PubMed · 12210617
Fixed-effect versus random-effect models for evaluating therapeutic preferences.
Abstract
A preference trial is a special form of cross-over trial where clinical conditions determine when patients change treatment, in a prescribed order. This leads to binary responses with variable lengths. In cross-over trials with normal responses, patient effect may be treated as either fixed or random. However, with binary responses, random- and fixed-effect assumptions may lead to very different conclusions, so that one is no longer an alternative to the other.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Y Lee. 2002-08-30. Fixed-effect versus random-effect models for evaluating therapeutic preferences.. https://doi.org/10.1002/sim.1049. https://searchpubmed.org/record/pubmed%3A12210617/fixed-effect-versus-random-effect-models-for-evaluating-therapeutic-preferences
Citations retain the original authors and DOI and include this site’s record URL. If this discovery service contributed to your research, you can also cite Search PubMed by Indiciae: BibTeX · RIS.