[Rasch] Investigating Dimensionality

Mike Linacre (RMT) rmt at rasch.org
Wed Jul 21 15:00:28 EST 2010


Thank you for your question, Jacob.

Yes, you can split the items into two subsets based on their loadings on 
contrast (component, factor), and compare the person abilities on the two 
sets of items.

In Winsteps:
1. Put a code into the item label to indicate the subset to which the item 
belongs.
2a. Use ISELECT= for each subset code, and produce a person measure 
(PFILE=). Cross-plot the person measures.
or
2b. Do a Differential Person Functioning (DPF=) analysis based on the 
subset code. Table 31.1 will give you an inter-item-subset t-test for each 
person.

Mike Linacre
mike at winsteps.com

At 7/20/2010, you wrote:
>I'm interested in investigating dimensionality in a data set and see many 
>references to E.V. Smith's (2002) paper in the Journal of Applied 
>Measurement on this topic. It appears that RUMM2030 has a standard routine 
>to do this analysis, but I'm unsure if the analysis can be done in 
>Winsteps. As an alternative, could one split items into subtests based on 
>the positive and negative loadings on a particular contrast and compare 
>the item difficulties for those two sets?
>
>Any input is appreciated.
>
>Thanks,
>
>Jacob Kean PhD
>Indiana University




More information about the Rasch mailing list