Hello all, I would like to plan and analyse a study with "k" treatments (one of which is a "control"), with some binary outcome, in order to find the "best" treatment (e.g: the one with a high number of "successes"). If this was done with a fixed sample size, the analysis is well known. However, I would rather be able to "drop" treatment(s), if at any (or some specific) point in the analysis, I find it (or them) inferior to the control.
*What correction/analysis might I use in order to find the "best treatment", while dropping "bad treatments" during the experiment?* After searching through google scholar, the most relevant article I found was "Drop-the-losers design: Binomial case" by Michael W. Sill , Allan R. Sampson - yet I was not able to find an implementation for their ideas. Thanks up front for any lead/idea on this topic. (p.s: this question was also cross-posted to http://stats.stackexchange.com/questions/22355/suggestion-for-drop-the-loser-design-and-analysis-in-r ) With regards, Tal ----------------Contact Details:------------------------------------------------------- Contact me: tal.gal...@gmail.com | 972-52-7275845 Read me: www.talgalili.com (Hebrew) | www.biostatistics.co.il (Hebrew) | www.r-statistics.com (English) ---------------------------------------------------------------------------------------------- [[alternative HTML version deleted]] ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
