In practice, behavioral phenomena and procedural aspects are often the most important factors determining the overall success in model based problem solving. Earlier literature has discussed procedural practices and behavioral phenomena such as cognitive biases. However, little attention has been paid to the interdependence between behavioral phenomena and the problem solving process. This Dissertation introduces the idea of path dependence in model based problem solving and in decision analysis, which is a branch of model based problem solving. This idea offers a systemic perspective for capturing the overall impact of cognitive biases and other behavioral phenomena. The term path refers to the sequence of steps taken in the problem solving process. There are usually alternative paths to be followed and the choice of the path can matter. The factors affecting the path include the behavior of the problem solving team, as well as the processes followed, the modeling techniques used, and the problem solving environment, for instance. The idea of paths draws attention to the dynamic interaction of these factors. This Dissertation includes considerations of the effect of the starting point, the accumulation of behavioral effects, and difficulties in changing the path. Taking the path perspective can support the management of model based problem solving projects. This Dissertation provides a checklist to help the problem solving team to reflect on their path and to be aware of its drivers. Procedures are described to help reduce the risk that the problem solving project gets stuck on a poor path. In decision analysis, the path perspective can help in mitigating the effects of cognitive biases. Biases are a concern especially when their effects build up along the path followed in the decision analysis process. This Dissertation shows that it is sometimes possible to find paths along which the effects of biases cancel out each other. In general, one should try to avoid paths where the effects of biases build up in favor of certain alternatives. This Dissertation introduces new bias mitigation techniques. These techniques are shown to be effective in a decision analysis process. Portfolio decision analysis is another systemic perspective discussed in this Dissertation. Environ-mental decisions are often portfolio problems where the task is to find a combination of actions, i.e. a portfolio, to meet the overall objectives. In these decision problems, the traditional approach has been to follow a standard decision analysis process to evaluate alternative portfolios generated manually by experts. This Dissertation describes how biases and path dependence create risks in such processes. The portfolio approach helps avoid these risks and creates new possibilities for stakeholder engagement. This Dissertation presents a review and a synthesis of alternative portfolio modeling approaches. A framework is developed to help environmental modelers use portfolio decision analysis.
|Publication status||Published - 2017|
|MoE publication type||G5 Doctoral dissertation (article)|
- behavioral operational research, multi-criteria decision making, environmental modelling, portfolio decision making, path dependence, systems perspective, cognitive biases, debiasing