-          previous types of analyses show that the distribution of impacts of public sector research is extremely asymmetric : across all programs or all organizations, a very small proportion of activities or projects generate the largest part of impacts. Thus, we can achieve a fairly complete analysis of an organization’s overall impacts based on the analysis of a relatively limited number of cases;
-          from this qualitative analysis of the role of research put in context and an analysis of the research processes, a case studies methodology will allow us to identify the mechanisms that generate impacts and thus explain the specific contribution of INRA in the changes observed;
-          a method of standardized case studies allows us to link the advantages of qualitative methods (the detailed comprehension of a single case, the story telling) to quantitative methods (the combination of a series of cases and the production of common descriptors and revelation of shared trends).
 
This approach also faces three main constraints:
-          the selection of a sample of cases that effectively illustrate the diversity of mechanisms and pathways producing different types of impacts ;
-          the standardization of analytical methods that requires the mobilization and adaptation of common concepts of tools ;
-          the production of crosscutting analyses that allow a scaling-up from the case study to the level of the organization.