Safety Sciences
Content
Safety sciences covers causation models, safety culture and climate, human and organizational factors, and the evidence behind common interventions.
Causation models. Heinrich's domino theory and the 1:29:300 ratio still appear on exams, but treat them as historical: the ratio is not a law and the 88 percent unsafe acts figure is not supported. Reason's Swiss cheese model explains defenses and latent conditions. Rasmussen's framework and AcciMap address system levels. Behavior-based safety focuses on observable acts but drifts into blame when it is not paired with engineering controls and leadership commitment.
Culture and climate. Culture is shared assumptions and behavior; climate is the measurable perception of it. Look for reporting, just culture, learning, and leadership commitment rather than posters.
Evidence and statistics. Incidence rates normalize for exposure: (cases x 200,000) / hours worked. Correlation is not causation; watch for confounding, selection bias, and regression to the mean when a bad year is followed by a celebrated program.
Traps. Claims that most injuries come from unsafe acts, that near-miss pyramids hold constant ratios across industries, or that a zero-injury record proves a safe system. Zero injuries can simply mean underreporting.
Worked example
A site with 300,000 hours worked records 3 recordable cases. Rate = (3 x 200,000) / 300,000 = 2.0 per 200,000 hours, comparable to a TRIR of 2.0.