SafetyInsights.org

This was really interesting. Via document analysis, it explored the origin of the figure that 80% of maritime accidents were due to or largely involved “human contribution” or “errors”. They note that few sources provide any actual evidence supporting this assertion.

292 documents were reviewed.

Note that this paper was interested in exploring the origin of the 80% figure rather than calculate reliable values that would replace it, nor argue for or against the construct of human error as a cause (which, cards on the table, I’m not a strong supporter of the construct). This is worth reading whatever your views.

I highly recommend you read the original source over my summary, since I can only highlight a few (cherry-picked) points.

They first cover some background on the construct. The idea that “learning from other’s mistakes”, although normally framed positively, is said to imply, at least to some degree, that a mistake occurred. Considering mistakes may be a slippery slope to blame – ending the investigation at error rather than digging deeper for other sociotechnical factors.

Further, inconsistencies in the terminology used (which may have different definitions) – human error, human factor, human action, human element, may lead to confusion. In any case, the author nicely points out that because humans are central parts of sociotechnical systems—e.g. designed, managed, maintained and operated by them—it’s not surprising that human performance considerably affects the performance of the entire system.

Results:

Overall it’s said that despite considerable research agreeing that “human error constitutes a significant contribution to the maritime accident occurrence” (p3), the widely accepted figure itself is “unsubstantiated” and the number hasn’t been “put to the test or rigorous quantitative analysis to date” (p22).

As they show, estimates of error or human contributions in maritime range from lower than 15% to up to 100%. Thus, “statements on the magnitude of human error contribution to accidents shall be made with a greater caution as they may be found misleading” (p23). In papers using “human error” terminology, the average was 70.2% involvement, whereas 88.9% was found for research using “human factor” as the term. In some cases, authors used human error and human factor interchangeably whereas other authors recognised the differences and avoided conflating them.

In the cases of 100% of human involvement in maritime accidents, these usually involved the use of the HFACS method (Human Factors Analysis and Classification System). They note that “the findings of investigations performed are strongly affected by the setup of the investigations themselves” (p19). This is supported by the what-you-look-for-is-what-you-find-principle, where they say it’s not surprising that HFACS and human error investigation models inevitably conclude that most accidents involve humans/error.

Other research highlighted in many cases that “human element was found to be present – not responsible, but merely present” (p21), where it is also “hard to separate the human from the technical elements in a total system” (p21).

Noted is that blaming individuals can divert attention away from problematic and defunct processes, where unsafe systems are maintained.

In other research, the issues around human biases and practices influence the outcome of the investigation. In one example it’s discussed that it’s far easier to identify “unsafe acts” or human action/inaction compared to underlying system/structural factors.

Another highlights the legal history of supposedly human contributions to accidents in maritime. When an accidental grounding occurs, it’s said that a presumption of negligence exists in the US legal setting by default (unless proven otherwise). Court documents support this position and is traceable back to at least 1933 in the US which set the legal precedence of blaming the boat’s operator. Investigations likewise concluding human error or a human factor are not surprising.

In concluding, they say that “the issue of how big is the human contribution to maritime accidents is complex and more-faceted to the extent that a single number can hardly describe this complexity” (p24).

Importantly, in maritime (and probably many other domains), much of the research effort is said to be “focused on the negative aspect of human involvement in the system, even despite [International Maritime Organization] specifically naming the importance of human element in prevention of maritime casualties” (p24).

Author: Krzysztof Wrobel, 2021, Reliability Engineering and System Safety.

Shout me a coffee

Study link: https://doi.org/10.1016/j.ress.2021.107942

Link to the LinkedIn article: https://www.linkedin.com/pulse/searching-origins-myth-80-human-error-impact-maritime-ben-hutchinson