Re: Re: Re: st: replicating 2 X 2 data from a paper

[Clyde B Schechter <clyde.schechter@einstein.yu.edu>]

In response to a question from Ariel Linda, David Hoaglin wrote, mostly as an afterthought:

"As a statistician, I think you expect too much of the "statistical
reviewers" or of peer review more generally.  Many papers that I read
contain rough edges and errors that, to my amazement, the authors and
the reviewers somehow failed to notice."

Here's my two cents on this generic issue.  Nearly all the journals I have dealth with have stringent word limits for articles.  It is almost never the case that a truly satisfactory description of the methods and important results can be accommodated within those limits, and when the time comes to cut, it is usually the methods that get the meat ax, at least in clinincal journals.  

It may well be that the authors provided a fuller explanation to the reviewers "behind the scenes" but it never made it into the article because of space limitations.  I have experienced this phenomenon myself both as an author and as a reviewer.

My hope and expectation is that as publication moves away from paper and comes to rely more on electronic formats these word limits, relics attributable to the expense of paper, will disappear (or at least greatly relax) and that on-line articles will include much better descriptions of methods.  It seems to be happening already: I have encountered several articles where the hard copy gives the "highlights" of the methods, but an "on-line appendix" has all the detail one could ask for.


Clyde Schechter
Dept. of Family & Social Medicine
Albert Einstein College of Medicine
Bronx, NY, USA



*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/