| Abstract: | The validity of indices of departmental research performance based on three different types of peer review is critically examined. The three types of peer review under investigation are: journal peer review, grant-giver peer review, and impressionistic peer review. It is shown that, of the indices considered, only those based on journal peer review constitute true performance indicators that are capable of yielding a reasonably valid measure of departmental performance. A basic requirement of a performance indicator is that it should relate output to input. None of the indices based on grant-giver peer review or impressionistic peer review satisfies this criterion. They are therefore unable to provide the information that the UGC seeks to obtain in its 1989–90 evaluation, namely, an indication of the cost-effectiveness of departmental research programmes. On both a priori and empirical grounds, indices based on grant-giver peer review and impressionistic peer review are shown to yield unsatisfactory measures of departmental research performance. Fifteen serious defects are identified in indices based on grant-giver peer review, and nine major flaws are highlighted in indices based on impressionistic peer review. Implications for the conduct of the UGC's 1989–90 evaluation are discussed. |
