[Rasch] Q3 index and local dependence

Mike Linacre mike at winsteps.com
Tue Mar 27 13:14:46 AEDT 2018


Q3 is the correlation between the raw residuals of pairs of items. These 
are listed, for instance, in Winsteps Table 23.99 with PRCOMP=R

Summarizing the Q3's across all items with PCA makes the dependency 
pattern clearer. See, for instance, Winsteps Table 23.1

Hope this help.

Mike L.

On 3/27/2018 10:38 AM, Bond, Trevor wrote:
> Dear Colleagues,
> It has been claimed that Yen’s Q3 index (1984) can be used to identify 
> local dependence in Rasch analyses.
> e.g., Chou, Y.-T., & Wang, W.-C. (2010). Checking dimensionality in 
> item response models with principal component analysis on standardized 
> residuals. Educational and Psychological Measurement, 70(5), 717-731.
> Christensen, K. B., Makransky, G., & Horton, M. (2017). Critical 
> values for Yen’s Q3: Identification of local dependence in the Rasch 
> model using residual correlations. Applied Psychological Measurement, 
> 41(3), 178-194.
> Have you used Q3 for that purpose? Do you know of papers that have 
> adopted it? Is it helpful/effective? Is it available in the software 
> that you use, or know about?
> Looking forward to your advice and guidance.
> Collegially
> Trevor

Mike Linacre, mike at winsteps.com or winsteps1234 at gmail.com
Winsteps 4.0.1 and Facets 3.80.4 - www.winsteps.com 
Concerned about aches, pains, youthfulness? Mike suggests Liquid Biocell 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20180327/61a2b631/attachment.html>

More information about the Rasch mailing list