How to Measure the Effectiveness of OHS Training

Measuring OHS training effectiveness: why is the exam score not enough?

If a maintenance technician who scored 90 in lockout-tagout training changes a press die the following week just by pressing the emergency stop button, the training is successful on paper and a failure on site. An exam measures what a person knows, not what they do. Measuring OHS training effectiveness is done to close this gap: it looks at what has changed in the work after training.

This is no longer just good practice but a regulatory requirement. Art. 16 of the Regulation on the Procedures and Principles of Occupational Health and Safety Training for Employees, which entered into force on 2 April 2026, regulates assessment and evaluation:

For the Regulation's other innovations, see our article on the new OHS training regulation. The Regulation does not describe how effectiveness is to be measured; the employer chooses the method. The framework below offers a practical, documentable approach for manufacturing businesses in Bursa and the surrounding provinces.

How can the Kirkpatrick model be adapted to OHS training?

The Kirkpatrick model, widely used in training evaluation, consists of four levels. The Kirkpatrick model can be adapted to OHS training as follows:

The levels do not replace each other. If satisfaction is high and exam scores are high but site observation is weak, the problem usually lies not in the training content but in the equipment, work pressure or a supervisor who does not set an example. To make this distinction, all four levels need to be examined.

Levels 1 and 2: the training evaluation survey and pre/post tests

The training evaluation survey should be short; a ten-question form is often filled in carelessly. We find five questions are enough:

1. What was covered in the training was relevant to my job. (1–5)
2. The presentation and language were understandable. (1–5)
3. Enough time was allowed for the practical section. (1–5)
4. There is something I will do differently in my job after the training. (Yes / No; if yes, please write it)
5. A question not answered in the training, or a hazard I have seen in the workplace: (open-ended)

The last question is the most valuable part of the survey; it often reveals a hazard that is not in the risk assessment.

At the learning level, the level assessment required by the Regulation can be used as a pre-test. When a pre-test and post-test measuring the same topics are compared, it can be seen question by question which topic was not successfully taught. For example, if in a group of 40 only 18 people answer the question "zero-energy verification after isolation" correctly in the post-test, this topic should be covered again with a practical demonstration in the next training. For how to keep exam documents, see our article on OHS training certificates and records.

Level 3: post-training site observation

The only way to measure behaviour is to look at the site. For post-training site observation, prepare a list of 5–8 items derived directly from the training content that can be marked "yes/no". Example for lockout-tagout training:

• Were all the machine's energy sources isolated before starting maintenance?
• Were a personal lock and tag fitted, and was the tag completed?
• Was zero energy verified by trying to start the machine?
• If more than one person was working, did each fit their own lock?
• At the end of the job, was energy not restored until the guards were refitted?

The observation can be carried out by the specialist, a foreman or a trained employee; what matters is that the same list is applied in the same way. If observation is used as a disciplinary tool, employees change their behaviour when the observer arrives and the result becomes meaningless. Respond to correct behaviour with thanks on the spot and to missing behaviour with a short reminder on the spot.

Track observation results with a simple rate: compliance rate = number of safe observations / total number of observations × 100. If 42 of 60 observations before training are safe, the rate is 70 per cent; if 54 of 60 observations six weeks after training are safe, the rate has risen to 90 per cent. If the rate does not rise, the equipment, time pressure or the workability of the procedure should be questioned rather than the training. To add observations to your general site walks, you can use our article on the OHS site walk checklist.

Level 4: near-miss reports and outcome indicators

At the results level, the workplace's indicators are examined. Avoid two misinterpretations here:

• An increase in near-miss reports is not bad news. An increase in the number of reports after hazard recognition training may show that employees are starting to see and speak up about more things. The content of the reports, that is, whether the hazards covered in training are being reported, matters more than the number. For setting up the system, see our article on the near-miss reporting system.
• Short-term changes in accident numbers can mislead. In small workplaces, a few months of accident data are highly subject to chance. Read accident indicators over a one-year period and together with other indicators; our article on accident frequency and severity rates guides the calculation.

We explained how to track leading and lagging indicators together in our article on OHS performance indicators.

Training effectiveness form: sample template

For each training topic, a one-page training effectiveness form brings the four levels together:

The decision on the form is the evidence for the "change or repeat if necessary" provision in Art. 16 of the Regulation. Art. 11 of the Regulation requires annual training programmes to be prepared and approved by the employer, and the views of employees or their representatives to be obtained when preparing them; effectiveness results should be an input to the following year's programme. For designing the programme, see our article on the annual OHS training plan. In workplaces with an OHS committee, Art. 8 of the Regulation on OHS Committees gives the committee the duty of monitoring the implementation of training programmes and giving feedback when it sees deficiencies; putting a summary of the form on the committee's agenda makes this duty concrete.

The mistakes we often see in effectiveness evaluation are:

• Looking only at the satisfaction survey. Liking the training does not mean learning or applying it.
• Not defining the target behaviour. If the question "what do we want to see differently on site?" is not answered before training, the observation list cannot be written either.
• Deciding on a single observation. Calculating the rate without collecting enough observations from different shifts and people is misleading.
• Leaving the result in the file. If a "partly effective" decision is not followed by a corrective action with an owner and deadline, the evaluation stays on paper.

In our OHS training service, we also prepare the site observation list after training and link the results to the next training plan. You can plan training and observation dates with the OHS calendar.

Frequently Asked Questions

Is measuring the effectiveness of OHS training mandatory?

Yes. Art. 16 of the 2026 training regulation requires the employer to evaluate the effectiveness of training and, if necessary, to change the programme or trainer or repeat the training. The employer determines the method.

Does scoring 60 in the exam show that training was effective?

No. The exam measures the level of knowledge. For effectiveness, you must also look at whether what was learned is applied on site, that is, at behaviour and outcome indicators.

How long after training should site observation be carried out?

Generally between 2 and 8 weeks is appropriate. Observation done too early may measure a short-lived effect, and observation done too late may measure the effect of other changes. Taking a measurement with the same list before training also makes comparison easier.

If near-miss reports increased after training, did the training fail?

No. An increase in reports may show that employees recognise hazards better and are willing to report them. The subject of the reports and the closure rate should be assessed together.