Re: st: continuous data measured as ordinal catogories

[Richard Williams <Richard.A.Williams.5@ND.edu>]

At 05:05 PM 4/8/2010, Shehzad Ali wrote:
> Hi listers,
>
> This question is probably beyond the scope of this list (so 
> apologies in advance). I am dealing with an odd dataset in which 
> willingness-to-pay has been measured using the following categories:
>
> $1 per month
> $5 per month
> $10 per month
> Any other amount/month
>
> I find this data very odd for the purpose of analysis (regression or 
> other statistical methods). I want to model the relationship of 
> willingness-to-pay and respondent characteristics. I wonder if 
> someone on the list has any thoughts.

That last category is really weird.  Does it include both less than 
$1 and more than $10? Do people maybe write in what the amount is?

If you want it as an independent variable, my first impulse is to 
break it into dummies.  Also, Long & Freese's Regression Models for 
Categorical Dependent Variables Using Stata shows how to test whether 
it is legit to treat an ordinal independent variable as a continuous 
variable.  See pp. 421-422.  Basically, you run a model with the 
single variable treated as continuous, and then a model with the 3 
dummies, and then test whether the 3 dummies approach provides a 
significantly better fit than the single variable approach.  If not, 
then treat the var as continuous.

If you want it as a dependent variable, you could use mlogit -- or 
possibly one of the ordinal regression techniques if you think the 
last category always means more than $10 (or less than $1).


-------------------------------------------
Richard Williams, Notre Dame Dept of Sociology
OFFICE: (574)631-6668, (574)631-6463
HOME:   (574)289-5227
EMAIL:  Richard.A.Williams.5@ND.Edu
WWW:    http://www.nd.edu/~rwilliam

*
*   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/