What if you narrow the confidence intervals? For example, how
about using an 80% CI instead of a 95% CI? (just for clarifying
the ternary diagram) Also, you can modify it so that you
produce different types of symbols/colors for each point.
----- Original Message -----
From: "Andrew Reeves" <reeves(a)fas.harvard.edu>
To: "gov2001" <gov2001-l(a)fas.harvard.edu>
Sent: Thursday, May 08, 2003 4:19 PM
Subject: [gov2001-l] Histograms for Ordered Probit First
Differences?
I've generated first differences for an ordered
probit model
with three
values (0,1,2). Is it okay to present
compare/histograms of
the
probabilities for each value (0,1,2) as opposed to
ternary
plot? Each
probability has rather large confidence intervals
which make
the ternary
less clear than it would be otherwise.
Any suggestions would be appreciated!
Andrew
--
Andrew Reeves
reeves(a)fas.harvard.edu
617.493.3485 tel.
301.639.8369 cell.
http://people.fas.harvard.edu/~reeves/
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l