• (4)  

    The committee did not investigate systematically and carefully the advantages and disadvantages of alternative mechanisms of predoctoral support. The only factual evidence pertinent to this issue (presented in chapter 5) comes from a 1984 Research Council study, The Career Achievements of NIH Predoctoral Trainees and Fellows. This study explicitly stated that "it cannot be determined whether [trainees'] superior records of achievement may be attributed to the selection process, the training they received, or a combination of these and other factors." Thus, any conclusion drawn from this study that training grants are a more effective training mechanism than research grants is unfounded.

  • (5)  

    The report's stated preference for training grants over research grants is not based on hard evidence of superiority, but rather on the opinions of individual committee members "with direct experience with training grants". Since the study charge does not encompass an evaluation of alternative mechanisms for graduate student support, it is not surprising that a majority of the committee do not have such ''direct experience". They are therefore not in a position to make independent judgments about the relative merits of these two training mechanisms and were not appointed with this task in mind.

  • (6)  

    The advantages and disadvantages of alternative support mechanisms were never fully discussed by the committee. Had the study called for a comparison of alternative mechanisms for predoctoral support, a much more detailed analysis would have been required, including an examination of the cost implications for different institutions and federal sponsors. (NIH training grants do not pay full indirect costs, while research grants do; and training grants also limit trainees' tuition reimbursement to the university.)

  • (7)  

    The proposal to substitute traineeships for research assistantships presents a particular problem for institutions that do not have training grants, yet have faculty members who are successful in obtaining NIH research awards. These investigators would be unable to make the recommended substitution, yet the quality of their research can be assumed to be as good as the research funded at universities that do have training grants.

  • (8)  

    From the perspective of federal policy-makers, the recommendation to increase training grant support may appear nonsensical—especially in light of the overwhelming evidence that universities are already training too many PhDs for the research positions available. Why should Congress appropriate more funds for training grants when there is already an overabundance of trained life scientists?

I want to emphasize that I have these reservations about the training-grant recommendation because of the totally inadequate evidential basis for the recommendation and because of the consequences it would have—not because I hold strong views on the intrinsic merits of either training grants or research assistantships. For several years, I chaired the aforementioned Committee on National Needs for Biomedical and Behavioral Research Personnel, which recommended annually to Congress the number of training-grant positions to be supported under the National Research Service Awards Act. Earlier, I served as associate director of the National Science Foundation with particular responsibility for the education and training of scientists (in all scientific disciplines). These experiences have made me keenly aware of the difficulty of making a valid comparison between alternative support mechanisms, as well as the multiple difficulties of implementing the changes recommended in this report. Without considerably more evidence on the relative merits of alternative mechanisms for supporting graduate students, a recommendation to increase training grants and

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement