Assessing Individual Intellectual Output in Scientific Research: Mexico's National System for Evaluating Scholars Performance in the Humanities and the Behavioral Sciences.

Assessing the research of individual scholars is currently a matter of serious concern and worldwide debate. In order to gauge the long-term efficacy and efficiency of this practice, we carried out a limited survey of the operation and outcome of Mexico's 30-year old National System of Investigators or SNI, the country's main instrument for stimulating competitive research in science and technology. A statistical random sample of researchers listed in the area of Humanities and Behavioral Sciences-one of SNI's first and better consolidated academic divisions comprising a wide range of research disciplines, from philosophy to pedagogy to archaeology to experimental brain research-was screened comparing individual ranks or "Levels of distinction" to actual compliance with the SNI's own evaluation criteria, as reflected in major public databases of scholarly production. The same analysis was applied to members of a recent Review Committee, integrated by top-level researchers belonging to that general area of knowledge, who have been in charge of assessing and ranking their colleagues. Our results for both sets of scholars show wide disparity of individual productivity within the same SNI Level, according to all key indicators officially required (books issued by prestigious publishers, research articles appeared in indexed journals, and formation of new scientists), as well as in impact estimated by numbers of citations. Statistical calculation from the data indicates that 36% of members in the Review Committee and 53% of researchers in the random sample do not satisfy the official criteria requested for their appointed SNI Levels. The findings are discussed in terms of possible methodological errors in our study, of relevance for the SNI at large in relation to independent appraisals, of the cost-benefit balance of the organization as a research policy tool, and of possible alternatives for its thorough restructuring. As it currently stands SNI is not a model for efficient and effectual national systems of research assessment.