simul_confusion/Modules/ado/plus/b/batplot.hlp

{smcl}
{* 9May2008}{...}
{cmd:help batplot}
{hline}

{title:Title}

    {hi: Produces a Bland-Altman plot when there is a relationship between paired differences and their average}


{title:Syntax}

{p 8 17 2}
{cmdab:batplot} {it:varname1 varname2} [if] [in]
[{cmd:,} {it:options}]

{synoptset 20 tabbed}{...}
{synopthdr}
{synoptline}
{syntab:Main}
{synopt:{opt info:}} specifies that the percentage of points outside the limits of agreement are displayed as a subtitle.{p_end}
{synopt:{opt val:abel}({varname})} specifies that the points outside the limits of agreement be labelled using the variable {it:varname}.{p_end}
{synopt:{opt shading:}(min max)} specifies the extent of shading beyond the range of the data.{p_end}
{synopt:{opt notrend:}} specifies that the original Bland-Altman plot (without a trend) be plotted.{p_end}
{synopt:{opt moptions:}} specifies options for the markers that lie outside the limits of agreement.{p_end}
{synopt:{opt sc:atter}} specifies options for the scatter part of the final plot, see {help scatter}.{p_end}
{synopt:{help twoway_options}} specifies options for example titles and labels.{p_end}
{synoptline}
{p2colreset}{...}


{title:Description}

{pstd}
{cmd:batplot} produces a Bland-Altman plot adjusted for trend.

{pstd}
The standard Bland-Altman plot is between the difference of paired variables versus the average, this is produced
using the {bf:notrend} option. The main addition that this command handles is when there is a linear relationship between
the the paired difference and the paired average. A regression model is used to adjust the limits of agreement accordingly. 
This is particularly useful when the two variables might be measured on different scales and hence a straight conversion 
factor would recalibrate the two variables.

{title:Options}

{dlgtab:Main}

{phang}
{opt info:} specifies that the percentage of points outside the limits of agreement are displayed as a subtitle. Additionally
when using the notrend option the limits of agreement and mean difference are included in the subtitle.

{phang}
{opt val:abel}({varname}) specifies that the points outside the limits of agreement be labelled using the variable {it:varname}.

{phang}
{opt shading:}(min max) specifies the extent of shading beyond the range of the data. The default is that the limits of 
shading is determined by the values in the {hi:xlabel} option.

{phang}
{opt notrend:} specifies that the original Bland-Altman plot (without a trend) be plotted.

{phang}
{opt moptions:} specifies options for the markers that lie outside the limits of agreement, the options can be anything from
the scatter marker options {help scatter##marker_options}.

{phang}
{opt sc:atter} specifies options for the scatter part of the final plot.

{title:Examples}

{pstd}
Using the {hi:auto.dta} dataset supplied with STATA 8 this command can check whether there is agreement between 
turning circle (ft) and miles per gallon, click the highlighted text in order,

{phang}{stata sysuse auto, clear}

{phang}{stata batplot mpg turn}

{pstd}
This is the most basic graphic and using twoway options the look can be improved by clicking below,

{phang}
{stata batplot mpg turn, title(Agreement between mpg and turn) xlab(26(4)38) }

{pstd}
By specifying extra options the outlying points can be labelled and identified by the car make,

{phang}
{stata batplot mpg turn, title(Agreement between mpg and turn) info valabel(make) xlab(26(4)38) }

{pstd}
To obtain the original Bland Altman plot use the notrend option,

{phang}
{stata batplot mpg turn, title(Agreement between mpg and turn) info valabel(make) notrend xlab(26(4)38) }

{pstd}
To improve the labelling of the point VW it may be preferable to change the clock position of the label i.e.
labels could appear to the left of the point. This is handled below with {hi: moptions()}.

{phang}
{stata batplot mpg turn, title(Agreement between mpg and turn) info valabel(make) notrend xlab(26(4)38) moptions(mlabp(9))}

{pstd}
Additionally in the case of multiple scatter points by using the {hi:scatter()} option the user can specify to "{hi:jitter}" datapoints

{phang}
{stata batplot mpg turn, notrend xlab(26(4)38) moptions(mlabp(9)) sc(jitter(4))}


{title:Author}

{p}
Adrian Mander, MRC Human Nutrition Research, Cambridge, UK.

Email {browse "mailto:adrian.mander@mrc-hnr.cam.ac.uk":adrian.mander@mrc-hnr.cam.ac.uk}

{title:Also see}

Related commands

HELP FILES
{help baplot} (if installed)