Streamlining Grant Applications - What are the probabilities a streamlined grant is fundable and that a fundable grant is streamlined?

Read the full article

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Background: Securing qualified peer reviewers for public granting agencies is challenging and to avoid needlessly overworking these volunteers, there is increasing reliance on triaging grants deemed unlikely to be competitive. For example, Canadian Institute for Health Research (CIHR), our largest public funder, has a stream-lining process whereby if the average of three initial reviewer scores is < 4.0, the submission is automatically triaged.Methods: Using Bayesian methods to account for the uncertainty in the re-viewers’ score, small sample sizes and vague prior beliefs, the true mean posterior(updated) distribution of a streamlined grant was calculated, including the probability it exceeds the threshold. The predictive distribution a new single reviewer, or a group of reviewers, might accord the streamlined grant while incorporating the aforementioned uncertainties was also calculated. Finally, the probability a submission with a true value > 4 could be streamlined was also calculated.Results: Simulations suggest that grants with mean scores slightly below 4 may have significant probabilities (30-50%) of being fundable, when reviewer variability is high. The mean probability that a new reviewer would rate a previously streamlined grant with a score > 4 was 48% (95% confidence interval (CI)0.13 - 0.86). The average probability that 15 new reviewers would rate the sestreamlined grants > 4 was 44% (95% CI 7 - 93). Simulations suggest that even streamlined grants with scores of 3.7 or 3.8, if associated with high reviewer variations, may have at least a 30% probability that the next reviewer score is > 4. Simulations also show that fundable grants, those with true scores in the range of 4.1 - 4.3, have a 20-40% probability of being streamlined.Conclusions: This study emphasizes the need to reevaluate streamlining mechanisms to increase the fairness and accuracy of the evaluation process. Agencies should consider adjusting thresholds or requiring additional reviews for streamlined applications with high variability to avoid excluding potentially fundable grants from full committee consideration.

Article activity feed