tag:blogger.com,1999:blog-15654301.post8707261429744114874..comments2024-02-24T04:43:56.829-05:00Comments on On Biostatistics and Clinical Trials: FDA Issued Draft Guidance for Industry for HIV drug "Human Immunodeficiency Virus-1 Infection: Developing Antiretroviral Drugs for Treatment"Web blog from Dr. Denghttp://www.blogger.com/profile/11917138094035874938noreply@blogger.comBlogger4125tag:blogger.com,1999:blog-15654301.post-83982261455696287592021-04-27T05:25:56.342-04:002021-04-27T05:25:56.342-04:00niceniceVishal Sharmahttps://www.blogger.com/profile/08861709834808119636noreply@blogger.comtag:blogger.com,1999:blog-15654301.post-40691356540186927462015-11-03T21:55:01.115-05:002015-11-03T21:55:01.115-05:00for good statistical practice, the list of covaria...for good statistical practice, the list of covariates should be pre-specified in the SAP. <br />For mixed model, the covariate structure should also be pre-specified in SAP. However, the SAP does not need to be finalized prior to the collection of data. In my opinion, it will be fine if you change the covariate structure in the SAP prior to the unblinding of the treatment assignment. Web blog from Dr. Denghttps://www.blogger.com/profile/11917138094035874938noreply@blogger.comtag:blogger.com,1999:blog-15654301.post-34607650454418325212015-11-03T11:22:04.494-05:002015-11-03T11:22:04.494-05:00This comment has been removed by a blog administrator.Redmanhttps://www.blogger.com/profile/08832521283897349544noreply@blogger.comtag:blogger.com,1999:blog-15654301.post-41149588055825346012015-11-03T11:21:28.615-05:002015-11-03T11:21:28.615-05:00If I were to use a forward selection method to sel...If I were to use a forward selection method to select my covariates for a mixed model (repeated measures) in my analysis, would I need to list out all the prospective covariates in the SAP?<br /><br />Also, regarding the use of covariance structure in the mixed model, would I need to describe which one I intend to use, prior to the collection of data? This might be a little bit more difficult as the nature of the data would drive this decision.<br /><br />Redmanhttps://www.blogger.com/profile/08832521283897349544noreply@blogger.com