Filed Under : Dougherty, Edward Rusell | 2012 | Feature Selection | Genomic Applications | BioEssays
Type: Journal Publication
Web: http://onlinelibrary.wiley.com/doi/10.1002/bies.201200003/abstract
Abstract: In searching for biomarkers, one might maintain that it is prudent to have preliminary studies based on small samples. Given a satisfactory result, a large-sample follow-on study can then check reproducibility. However, is this two-fold approach, namely the decision to conduct the small study and the subsequent decision to spend significant resources on a large study, actually prudent? There is the risk that in the first study nothing at all might be discovered while in the second study the first results are not reproducible
Cited as: Dougherty, E. R., "Biomarker development: Prudence, risk, and reproducibility", BioEssays, Vol. 34, No. 4, 277-279, 2012