THE PREDICTIVE ABILITY OF PEER-REVIEW OF GRANT PROPOSALS - THE CASE OF ECOLOGY AND THE UNITED-STATES-NATIONAL-SCIENCE-FOUNDATION

Analyses of temporal correlations in measures of scientific productivity and quality suggest that individuals who have produced good science in the recent past will continue to do so in the near future. However, previous work has shown little correlation between measures of the quality of previous publications by a scientist and the ratings received by that individual's research proposals. This suggests that evaluating previously published work may be a better means of predicting the quality of future work than is evaluating research proposals. Possible explanations for this finding are explored, using the Ecology Panel of the US National Science Foundation as an example.