Premium
Four case studies in using expert opinion to inform priors
Author(s) -
Kuhnert Petra M.
Publication year - 2011
Publication title -
environmetrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.68
H-Index - 58
eISSN - 1099-095X
pISSN - 1180-4009
DOI - 10.1002/env.1115
Subject(s) - prior probability , expert opinion , expert elicitation , inference , computer science , data science , interpretation (philosophy) , management science , artificial intelligence , bayesian probability , engineering , mathematics , statistics , medicine , intensive care medicine , programming language
Priors gained from expert opinion can be the key to effective decision‐making. Yet, there is continuing controversy with its use because of its subjective and potentially biased nature. I examine the use of expert opinion through four environmental case studies in which one or more experts participated in an elicitation exercise to provide inference on ecological problems. In each case study, I examine how expert opinion informed the model and the potential pitfalls that could result, especially in data‐limited situations. I discuss two opposing schools of thought: (1) experts provide a valuable source of information that can offer useful insights into a model and (2) expert priors are most times biased, leading to incorrect results and bad decisions. I show that expert opinion has a place in ecological analyses if carefully structured in a model. In situations in which data are limited or simply not available, steps can be taken to ensure its proper use and interpretation from models so decisions can be made urgently and updated when new data becomes available. Copyright © 2011 John Wiley & Sons, Ltd.