School of BioSciences - Theses
Permanent URI for this collection
Search Results
1 - 1 of 1
-
ItemWho to trust? Assessing and improving expert judgement in ecological domains( 2019)Managing ecosystems and species involves estimation and prediction of complex ecological interactions and their response to management intervention. While data are invaluable for informing management decisions, often the data required are unavailable, incomplete or uninformative. In these contexts, expert judgement is routinely utilised to inform critically important decisions. It is important that these judgements represent the best possible foundation for decisions and assessments. The application of structured expert elicitation protocols has been advocated as a means by which expert judgement may be improved. While these protocols have been substantially discussed in the ecological literature, their widespread application in ecological domains has remained limited. It is suggested that a lack of prescriptive guidance, supported by evidence that structured protocols can be applied within the practical and financial constraints of most ecological contexts, may be impeding the widespread adoption of these protocols. This thesis explores how barriers to the implementation of structured elicitation protocols may be overcome, and the extent to which their application improves judgements. It provides prescriptive guidance on the application of the IDEA protocol for structured expert elicitation, and presents evidence that the application of the entire protocol is both practical and improves judgements. It examines how judgements derived from the IDEA protocol may be further improved through performance weighted aggregations of the Classical Model, and the extent to which performance weighting may be practically applied.