ABSTRACT
The connection between grant funding and research productivity has not been well established.
Objective to examine the impact of grant funding on the publication activity of awarded applicants.
Methods a systematic review of results from comparative studies on the publication activity of applicants (awarded vs rejected) both prior to and after the award process. All pool estimates (weighted mean difference) were based on random-effects models.
Results revealed 16 relevant publications (grant funding from 14 funds, 1980 to 2007 years), all with results from quasi-experimental studies. 45 paired values (ex ante – ex post) for the number of articles published by awarded and rejected applicants were used in the quantitative synthesis. The median average publication activity of awarded applicants before the award process was 2.4 (1.3; 3.4) and after the award process 3.1 (1.7; 4.3) publications per year, for rejected applicants was 1.8 (1.0; 2.9) and 2.4 (1.1; 3.8) respectively. The summation of the results from these studies using the difference-in-differences approach showed that awarded applicants published 0.14 articles per year (95% Cl 0.07 to 0.21) more than rejected applicants (adjusted for publication bias). A meta-regression analysis made it possible to tie together the revealed small difference with the difference-in-differences approach bias − the subsequent differences in the groups are determined by the scale of the initial differences in their publication activity.
Conclusion awarded applicants published slightly more often than their rejected opposites. However, this effect may be the result of a bias caused by the shortcomings of the difference-in-differences approach.