Nongye tushu qingbao xuebao (Feb 2022)

The Validity of Peer Review Results of DEA Based Super Efficiency Projects

  • WAN Hao, ZHANG Fujun, LV Qianqian

DOI
https://doi.org/10.13998/j.cnki.issn1002-1248.21-0528
Journal volume & issue
Vol. 34, no. 2
pp. 88 – 101

Abstract

Read online

[Purpose/Significance] This article uses some projects'peer review results as an example to explore the effectiveness of peer review based on expert knowledge. [Method/Process] It uses data envelopment analysis to measure the R&D efficiency of the projects. The evaluation model is a CCR model modified by super-efficiency, and the efficiency index is used as a verification standard for the effectiveness of peer review results. The data source is 126 "948 projects" implemented by the State Forestry Administration, covering 13 input and output evaluation indicators. [Results/Conclusions] The experiment found that there was a moderate positive correlation between the project peer review results and the actual R&D efficiency (DEA super efficiency), and the Spearman correlation coefficient ρ=0.250 (Sig.<0.01). The percentages of type I errors and type II errors made by subjective review were 17.5% and 18.2%, respectively, and effective peer review decisions accounted for 64.3%. It is confirmed that there are certain systematic errors in peer review, but they are inevitable.

Keywords