We used for the evaluation the number of papers that were published by the applicants subsequent to application and the citations of these papers.
With the bibliometric data of the applicants subsequent to application we were able to calculate the extent of over- and under-estimation (type 1 and type II errors) of the future success of the applicants.
The results of the regression analyses show that the mean scientific performance of approved applicants is higher subsequent to application than the mean performance of rejected applicants.
For an approved applicant, the expected scientific mean performance is increased by 31% (number of papers), by 41% (citations for papers published prior to application) and by 49% (citations for papers published subsequent to application) against a rejected applicant, holding all other variables in the models (models 2 and 3) constant.
The bibliographic data of the applicants' papers (published prior and subsequent to application) were taken from the SCI and were double-checked in the Medline database (provided by the National Library of Medicine, NLM, Bethesda, MD, USA) and with the applicants' lists of publications.
In type I error (also called false positive error), the EMBO selection committee concluded that an applicant had the scientific potential for promotion and was approved, when he or she actually did not, as reflected in an applicant's low scientific performance subsequent to application.
In the interpretation of the results of the regression analyses it cannot be ruled out that the applicants who received funding from EMBO may have published more subsequent to application because they received funding and not necessarily because the committee made the right choice about who received funding.
Nunome et al. (2001) also reported the contamination of Cd, Cu, Cr, Zn and Pb in soil and S. melongena subsequent to application of sugarcane pressmud and effluent.
In other words, we evaluated the committee's decisions with the following bibliometric indicators: (1) number of papers that were published subsequent to application, (2) citation counts for papers that were published prior and (3) subsequent to application.
To test whether indeed young scientists were selected for funding who subsequent to application developed better than the rejected ones requires a generally accepted criterion for scientific merit.
