Braun Tibor, Schubert András (szerk.): Szakértői bírálat (peer review) a tudományos kutatásban : Válogatott tanulmányok a téma szakirodalmából (A MTAK Informatikai És Tudományelemzési Sorozata 7., 1993)
RUSTUM ROY: Alternatives to Review by Peers: A Contribution to the Theory of Scientific Choice
149 ROY: ALTERNATIVES TO REVIEW BY PEERS new members should receive one extra share for each new member equal to the average allocation for research for each member of the unit. If the newly appointed person is a replacement, there is presumably already enough money to pay the salary and the expenses of the research. This averaging scheme in my view is less perturbing to the pattern of any particular university than a fixed award for young scientists. An award of $50,000 per year for a new assistant professor who has recently been awarded the doctorate might not be unreasonable at the major universities, but it could cause jealousy at some others. A comparison of the scheme I propose with the existing system of awards on the basis of review by peers shows that the proposed scheme rests on a clear principle. It is a proportional award by ihe representatives of society for delivering the "products" those representatives desire. Moreover, it is based on recorded, quantifiable performance, not on promises made in an essay. The "best science" is defined as that which honours both scholarship and the public interest, and is "quantified" in the four terms of the formula. There are, moreover, different sets of peers for different evaluations. In most of these, the particular group of peers is much more broadly cognisant of other fields of the national interest, or industry's wants, than in the prevailing system of peer review. In addition, conflict of interest is virtually eliminated. Whereas in the system of peer review, a single negative review by one of the peers can block the entire research of a senior scientist with less than six months' notice, no single person can affect the immediate future of any individual drastically. This elimination of conflict of interest goes far to remove the corrosion of the integrity of the community caused by the present system. It goes much further than the commendable though mild effort by the National Institutes of Health in the warning statement which its reviewers are asked to note , but not even to sign and return. Because the recipient is not constrained by any proposal, or in ány fashion in the National Science Foundation, the National Institutes of Health or among his colleagues, he can follow any unforeseen lead or any unexpected opening. He can chose to do the most risky experiments in his genuinely atelistic work, i.e., work without a particular goal in view, by balancing it with work for a mission-oriented agency and industry. Dr Weinberg's proposal made 20 years ago that atelestic research should be an "overhead" on mission-oriented research will have been realised. 1 1 The magnitude of an individual's research grants will vary gradually over time, guaranteeing every scientist a certain proportion of the stable allocation of funds to do some of the difficult work requiring five or more years, which is now shunned for the quick production of papers. The beneficial feature of the formula is that it provides no sharp line of demarcation with respect to quality of individuals or universities. Thus it
