R help - Aug 2016 - how to prepare my data to be used in the "survey" package

Looking for some help in using the "survey" package
I have gone through the documentation but still find problems in 
understanding how to prepare my data to be analyzed wih this package.

In particular I dont understand how to prepare the fpc data.
The documentation provides a "small survey example" where 6 variables 
are given:
stratid,psuids,weight,nh,Nh and x

My survey is a two stage stratified sampling so I have for each stratum 
the number of primary
and secondary units in the sample and in the population, and using these 
data I can prepare such
data frame, except for the last variable "x".
I cant understand what does it means and which information I have to put 
there.

A second (minor) problem is that my stratification is based on two 
variables, but it looks like it is only possible to give 1 variable as 
strata parameter.
Is it possible to give two ?

thanks in advance for any indication
David Remotti
Rome
Italy

> On Aug 29, 2016, at 9:48 AM, David Remotti <david at agros.it> wrote:
> 
> Looking for some help in using the "survey" package
> I have gone through the documentation but still find problems in
understanding how to prepare my data to be analyzed wih this package.
> 
> In particular I dont understand how to prepare the fpc data.
> The documentation provides a "small survey example" where 6
variables are given:
> stratid,psuids,weight,nh,Nh and x
> 
> My survey is a two stage stratified sampling so I have for each stratum the
number of primary
> and secondary units in the sample and in the population, and using these
data I can prepare such
> data frame, except for the last variable "x".
> I cant understand what does it means and which information I have to put
there.
> 
> A second (minor) problem is that my stratification is based on two
variables, but it looks like it is only possible to give 1 variable as strata
parameter.
> Is it possible to give two ?
This tutorial by Lumley illustrates both data preparation and the use of a
two-stage clustered design:

https://www.unc.edu/courses/2010spring/ecol/562/001/docs/lectures/lecture20.htm#svydesign


Found with a google search on "primary sampling unit secondary sampling
unit r survey package lumley"


-- 

David Winsemius
Alameda, CA, USA