Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
| From | Austin Nichols <austinnichols@gmail.com> |
| To | statalist@hsphsun2.harvard.edu |
| Subject | Re: st: spike model for DB CV survey |
| Date | Sun, 25 Sep 2011 08:06:08 -0400 |
Kifle Wondemu <kiflea@gmail.com> : http://fmwww.bc.edu/repec/bocode/d/doubleb.html indicates that this is an interval regression model assuming normality. It seems official -intreg- could be used to estimate this model. If the unobserved true outcome (WTP) cannot be negative, then the assumption of normality is untenable, and you should specify an alternative distribution when you write out a -ml- model. Or findit censored poisson On Sun, Sep 25, 2011 at 12:01 AM, Kifle Wondemu <kiflea@gmail.com> wrote: > Hi folks, > > I want to use a spike model to deal with zero response for double > bounded dichotomous choice contingent valuation survey. The existing > DOUBLEB user written program, if I am correct, does not differentiate > genuine zeros from the protest ones. When I use doubleb on the whole > sample, I am getting a negative willing to pay which is not > theoretically tenable in my case. > I would appreciate if anyone is aware of any user written program that > can resolve my problem. > Thanks > Kifle Asfaw * * 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/