[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
st: How to correct the sampling frame?
I am working on a small country that has not had a reliable census since 1981.
A household budget survey is administered every year, as a sample of 300 EAs is drawn every year from the 1981 frame. Then the EAs are fully relisted and 2400 households are drawn with pps. The results indicate that our sampling frame is not representative -- population drops for about 20% in 3 years and household size also decreases significantly.
I can also obtain the data on the relisted 1600 EAs over the last 4 years (a total of 400 EAs are drawn each year and only 300 used for the survey).
Could you recommend studies that could address this problem? So far I have zeroed down the Peter Lynn's paper in Statistics in transition about Bosnia & Herzegovina (the method is to assume 1600 EAs are randomly drawn from the population and then adjust the weihgts accordingly) and Demery & Grooert's paper on Burkina Faso of adjusting the weights to make the distribution of household size across years the same.
Any ideas, especially from other disciplines, would be greatly appreciated.
* For searches and help try: