Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Comparing bootstrapped estimates across two models


From   Steven Samuels <[email protected]>
To   [email protected]
Subject   Re: st: Comparing bootstrapped estimates across two models
Date   Sat, 5 Feb 2011 09:53:36 -0500

-

Since you want to compare results from two commands run on the same data, you will need to write a single program that bootstraps both. For an example, see: http://www.ssc.wisc.edu/sscc/pubs/4-27.htm.

The quantity to bootstrap would be wtp1 - wtp2. I don't know anything about this field, but if WTPs are always positive or always negative, perhaps their ratios are more informative; if so, consider bootstrapping log(wtp1/wtp2).

I suggest that you base inference on the built-in bootstrap confidence intervals. If you want to do a hypothesis test, you might have to build your own; see e.g. http://portal.acm.org/citation.cfm?id=1280646 and an older paper: Peter Hall and Susan R. Wilson Two Guidelines for Bootstrap Hypothesis Testing, Biometrics Vol. 47, No. 2 (Jun., 1991), pp. 757-762.

Steve
[email protected]



On Feb 4, 2011, at 12:47 PM, Matthew Wibbenmeyer wrote:

Hi all,

I'm engaged in a choice modeling study in which wildfire managers were
asked to select their **preferred** management strategy and their
**expected** strategy  given current political and social constraints.
I want to compare willingness-to-pay values derived from two models
with different dependent variables: choice_expected and
choice_preferred. WTP values are ratios between coefficients so they
have no distribution and a distribution must be estimated using
bootstrapping (or parametric bootstrapping).

My question has two parts:
1) What formal test can I use to test the significance of the
difference between bootstrapped WTP estimates from two models with
different dependent variables?
2) How can I implement this following my bootstrap commands, which
look like this:

bootstrap wtp1 = ((-1)*(_b[homes]/_b[cost])), reps(100) seed(12345)
cluster(obsid) dots: ///
clogit choice_exp homes watershed aviationhours grounddays duration cost ///
, group(obsid)

bootstrap wtp2 = ((-1)*(_b[homes]/_b[cost])), reps(100) seed(12345)
cluster(obsid) dots: ///
clogit choice_pref homes watershed aviationhours grounddays duration cost ///
, group(obsid)

Thanks in advance for any help anyone can offer!

Matt Wibbenmeyer
USDA Forest Service
*
*   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/

*
*   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/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index