The AIBS dataset (Gallo, 2020) comes from the scientific
peer review facilitated by the American Institute of Biological Sciences (AIBS)
of biomedical applications from and intramural collaborative biomedical research
program for 2014--2017. For each proposal, three assigned individual reviewers were
asked to provide scores and commentary for the following application criteria:
Innovation, Approach/Feasibility, Investigator, and Significance (Impact added
as scored criterion in 2014). Each of these criteria is scored on a scale from
1.0 (best) to 5.0 (worst) with a 0.1 gradation, as well as an overall score
(1.0--5.0 with a 0.1 gradation). Asynchronous discussion was allowed, although
few scores changed post-discussion. The data includes reviewers' self-reported
expertise scores (1/2/3, 1 is high expertise) relative to each proposal reviewed,
and reviewer / principal investigator demographics. A total of 72
applications ("Standard" or "Pilot") were reviewed in 3 review cycles. The
success rate was 34--38%. Application scores indicate where each application
falls among all practically possible applications in comparison with the
ideal standard of quality from a perfect application. The dataset was used by
Erosheva et al. (2021a) to demonstrate issues of inter-rater reliability in
case of restricted samples. For details, see Erosheva et al. (2021b).
data(AIBS)AIBS is a data.frame consisting of 216 observations on
25 variables. Data describes 72 proposals with 3 ratings each.
Proposal ID.
Year of the review.
Proposal type; "Standard" or "Pilot".
Anonymized ID of principal investigator (PI).
PI's organization type.
PI's gender membership; "1" females, "2" males.
PI's rank; "3" full professor, "1" assistant professor.
PI's degree; "1" PhD, "2" MD, "3" PhD/MD.
Innovation score.
Approach score.
Investigator score.
Significance score.
Impact score.
Scientific merit (overall) score.
Average of the three overall scores from three different reviewers.
Average of the three overall scores from three different reviewers, increased by multiple of 0.001 of the worst score.
Project rank calculated based on ScoreAvg.
Project rank calculated based on ScoreAvgAdj.
Reviewer's ID.
Reviewer's experience.
Reviewer's institution; "1" academia, "2" government.
Reviewer's gender; "1" females, "2" males.
Reviewer's rank; "3" full professor, "1" assistant professor.
Reviewer's degree; "1" PhD, "2" MD, "3" PhD/MD.
Reviewer code ("A", "B", "C") in the original wide dataset.
Gallo, S. (2021). Grant peer review scoring data with criteria scores 10.6084/m9.figshare.12728087
Erosheva, E., Martinkova, P., & Lee, C. (2021a). When zero may not be zero: A cautionary note on the use of inter-rater reliability in evaluating grant peer review. Journal of the Royal Statistical Society - Series A. Accepted.
Erosheva, E., Martinkova, P., & Lee, C. (2021b). Supplementary material: When zero may not be zero: A cautionary note on the use of inter-rater reliability in evaluating grant peer review.