Double Bounded WTP Coding issues

[Soumi Roy Chowdhury]

Hello Everyone,
I do have the following question on interval regression.

In my research project, I have a double bounded willingness to pay data for cancer screening technology and I am trying to run an Interval regression model ( intreg ).
I am facing some difficulty in coding the maximum WTP limit for those individuals who have responded "Yes" and "Yes" to both my initial and follow up bids.
( Technically they should be coded as infinity but how can I model it?)

P.S: I do not have any open ended question in my data which asks for the maximum willingness to pay.

Any help will be highly appreciated.

[Carlo Lazzaro]

Soumi:
reply #3 at http://www.statalist.org/forums/foru...normality-test may help.
I will also search on and off the list for Arne Risa Hole's contributions on WTP.

[Jeff Wooldridge]

Hello Everyone,
I do have the following question on interval regression.

In my research project, I have a double bounded willingness to pay data for cancer screening technology and I am trying to run an Interval regression model ( intreg ).
I am facing some difficulty in coding the maximum WTP limit for those individuals who have responded "Yes" and "Yes" to both my initial and follow up bids.
( Technically they should be coded as infinity but how can I model it?)

P.S: I do not have any open ended question in my data which asks for the maximum willingness to pay.

Any help will be highly appreciated.

This answer is really off the cuff, but in the intreg command you indicate that a response is above the upper bound by setting the variable, say "upper", to missing. For example, if the cost is 40, the answer is yes, and the next cost is 60, and the answer is yes, I believe you would use lower = 60 and upper = . for this respondent.

JW

[Soumi Roy Chowdhury]

Thank you Professors so much for responding. I really appreciate.