PrimeDSF methods compendium, including sample data, test runs, and comparative analysis
The data used in the various tasks of PrimeFish comes from many sources. The approach taken in the analysis within each task depends both on the objective of the task at hand, as well as the data available. The methodology used in PrimeFish therefore spans a wide range of methods; from quantitative and qualitative analysis of interview data and simple descriptive statistics to state-of-the-art advanced statistical models.
This deliverable discusses the various methods used in the various tasks and deliverables. Following a short introductory section, Sections 2-4 present the methodology used in work package 2 (WP2). Section 2 presents the growth accounting used to assess productivity in the harvesting sectors of Denmark, the Faroe Islands, Iceland, Newfoundland and Norway, while Section 3 describes the data envelopment analysis to study efficiency and productivity in aquaculture in several European countries and Vietnam. Section 4 presents the Kalman filter method which is used to study the occurrence of boom-and-bust cycles in various seafood markets. Section 5 discusses the method of co-integration which is used to analyse price transmission and market integration. The global value chain analysis, upon which most of the analysis in WP3 is based, in presented in Section 6 and Sections 7 and Section 8 present the approach taken to study the how market institutions and labelling and certification affect seafood and aquaculture firms. The analysis of European seafood products innovations is discussed in Section 10, while the microeconomic models applied in WP4 are illustrated in Sections 11 and 12. The methods chosen to analyse social awareness, attempts to stimulate fish consumption, and negative press, are presented in Section 13. The choice experiments conducted in WP4 are described in Section 14, while the fisheries and aquaculture competitiveness index developed in WP5 is visited in Section 15. The methodology behind the boom-and-bust model developed in WP5 is presented in Section 16 and the latent class analysis and multinomial logistic regression used in WP5 is discussed in Section 17.
The wide variety of approaches shown in this deliverable show the breadth of the analysis undertaken in PrimeFish, and the range of tools used to study the competitiveness of the European seafood and aquaculture industries in this project.