No abstract

No abstract

Behavioral adaptation describes the collection of behaviors that occur after a change in the road traffic system. Typically, those behaviors not intended by the initiators of the change having a negative impact on safety are of particular interest. Although behavioral adaptation is frequently cited as an explanatory variable for observed discrepancies between engineering estimates and actual outcomes of safety interventions, a thorough understanding of behavioral adaptation does not, at present, exist. Most theories posit that a driver’s goal to maintain an acceptable level of risk will determine if and when behavioral adaptation will occur; few models incorporate individual driver characteristics into their explanation of behavioral adaptation. Recently, a qualitative model of behavioral adaptation was proposed. The model predicts that the degree of behavioral adaptation to a novel road safety intervention depends on several psychological characteristics of the individual, including propensity to trust automation, “locus of control,” and inclination toward sensation-seeking. To test the predictions of this model, simulator and test-track studies were conducted to investigate the ability of lane departure warnings to induce behavioral adaptation in drivers performing a secondary number-entry task. While the presence of reliable warnings in both settings improved lane-keeping performance, drivers tended to report a high degree of trust in both accurate and inaccurate systems, despite the intentional infidelity of the latter. Externals and low sensation-seekers were more likely to report an increase in trust in the system, regardless of its accuracy. The collective results from both studies indicate that, because of the propensity of some people to trust unreliable or faulty devices, caution should be used in attempting to predict the aggregate safety benefits of these systems.

Safety-compromising accidents occur regularly in the led outdoor activity domain. Formal accident analysis is an accepted means of understanding such events and improving safety. Despite this, there remains no universally accepted framework for collecting and analysing accident data in the led outdoor activity domain. This article presents an application of Rasmussen's risk management framework to the analysis of the Lyme Bay sea canoeing incident. This involved the development of an Accimap, the outputs of which were used to evaluate seven predictions made by the framework. The Accimap output was also compared to an analysis using an existing model from the led outdoor activity domain. In conclusion, the Accimap output was found to be more comprehensive and supported all seven of the risk management framework's predictions, suggesting that it shows promise as a theoretically underpinned approach for analysing, and learning from, accidents in the led outdoor activity domain. STATEMENT OF RELEVANCE: Accidents represent a significant problem within the led outdoor activity domain. This article presents an evaluation of a risk management framework that can be used to understand such accidents and to inform the development of accident countermeasures and mitigation strategies for the led outdoor activity domain.

No abstract