Functions useful for producing estimates from data that were collected using network reporting techniques like network scale-up, indirect sampling, network reporting, and sibling history.
The goal of the
networkreporting package is to help people use network reporting methods.
In traditional survey methods, respondents report about themselves. In network report methods, survey respondents report about the people to whom they are socially connected. Thus, network reporting methods can be thought of a generalization of traditional surveys. Many existing methods rely on this kind of indirect reporting and are thus network reporting methods:
Although many of the methods that rely on indirect reporting have been regarded as distinct in the past, there is great value in developing a unified statistical framework to deal with common problems that all of these methods share. This unified framework also enables us to develop new network reporting methods. This package will help you use both existing network reporting methods and any new network reporting methods that you develop.
For more information on network reporting methods see:
The development of this software was supported by a grant from the National Institutes of Health (R01-HD075666).
You can install:
the latest released version from CRAN with
the latest development version from github with
networkreporting package enables you to use several methods that many people currently think of as distinct. Here are some vignettes for how to use the package:
If you would like to suggest a feature or report a bug, please create an issue
If you use our package for your research, please cite it so that we can continue to develop it.
moved C++ code and other bootstrap code into a separate package, called surveybootstrap
some new unit tests and vignette