Data Weighting in Contemporary Stock Assessment Models That Integrate Information from Multiple Diverse Data Sets

Contemporary fisheries stock assessment models often use multiple diverse data sets to extract as much information as possible about all model processes. However, models are, by definition, simplifications of reality, and model misspecification can cause the inclusion of data sets that degrade the model results. The process, observation, and sampling components of the model must all be approximately correct to minimize biased results. The data set could be analyzed separately from the integrated model and the resulting parameter estimates and their uncertainty used in the integrated model, but this will likely involve assumptions inconsistent with the integrated model. Other approaches to reduce the influence of a data set include arbitrarily changing the variance parameter of the likelihood function, adding process variation, and using information-focusing estimation. Model misspecification and process variation can be accounted for in the variance parameters of the likelihoods, but it is unclear when this is appropriate. I recommend external estimation of the sampling error variance used in likelihood functions, including process variation in the integrated model, and internal estimation of the process error variance. These concepts are illustrated in the context of estimating selectivity in integrated fisheries stock assessment models.