Decision making process: A data-driven approach to understand expert grading processes

19:00 Tuesday 28 May




Finn Laurien (Austria) 1; Stefan Hochrainer-Stigler (Austria) 1; Stefan Velev (Austria) 1

1 - International Institute for Applied System Analysis

Community resilience against flood risk is multi-faceted and there is no simple set of indicators that can be easily measured. In fact, performance measurement falls into the realms of combining many different data sources together. Risk engineers working in the insurance or risk reduction sector are used to this kind of environment. In the same way that resilience comes from many and varied sources, so too does risk.  Usually risk engineers use a document that describes the risk, called a Technical Risk Grading Standard (TRGS), to help them make sense of the data they gather about the situation they are assessing.  The TRGS provides a consistent benchmark against which to quantify risk. Based on this idea, we address the issue how the grading of resilience dimensions against flood events on the community level done by an expert can be related to the underlying data he is using as information. The TRGS has been adapted to the context of communities’ enhanced resilience to flooding.

The approach brings together quantitative and qualitative data about the attributes, resources, and capacities that contribute to resilience, allowing experts to “grade” these factors (based on the TRGS approach) and to help identify actions for enhancing resilience. Based on big data and content analysis approaches we analyzed different sets of resilience questions and dimensions to answer the question for which of them expert judgment grading is more needed or not. This should help to answer the question which resilience dimensions can be addressed with standardized questionnaire tools and for which dimensions also qualitative approaches for grading are needed. We found that rather than the resilience dimensions itself the specific type of grading of a given source of resilience is important, especially if the grading can be related to some easily quantifiable categories and if the underlying data is giving specific information on the whole community level or only part of it. Our analysis gives important indications for which type of resilience dimensions a standardized approach similar to the TRGS can be applied and for which expert judgment is necessary.