St. James Ballroom
Improving Reproducibility via Experimental Design in the Life and Social Sciences (303875)
*Ben George Fitzpatrick, Loyola Marymount UniversityKeywords: experimental design, reproducible research
Reproducibility of preclinical animal research has become a source of great concern in recent years. A number of studies have demonstrated difficulties in replicating high-profile results. To address these concerns, NIH has developed new Rigor and Reproducibility guidelines for the grant applicants and reviewers. Among other things, these guidelines call for careful development of the experimental design.
As part of an NIH-sponsored project on improving access to experimental design expertise, we interviewed 131 academic and industry researchers, program managers from government and private foundations, and regulatory and administrative personnel involved in animal research for basic, preclinical, and clinical studies, to find out their concerns about reproducibility in general and experimental design challenges in particular.
Among the more problematic issues uncovered in these interviews is choosing sample sizes. We are working with multiple teams of biomedical faculty, postdocs, and students to develop tools for understanding things like power and effect sizes that are crucial for designing experiments in a statistically rigorous manner.