Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | Steve Samuels <sjsamuels@gmail.com> |
To | statalist@hsphsun2.harvard.edu |
Subject | Re: st: Sample Wegihts |
Date | Wed, 10 Mar 2010 09:42:05 -0500 |
Stas's is absolutely correct. Probability weights must be based on the overall probability of selection. The same principle is the basis of network sampling. A few years ago, a telephone survey company supplied us two sets of weights, because they had taken a second survey to bring the sample size up. I learned that this had been standard practice for years. Not surprisingly, their staff did not include a statistician. The second set of weights would have been correct if they had added a new stratum to the population and done the second survey in that stratum only. -Steve On Wed, Mar 10, 2010 at 12:47 AM, Michael Lichter <mlichter@fastmail.fm> wrote: Stas said: >>overall P[ selection ] = P[ to be selected in the first sample ] + P[ to be selected in the second sample ] - P [ to be selected in both ] = 1 - (1-P[first])*(1-P[second]) Is this correct even with if the first sample is an SRS and the second is clustered? I can't show otherwise, but it doesn't feel right that every case in a specific urban area should have the same weight regardless of which sample it was drawn from. -- Steve Samuels sjsamuels@gmail.com 18 Cantine's Island Saugerties NY 12477 USA 845-246-0774 * * 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/