Second Workshop on Estimation with the RDBES data model (WKRDB-EST2; outputs from 2020 meeting)

This report shows how the new RDBES that is currently in development will be better able to support the recast EU Data Collection Framework (Regulation (EU) 2017/1004) than the existing RDB. The RDBES is an essential platform for MS and RCGs to fulfil their obligations towards documenting and improving data quality and designing and implementing regional sampling designs. The evaluation of data precision was performed using two complementary techniques. For relatively simple sampling designs it is possible to use analytical functions to calculate the precision (or a related statistical measure such as variance) of a statistical estimate. These calculations and implementations of them in R code are presented in this report. For more complicated sampling designs, the use of analytical functions is usually not feasible. In these cases, it is necessary to evaluate precision using numerical techniques, the main one of which is bootstrapping. This report discussed when bootstrapping is appropriate and gives several worked examples describing how bootstrapping can be applied in different cases. The evaluation of bias is a difficult subject and is hard to quantify. The approach followed in this report was to build on the previous work available in the ICES literature and identify and enumerate the main common sources of bias in catch sampling programs they describe. The information was collated and an evaluation performed as to whether data stored using the RDBES data format and reports issues from them can inform about the potential for bias in catch estimates. A set of example reports was coded that demonstrates the utility of the RDBES in relation to bias issues and can already help member states to identify how deviations in their sampling programmes and sampling variability may potentially lead to bias in their catch estimates.