Assessment of sectoral aggregation distortion in research productivity measurements

The reliability and correctness of the results of the assessment of research productivity in universities by means of bibliometric techniques depend on the level of data aggregation used for the analysis by disciplinary area. The variability among the university research fields, the varying prolificacy of the scientific disciplines and the different levels of representativeness by discipline of the journals covered in source databases are the main causes leading to distortions. Such effects can be reduced considerably by using a level of data aggregation by discipline which is as homogeneous and uniform as possible. This study compares the research productivity data of Italian universities at two different levels of aggregation of output and input data, thus showing and assessing the scale of the distortions that may result.