SafetyInsights.org

This study evaluated the impact of construction safety training on accident rates in Spanish industry.

Data from 1,283 Spanish construction companies over 11 years were analysed. They studied the effects of training in the same year of the accident, and the prior three years.

I’m not really sure what to make of these findings – covered later.

Results

Key findings:

·         “Contrary to our hypotheses, our results show that more hours of safety training of all types are associated with more accidents, questioning its effectiveness”

·         e.g. “companies that provided the most training to their workers were the ones that reported the most accidents”

·         “Although the effect size is very small, this result does not support the expectation that safety training reduces workplace accidents”

·         “safety training is not achieving the expected effect of reducing accidents”

They say this finding, while contrary to their hypothesis, supports a 2012 study which found no evidence of the effect of training on accident reduction.

Explaining this lack of effect is that “there is insufficient knowledge transfer to workers, among other possible factors”; predicated on the point that “the transfer of knowledge is crucial for altering workers’ risk behavior and thus reducing accidents”.

They also highlighted that most companies in the Spanish construction sector, like elsewhere, are small to medium-sized companies, who generally lack the resources to design or adapt effective training to the operational needs.

Importantly, and a source of my own uncertainty about what to make of the findings, is where they point out the “possible reactive training practice”. That is, companies who report the most accidents also deliver the most training in response to the accidents (hence finding a statistical relationship).

Quoting the paper, “the company provides safety training to its workers once they have suffered an accident, aiming to correct possible errors and improve the knowledge and behavior of the same or other workers to avoid future accidents”, or trying to comply with some legal requirements.

Hence, it’s difficult to elucidate the relationships between training and performance (and prior research has been mixed on the effects).

[** Also, there’s the old chestnut about whether incident data is even statistically robust enough to show changes based on predictor variables. In many cases, these data are likely underpowered, and heavily influenced by unmodelled and largely statistically random variables…but, whatever.]

They also point out that even though there is mandatory training required in Spanish construction, this need isn’t being met by many companies in their sample. Further, “in many cases [the training] are not even taught in the languages of foreign workers”.

They unpack some specific findings around accumulated training over years of work. From what I can tell, the findings suggest that accumulated training of certain types >3 years may reverse the trend, and potentially show some improvements (I could be mistaken here, though).

Outside the scope of the study, but they cite work suggesting that the incorporation of more practical training can improve measurable outcomes.

In concluding, they state:

“This study provides empirical evidence of the unexpected effect of safety training on accidents in the construction sector in Spain. Contrary to our expectations, the training received by workers does not contribute to a reduction in accidents, suggesting that preventive measures are being implemented in a formalistic rather than effectively manner”.

Ref: Estudillo, B., Forteza, F. J., & Carretero-Gómez, J. M. (2025). Effectiveness of training in reducing accidents in construction companies. Journal of Safety Research, 92, 283-291.

Study link: https://doi.org/10.1016/j.jsr.2024.12.005

My site with more reviews: https://safety177496371.wordpress.com

LinkedIn post: https://www.linkedin.com/pulse/effectiveness-training-reducing-accidents-companies-ben-hutchinson-gaszc