Safety Training and Communication
Training that actually changes behavior
- Needs assessment compares required knowledge and skill against current capability, derived from task analysis and audience characteristics.
- Objectives must be observable and measurable - "identify," "demonstrate," "calculate" - never "understand" or "be aware of."
- Adult learning depends on relevance to the task, practice with feedback, reinforcement over time, and respect for the learner's prior experience.
- Kirkpatrick's four levels: 1 reaction, 2 learning, 3 behavior, 4 results.
- Competency is knowledge plus skill plus demonstrated performance. A signed attendance roster documents attendance, not competence.
- Retraining triggers: new or changed hazard, incident, process change, observed performance deficiency, or a regulatory requirement.
- Communication barriers: language, literacy, hearing and noise, shift and remote work, culture, and lack of feedback channels.
Traps. The most common is a stem stating that participants rated the class highly and asking what it proves - that is Level 1 reaction, not evidence of learning. Test scores are Level 2; observed field behavior is Level 3; reduced incidents is Level 4. Also watch for training offered as the corrective action for a design or engineering problem, and for confusing hazard communication (labels, SDS, and training) with training generally.
Rule of thumb. To verify training effectiveness, select a measure of behavior or results. Reaction data tells you the room was comfortable, not that the risk was reduced.
Worked example. After confined-space training, an auditor observes entrants testing the atmosphere before entry without prompting. That observation is Level 3 evidence - transferred behavior - and it outweighs any post-class survey score.