Various people have made interesting suggestions, but
I wouldn't start on this with either box plots or Somers' d!
The information that this comes from Haiti is helpful.
I would guess wildly that there is a range from 1 up
for each variable, that each variable is very skewed and
values may easily go up to 20 or so.
A scatter plot would show lots of overplotting, unless
you used jittering. A graphical rendering of a twoway table
is available from -tabplot- from SSC.
Nick
[email protected]
Todd Swarthout
Im looking for a bit of advice on analysis of a database
from a household survey in Haiti.
I have one variable 'q5rooms'that tells how many rooms are in the house.
I have a second variable 'q22occups' that tells how many persons livein
the house.
As you might imagine, I am using these to build a picure of household
economic status.
I am trying to see if there is a proportional trend in the number of
rooms and the number of occupants. That is to say, do the people
reporting a larger number of rooms also report a larger number of
occupants.
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
