> >Mark Fenton writes:
> >>Much of the statistics are presented confusingly. For
> >>example, in graphs which don't say anything and don't
> >>allow you to extract data from them; in trials where
> >>the drop out rate is greater than 50% and then don't
> >>report on which group they were in; trials which don't
> >>report on patients they randomised. I think what I'm
> >>getting at is any confusion is not all yours. Trialist's
> >>and Journals print articles which are confusing, (albeit
> >>some are improving)
Ian Bowns added:
> article From: Ian Bowns <[log in to unmask]>
> From the perspective of author, such apparent inconsistencies can arise
> when attempting to squeeze a lot of study into a little paper! The
> temptation is to avoid any repetition between tables and text, and even
> to put another variant of the summary statistics in the abstract. Major
> journals seem to be pushing authors to shorter and shorter papers,
> making it very difficult to put all the findings in one paper. This
> also causes difficulties for reviewers.
>
> Does anyone agree?
I think this works both ways: last week an editor asked me to
provide MORE details on the methodology and statistics (but to
economise on the background) for an article
Regards
Alan O'Rourke
______________________________________
Alan O'Rourke
Information Officer
Institute of General Practice
Community Science Centre
Northern General Hospital Sheffield S5 7AU
Tel: 0114 2714302 Fax: 0114 2422136
E-mail: [log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|