Re: st: presenting odds ratios using log scale

[martine etienne <martieluvsu@yahoo.com>]

thanks Maarten,
I decided to use a table instead of a graph, in which I did put the CI values.



----- Original Message ----
From: Maarten buis <maartenbuis@yahoo.co.uk>
To: statalist@hsphsun2.harvard.edu
Sent: Mon, October 11, 2010 4:42:52 AM
Subject: Re: st: presenting odds ratios using log scale

--- On Sun, 10/10/10, martine etienne wrote:
> I'm trying to graph my odds ratios using a logarithmic
> scale but my graph looks very weird because the odds
> values are so small and because my yscale min is .01 
> 
> (crosses 1) and max is 10. and I have log values like .98
> and 1.08. Can anyone help me figure out how to graph these
> values in a way that look nice graphically.

Typically, I would do something like that by creating a 
"dataset" containing the odds ratios, and plot them using
-yscale(log)- or -xscale(log)- option. Often I need to 
specify nice values for the axis labels using the -ylab()-
or -xlab- option, as the default labels are usually not
nice for log scales.

This is not very specific advise, but I need more specific
information on what you want your graph to look like in
order to give you more specific advise. For example, if
you are plotting an interaction, then you typically want
a line graph, otherwise you often want a scatter plot.
Do you want to include confidence intervals?

Hope this helps,
Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany

http://www.maartenbuis.nl
--------------------------


      

*
*   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/



      
*
*   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/