Re: st: -baplot- output

[Nick Cox <njcoxstata@gmail.com>]

-baplot- (STB-55) is a user-written program. Please remember that you
are asked to explain where user-written programs you refer to. Once
again,

1. This gives detailed context on your problem to people who might be
interested in trying to help you.

2. It stops people wasting time, especially new or naive users, by
typing -help baplot- and then wondering why it doesn't work.

3. It tells people about a program that may be useful to them.

4. It helps remind any program authors who may have forgotten about
some of their progeny.

5. It gives appropriate credit to those program authors, at least indirectly.

I can think of no reason not to do this. I assert without fear of
contradiction that there is no user-written program that absolutely
all Statalist members know about.

In this particular case, try -concord- (SJ 10-4) which will add that
regression line painlessly.

Proportional error is surely best examined after transformation to logarithms.

You can and should -update- from 10.0 to 10.1.

Nick

On Sat, Jan 14, 2012 at 3:38 AM, cecilia sam <samcecilia2010@gmail.com> wrote:
> Hi all,
>
> I am using -baplot- and don't know how to put the regression line on
> the plot, and get the p-value of the regression line.
>
> I am using Stata 10.0 for Windows. I typed -baplot- and have no
> problem to obtain the mean difference and the limit of agreement. I
> want to do further analysis to test whether there is proportional
> error.
>
> Is there a way to put the regression line onto the Bland and Altman
> plot, and also obtain the significance of the regression slope by
> using -baplot-?
>
> Or is there someway to test whether there is proportional error using
> the Bland and Altman method.
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/