[Rasch] Quadrature points and posterior weights

Mike Linacre (RMT) rmt at rasch.org
Tue May 12 12:25:33 EST 2009


Thank you for your question, Hu Nan.

Sorry, Winsteps does not estimate latent distributions.

 From Winsteps, you could output the person estimates from each group into 
an Excel file, and then use Excel to estimate the shape of their 
distribution: Tools/Data Analysis/Descriptive Statistics  (after installing 
the Excel "Data Analysis" Add-in).

The BILOG-MG latent distributions may be corrected for measurement error. 
This can be approximated by reducing the Excel-reported "sample variance" 
by the average error-variance of the person estimates in the group.

If your sample distribution is bimodal, then you can use Excel Solver to 
estimate the parameters of a suitable distribution.

OK?

Mike L.

At 5/11/2009, you wrote:
>Hello,
>
>Could anyone tell me if Winsteps (v 3.68.1) provide "Quadrature points and 
>posterior weights" of the estimated latent distributions?  If yes, from 
>which output table can I obtain them?  Thanks.
>
>(For example, BILOG-MG provides "Quadrature points and posterior weights" 
>of the estimated ltent distributions of the examinee groups in the output 
>file)
>
>Hu Nan

Mike Linacre
Editor, Rasch Measurement Transactions
rmt at rasch.org www.rasch.org/rmt/ Latest RMT:  22:4 Spring 2009

------------------------------------------------
Please consider the environment before you print

-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20090511/8b9380ca/attachment.html 


More information about the Rasch mailing list