Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | Nick Cox <n.j.cox@durham.ac.uk> |
| To | "'statalist@hsphsun2.harvard.edu'" <statalist@hsphsun2.harvard.edu> |
| Subject | st: RE: quick question about -concord- |
| Date | Fri, 13 Jan 2012 18:14:53 +0000 |
-concord- is a user-written program, last updated in SJ 10(4) 2010. Please remember to explain where user-written programs you refer to come from. Thomas Steichen was the senior author, but he's now retired, so I will speak as the young man on the team. You are right: -concord Y X, loa(noref)- plots Y - X vs (Y + X)/2. However, your reviewer may not be convinced just by the program author's say-so. There are easy demonstrations: sysuse auto gen mpgp2 = mpg + 2 concord mpg2 mpg, loa On Ricardo's hypothesis, and mine too, this graph should show a horizontal line at +2, and it does. However, the default wording on the axis plotting Y - X is "Difference of ... and ..." and could fairly be described as ambiguous. It was clearly intended to match the default wording of "Mean of ... and ...". Perhaps the thought was that for some of the likely readership + and - signs were a bit too mathematical.