Biostatistics Seminar: Daniel Tancredi, Pediatrics, UC Davis
DATE: Tuesday, May 14th, 2013
LOCATION: 1043 Gladys Valley Hall (map)
SPEAKER: Daniel J. Tancredi, Dept. Pediatrics, School of Medicine, UC Davis
TITLE:Practical considerations for sample size calculations in clinical research
ABSTRACT: “How many subjects do I need?” is a common question asked of medical statisticians. Appropriate power and sample size calculations may promote the scientific, ethical and economic soundness of clinical research studies. However, many clinical research studies are undertaken with seriously defective or even no power calculations. In this seminar, we will identify key design and analysis features that affect the statistical power of a study to detect clinically meaningful effects but that are often neglected. We will also illustrate some practical methods for accounting for such features, including longitudinal measurement, cluster randomization, unknown values for nuisance parameters, and the use of generalized linear models.(1-2) We will also discuss the “sample size samba” and some of the other flaws and potential harms that arise from contemporary sample size conventions.(3)
- Lachin JM. Maximum information designs. Clinical Trials. 2005;2(5):453-64.
- Lyles RH, Lin HM, Williamson JM. A practical approach to computing power for generalized linear models with nominal, count, or ordinal responses. Stat Med. 2007 Mar 30;26(7):1632-48.
- Bacchetti P. Current sample size conventions: Flaws, harms, and alternatives. BMC Med. 2010 Mar 22;8.