R help - Jun 2018 - Multinomial Logistic Regression with Complex Survey using 'Survey' Package in R

Dear R Users,

I want to use a multinomial logistic regression model with survey data in the
?survey? package. The original package did not have a function for multinomial
logistic regression, so Thomas Lumley suggested creating replicate weights for
the survey and doing a multinomial regression with frequency weights in the
mlogit package. See the below message for reference:

There isn't an implementation of multinomial regression in the survey
package.  The easiest way to do this would be to create replicate weights for
your survey if it doesn't already have them (with
as.svrepdesign()) and then use withReplicates() to do the regression using a
function that does multinomial regression with frequency weights, such as
mlogit() in the mlogit package.  The example on the withReplicates() help page
shows how to do this for quantile regression, and it should be similar.

However, there has been a more recent release of the ?survey? package in May
2018, so I am wondering if there is now a function that does multinomial
logistic regression with the survey. Please let me know if anyone knows of this
update, or has any additional advice on how to perform this function.

Thank you,
Mackenzie


	[[alternative HTML version deleted]]

hi, check out the news page..
https://cran.r-project.org/web/packages/survey/NEWS

On Tue, Jun 19, 2018 at 5:54 PM, Mackenzie Jones <mackenzietjones at
gmail.com>
wrote:
> Dear R Users,
>
> I want to use a multinomial logistic regression model with survey data in
> the ?survey? package. The original package did not have a function for
> multinomial logistic regression, so Thomas Lumley suggested creating
> replicate weights for the survey and doing a multinomial regression with
> frequency weights in the mlogit package. See the below message for
> reference:
>
> There isn't an implementation of multinomial regression in the survey
> package.  The easiest way to do this would be to create replicate weights
> for your survey if it doesn't already have them (with
> as.svrepdesign()) and then use withReplicates() to do the regression using
> a function that does multinomial regression with frequency weights, such as
> mlogit() in the mlogit package.  The example on the withReplicates() help
> page shows how to do this for quantile regression, and it should be
similar.
>
> However, there has been a more recent release of the ?survey? package in
> May 2018, so I am wondering if there is now a function that does
> multinomial logistic regression with the survey. Please let me know if
> anyone knows of this update, or has any additional advice on how to perform
> this function.
>
> Thank you,
> Mackenzie
>
>
>         [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/
> posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
	[[alternative HTML version deleted]]

Did you download the new package and check? Is there some reason you
shouldn't do this in any case?

There is also usually a News file in the package download that tells you
about new features.

-- Bert



Bert Gunter

"The trouble with having an open mind is that people keep coming along and
sticking things into it."
-- Opus (aka Berkeley Breathed in his "Bloom County" comic strip )

On Tue, Jun 19, 2018 at 2:54 PM, Mackenzie Jones <mackenzietjones at
gmail.com>
wrote:
> Dear R Users,
>
> I want to use a multinomial logistic regression model with survey data in
> the ?survey? package. The original package did not have a function for
> multinomial logistic regression, so Thomas Lumley suggested creating
> replicate weights for the survey and doing a multinomial regression with
> frequency weights in the mlogit package. See the below message for
> reference:
>
> There isn't an implementation of multinomial regression in the survey
> package.  The easiest way to do this would be to create replicate weights
> for your survey if it doesn't already have them (with
> as.svrepdesign()) and then use withReplicates() to do the regression using
> a function that does multinomial regression with frequency weights, such as
> mlogit() in the mlogit package.  The example on the withReplicates() help
> page shows how to do this for quantile regression, and it should be
similar.
>
> However, there has been a more recent release of the ?survey? package in
> May 2018, so I am wondering if there is now a function that does
> multinomial logistic regression with the survey. Please let me know if
> anyone knows of this update, or has any additional advice on how to perform
> this function.
>
> Thank you,
> Mackenzie
>
>
>         [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/
> posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>
	[[alternative HTML version deleted]]