This paper presents a dynamic hierarchical model for the production of university departments. It is a hierarchical model because technical coefficients of the academic production function are modeled as function of second level explanatory variables in such a way that such production function is allowed to change coherently from one department to another. It is dynamic in the sense that these technical coefficients are also allowed to change through time. Moreover, validation and outlier selection techniques are applied to feed an automatic model revision procedure that preserves continuity in the evaluation process.
Hierarchical Models; Quality Management; Academic Evaluation; Recursive Estimation