Using "Design Effect" weight variable... PreviousNext
Mplus Discussion > Multilevel Data/Complex Sample >
 Anonymous posted on Wednesday, August 24, 2005 - 11:06 am
In many datasets produced by NCES (ECLS-K, for example), they provide weight variables based on Taylor series. Is there a way to incorporate those variables in Mplus runs? Where do I specify and how? Thanks!
 Linda K. Muthen posted on Wednesday, August 24, 2005 - 2:47 pm
You can use the WEIGHT option of the VARIABLE command to incorporate a sampling weight in the analysis.
 Anonymous posted on Wednesday, August 24, 2005 - 6:32 pm
Thank you Linda!

NCES used "jackknife 2 method" to produce a series of 90 replication weights. They recommend AM software that reads in all the 90 replicate weights to obtain correct standard errors for significance tests. Is there a way we can do it with Mplus?
 Linda K. Muthen posted on Wednesday, August 24, 2005 - 6:46 pm
This cannot be done in Mplus as far as I know.
 Anonymous posted on Wednesday, August 24, 2005 - 8:04 pm
Thank you anyway!
 Kelly McGinn posted on Friday, May 01, 2015 - 6:21 am
Hello Drs. Muthen,

I am working with data that has high ICCs but cannot use MLM as I only have eight clusters. It was suggested to me to use Taylor Series Linearization for variance estimation. Can this be done in MPlus? I have not come across this in the manual. I currently only have access to MPlus 7.3 Base Program and Combination Add-on and the standard version of SPSS.
 Bengt O. Muthen posted on Friday, May 01, 2015 - 1:45 pm
I don't think that method would work well with only 8 clusters. See my other reply.
Back to top
Add Your Message Here
Username: Posting Information:
This is a private posting area. Only registered users and moderators may post messages here.
Options: Enable HTML code in message
Automatically activate URLs in message