As a result, an estimate of confidence was almost always assigned
when estimates of condition or trend were assigned. The absence of confidence assignment therefore mostly infers a lack of available knowledge, and is an estimate of information paucity. Three expert elicitation workshops were conducted, each over a 3-day period in Perth (Western Australia), Brisbane (Queensland) and Hobart (Tasmania). The locations were chosen to most effectively draw on local knowledge of experts about the nearby regions, and to maximise the prospects of full attendance by the experts at workshops. These workshops were attended by 40 invited experts PI3K Inhibitor Library cell assay from a range of backgrounds, disciplines and institutions (Ward, 2011). Each workshop was conducted
by a mix of plenary and small group discussions, with group consensus scores assigned directly into a spreadsheet Selleck Target Selective Inhibitor Library in plenary. Sub-groups were created as necessary if detailed discussions were required, or time was required to review additional literature. Each score/grade in the spreadsheet was assigned with comments, source citations, and any further information, and this was subsequently updated post-workshop where possible. Subjective bias in the process (sensu Martin et al., 2012) was recognised and managed as far as possible by the organisers and facilitators in both the workshops and post-workshop rounds. At the end of each workshop, and again about a week later, participants were provided with the full dataset from the workshop they attended, and invited to make any corrections, additions, or explanatory Dolichyl-phosphate-mannose-protein mannosyltransferase material. A small number of additional sources and clarifications were made, but less than 10 scores or grades were changed
as a result of this final consultation round, and these were all minor changes. Three data analyses are presented here (i) a summary overview of all workshop-derived data on condition, trends, pressures and confidence; (ii) condition and trend in biodiversity and ecosystem health parameters; and (iii) regional comparisons of condition and trend in biodiversity and ecosystem health parameters. The full workshop raw datasets are available at SoE, 2014b. All data for all biodiversity and ecosystem health components that were assigned a score or grade, including condition scores and trend and confidence categories (181 of the total 196 components, see Supplementary Material) were graphically summarised—median scores, percentage data densities, frequency analysis, and number of observations.