tag:blogger.com,1999:blog-1582131107850597899.post86850619642710869..comments2012-10-03T02:31:23.396-07:00Comments on Get this bloody PhD done!: Potential selection bias problemsRobhttp://www.blogger.com/profile/11286219689849525803noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-1582131107850597899.post-68474277771378162322012-10-03T02:31:23.396-07:002012-10-03T02:31:23.396-07:00They gave up on the first one (possibly the easier...They gave up on the first one (possibly the easier of the 2)!<br /><br />I like the suggestion re how to proceed. Was my preference.<br /><br />Will chat to you later about the randomisation. Have a meeting with Karen K on Monday.Robhttps://www.blogger.com/profile/11286219689849525803noreply@blogger.comtag:blogger.com,1999:blog-1582131107850597899.post-24871661875461228492012-10-02T14:40:46.063-07:002012-10-02T14:40:46.063-07:00Attrition is a problem, you can do some sensitivit...Attrition is a problem, you can do some sensitivity analysis to look at the impact of missing data after the event but it is preferable to avoid attrition if possible. Which of the two tests did participants give up on, was it the same one? Does there need to be two tests? <br />My preference would be to suggest to participants that the tests are tricky but to give it a go and not to worry about low test scores and explain the relevance in relation to this study.<br />I'm not sure I agree with the person that said simple randomisation is likely to be sufficient, we've discussed this before I'd need to know the context of their recommendation for simple randomisation being sufficient.<br />HAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-1582131107850597899.post-1540255594854884262012-10-02T11:50:40.566-07:002012-10-02T11:50:40.566-07:00This comment has been removed by the author.Robhttps://www.blogger.com/profile/11286219689849525803noreply@blogger.com