountable.
FAQs
1) What is an unsafe act rate?
It is the percentage of observed behaviors classified as unsafe for a period: (unsafe acts ÷ total observations) × 100. It helps compare periods even when inspection volume changes.
2) How many periods should I enter?
Use at least two to calculate a trend, but six to twelve periods usually gives a more reliable slope and moving average. Short series can swing due to random variation.
3) Why does my rate increase when counts stay similar?
If total observations drop while unsafe acts remain similar, the rate rises because the denominator is smaller. Track both counts and totals to understand whether the change is sampling or behavior.
4) What does a positive slope mean?
A positive slope means the unsafe act rate is trending upward over time, suggesting worsening behaviors or weak controls. Confirm with the moving average before escalating actions.
5) How should I set a target rate?
Base targets on historical performance and risk tolerance. Start with a realistic reduction goal, review monthly, and adjust as sampling improves. Avoid targets that discourage honest reporting.
6) Can I use this for different trades or areas?
Yes. Enter a site label and trade or area to keep reports comparable. Trend one crew or zone per dataset, or run separate submissions for each group to avoid mixing behaviors.
7) Do the CSV and PDF downloads include my inputs?
Yes. Exports include your identifiers, period type, entered rows, calculated rates, moving averages, and summary metrics so you can share consistent reports with supervisors and subcontractors.