Distinguished Lecture by Matthew Welsh, University of Adelaide
Abstract:
The oil industry regularly makes decisions that are high-stakes, requiring long-term forecasting, accurate estimation of unknown parameters and complex modelling of scenarios. Observational and experimental evidence, however, tells us that people's natural decision-making tendencies are reliant on simple, inaccurate estimation strategies that lead to systematic biases in judgments and decisions. Preventing people from using these processes, however, is very difficult - as evidenced by the fact that the oil industry has been discussing the problem of overconfidence in forecasting for more than 40 years without great improvements. In order to work out how to improve judgments and decisions, we need a better understanding of how people's cognitive tendencies and limitations lead to observed biases and errors.
By way of example, this lecture discusses a core aspect of cognition - memory - explaining how this differs both from people's intuitive assumptions about how it works and from how a purely rational process would. Key examples of biases that result from our inherent memory processes are discussed and the implications of this for decision making in oil industry contexts presented, leading into a discussion of how to design elicitation processes informed by cognitive science that can limit the impact of biases.
Members should come away from this lecture understanding that we should not expect people to be able to make better decisions if we do not, first, understand how and why people think the way they do. This is true for professionals at all stages of a decision process: from SMEs generating inputs for models; to team leaders selecting which development options to consider; to senior decision makers' assessments of these options relative to other opportunities.
Biography:
Dr Matthew Welsh has a PhD in Psychology and has worked for 16 years in the Australian School of Petroleum, applying his understanding of human cognitive processes to improving oil industry decisions. In this role, he has supervised 20 student projects, published 50 papers in psychology and industry outlets and written a book on how cognitive processes affect the judgments and decisions that scientists make: "Bias in Science and Communication: a field guide." He has consulted and collaborated with oil companies incuding BG, BHP, ExxonMobil, Santos and Woodside, as well as with researchers in diverse fields including psychology, geophysics, medicine and defence. His key focus is on the problem of elicitation - how we elicit estimates from experts in such a way as to avoid biasing their responses and to ensure that their estimates accurately reflect what they do and do not know.