Expert Elicitation Techniques for Social Scientists (bookings closed)
06/12/2018 - 07/12/2018
NCRM, University of Southampton
Professor John Gosling and Dr Jose Pina-Sanchez
Entry (no or almost no prior knowledge)
Training and Capacity Building Co-ordinator
National Centre for Research Methods
Phone: 02380 594069
View in Google Maps (LS2 9NL)
Leeds Institute for Data Analytics, Level 11, Worsley Building, Clarendon Way, Leeds
Increasingly decision makers are interested in knowing about the uncertainty in events that will impact their decisions. Traditional quantitative uncertainty modelling relies on vast amounts of data or assumptions to characterise the uncertainty reliably. Where data is lacking, an expert’s gut judgement is sometimes used, but questions of reliability and transparency of such judgements can always be raised.
In the context of uncertainty modelling, expert knowledge elicitation is the process of translating someone’s judgements about some uncertain quantities into something useful. Elicitation can help us to take stock of the uncertainty about quantities of interest without the cost of data collection whilst using processes that are transparent and reproducible. In this short course, we will introduce the principles of expert knowledge elicitation, demonstrate its use in probabilistic uncertainty modelling and focus on an establish elicitation protocol: the Sheffield method.
The course covers:
- 1: Principles and pitfalls of expert knowledge elicitation;
- 2: The difficulty of picking the right experts;
- 3: The importance of process and transparency;
- 4: The Sheffield method for capturing expert judgements;
- 5: Getting the questions right and being unambiguous: problem definitions and elaboration;
- 6: Facilitating a group to arrive at probabilistic judgements;
- 7: Software to help facilitate expert knowledge elicitation (the SHELF package in R);
- 8: Using the results of an elicitation for uncertainty modelling.
By the end of the course participants will:
- Be able to state the key features of expert knowledge elicitation;
- Have knowledge of the main biases that can arise when probabilitistic judgements are made;
- Be able to set up an expert knowledge elicitation exercise;
- Be able to carry out the model fitting aspects of the Sheffield method via the R package.
This course is mainly lecture-based, including two sessions using computers. During those sessions, we will use R, but no previous knowledge of the software or statistical programming is required.
Participants are encouraged to think of potential applications of expert elicitation techniques in their areas of research as part of their preparation for the course.
The fee per teaching day is:
• £30 per day for UK/EU registered students
• £60 per day for staff at UK/EU academic institutions, UK/EU Research Councils researchers, UK/EU public sector staff and staff at UK/EU registered charity organisations and recognised UK/EU research institutions.
• £220 per day for all other participants
All fees include event materials, lunch, morning and afternoon tea. They do not include travel and accommodation costs.
Full refund is offered for cancellations up to two weeks prior to the course taking place. No refund is available after this date.
R, Expert elicitation , Expert recruitment , Decision making under uncertainty , Uncertainty analysis , Subjective judgements
Related publications and presentations: