speaker: Tony O'Hagan
(University of Sheffield)
building: The Observatory
room: Seminar Room
see also: additional details
host/contact: Dr Rob Schick
Ever wanted to know how individual experts come up with their informed opinions? How they generate quantitative answers to difficult and uncertain problems? If so, then join CREEM on Wednesday December 2nd, from 1330-1600 when we host one the world’s leading experts on expert elicitation – Professor Tony O’Hagan from the University of Sheffield. Professor O’Hagan has consulted and instructed government, academia, and many corporations on the successful use of expert elicitation.
Ahead of Professor O’Hagan’s seminar, we will have two shorter presentations of case studies led by members of CREEM. Professor John Harwood (Biology, CREEM) will present work he has led on the use of expert elicitation to help inform policy and provide guidance on the possible impact of sound on marine mammals. Dr Rob Schick (CREEM) will present on the use of expert elicitation to help discern movements of North Atlantic right whales.
Schedule for the afternoon:
- 1330 Introduction
- 1335-1400: John Harwood – EE, marine mammals, and conservation policy
- 1400-1430: Rob Schick – EE, right whales, and the mid-Atlantic migratory corridor
- 1430-1500: Tea & coffee
- 1500-1600: Tony O’Hagan seminar
Location: Seminar Room, CREEM, The Observatory
Harwood: There is growing evidence that individuals of many marine mammal species show a marked change in behaviour when they are exposed to noise from activities such as pile driving and navy exercises. However, the biological significance of this disturbance is unclear. Together with other members of a working group funded by the US Office of Naval Research, we have developed a conceptual framework that can be used to forecast the potential population-level consequences of disturbance. Unfortunately, for most marine mammal populations there are insufficient empirical data to parameterise the mathematical functions that underpin this framework. However, there are a number of situations where regulators urgently require scientific advice on the potential effects of a particular development on specific marine mammal populations. In order to provide this advice, we have used expert elicitation to obtain estimates of the relevant parameters and the uncertainty associated with these estimates. In this talk I will describe how we have designed the expert elicitation process and how we have used the results from that process.
Schick: Approximately 500 North Atlantic right whales remain in the world, and despite decades of protection, their recovery continues to be slow. The migratory corridor in the mid-Atlantic ocean links the calving grounds off the southeastern United States with feeding grounds in and around the Gulf of Maine, yet is one of the most highly industrialised stretches of ocean in the world. Movements of animals through this area are poorly documented. We used expert elicitation to poll experts about two sources of information: 1) the seasonal distribution of right whales in the mid-Atlantic; and 2) certain movement transitions from/to the mid-Atlantic. Here we present results from the elicitation, and document how we will use information from # as priors in a statistical model for movement and health. We highlight important lessons learned - both in terms of how to conduct the elicitation, as well as what types of movement related information remains poorly known. In particular, movements of adult male right whales remains very uncertain. And in general, many experts have little idea of what is happening for right whales in the mid-Atlantic.
O’Hagan: There is no single, accepted state-of-the-art method for eliciting expert knowledge. Different practitioners advocate different methods. This talk will begin by outlining some leading approaches and highlighting the factors which might favour one method over others. I will then concentrate on my own preferred approach, presenting some recent developments.