In the context of road safety, evaluations are performed in order to provide knowledge on the safety strategies’ performances. This type of knowledge is needed by the stakeholders who are involved in the management of the road safety system. It allows them to control the activities they perform. For instance, national governments need information about the abilities of a new speed regulation to save people or to reduce injuries. Such results, which are provided by evaluations, help them to validate the strategy (a posteriori) or to decide its implementation (a priori). According to this statement, evaluation activity is one part of the safety improvement process. In this report, a model of the evaluation activity is presented. It aims to guide evaluators in the achievement of evaluations. In order to improve the evaluators’ monitoring on their activity and to perform relevant evaluations, we identify some issues that need to be handled. The current issues on evaluation activity concern its objectives, the indicators, the tools/methods and the unpredictable changes. Firstly, it is difficult for evaluators to identify what are the stakeholders’ expectations concerning evaluation. Their diversity implies diversity in their needs; they do not all want the same things. Some of them focus on the economic side, other on the public health or technological sides. We notice that for evaluators (who design and perform evaluations) this is quite a major issue because we have not found method or tools that could help them in identifying needs. Secondly, the major media used in evaluation to deliver the results is the “indicator”. It is a mathematical object that gives factual information. Related to the first point, the conception of indicators is dependent of the expressed needs. Therefore, according to issues in identifying needs, evaluators have difficulties to offer relevant indicators. They mainly used indicators that they are in the habit to use and that they are able to calculate. Moreover, we do not identify methods/tools that allow designing new indicators. Thirdly, evaluation is an activity that needs to be formalized in order to guide the evaluators’ work. We only identify some operational methods and tools but we do not know how they were build and if they are relevant according to the needs. We do not find a general evaluation model that could handle its definition, its realization, its valorization and its evolution. Finally, the road system is a complex system that is usually represented by the triptych: vehicle, user and environment. One can understand its complexity by the unpredictable behavior of each of its component. For the evaluators, who need to understand what they evaluate, complexity is an obstacle. They cannot foresee all the unpredictable changes that could affect performances of a safety strategy. For instance, the implementation of a system that automatically regulates speed of the vehicle could lead to the appearance of new driver behavior that could be dangerous. The drivers can take advantage of it to perform other task like phoning or reading. Complexity also implies a dynamic vision of the evaluation activity; it is not always the same. Evaluation activity evolves according to the changes of its context. However, how evaluators can make evolve evaluations? Following these observations, we developed a framework of the evaluation activity. This report introduces the representation of this framework through a systemic paradigm. Various functional and descriptive models are proposed. Evaluators used them as guidelines in order to model knowledge on study case and to design evaluations. Our general framework takes into account the various viewpoints of stakeholders and evaluators. It allows performing evaluations that are (1) relevant for all the various stakeholders and that (2) aim to assess performances according to various viewpoints (aggregation of various performances from road safety fields — accidentology, economy, biomechanics, etc.). (Author/publisher)
Samenvatting