Abstract

The National System of Researchers of Mexico (SNI) evaluates, selects, and recognized by an economic stimulus to national human capital that makes quality research. This logistics can be considered as a selection of projects, which leads, inevitably, to the choice of specialized human capital. This article uses the technique of analysis and clustering of data known as clustering (k Means) to deepen on the criteria followed by the NSR with regard to the choice of researchers. Once the productive profile of each appointment defined by SNI, and through the Hamming distance is known, is a comparison between the actual and estimated data associated with each appointment. Estimates allowed to conclude that it is not justified the current classification into four groups (appointments), perhaps this is due to that the evaluators of the SNI used information not collected on variables reported by requests. In addition, demonstrates the need for improved statistical information used as the database for the evaluation; the differences that exist in the ratings for the seven knowledge areas defined by the SNI and recommended some of the results to supplement assessments by peers today, provided that improvements are designated the quantity and quality of available information. Certainly, this should serve to streamline the future selection of projects of research and development concerning a programme of public policy research in Mexico.

Keywords: evaluation methods, scientific statistics; comparative analysis, research workers