[Rasch] Re: Rasch Digest, Vol 36, Issue 7
lavallee.jp at gmail.com
Mon Jul 7 15:14:22 EST 2008
Thanks, Mike, for the link to the ETS DIF criteria. I hope I can ask a
further question on the use of the DIF 'size' and 'significance' criteria.
I'm currently equating exams originally created by test editors drawing from
a pool of uncalibrated items. Each test was taken by more than 2000
students. I'm using WINSTEPS MForms to merge two forms at a time into a
single matrix, then using the DIF table to check the common items for DIF.
My question is this: are the ETS size/significance criteria 'either-or'? I'm
finding many items exhibiting DIF where the DIF contrast is small in
absolute size (0.25 logits or less) but nonetheless statistically
significant at the 0.01 level.
I'm not sure whether I should...
-simply use a cut-off point based on size, such as ETS's 0.43 'moderate DIF'
-investigate the causes of the high DIF significance/low DIF size items
-randomly select a smaller sample of, say 1000 students per test, and see if
the items still show DIF.
I'd be grateful for any and all input on this problem!
Ming Chuan University, Taiwan
> From: "Mike Linacre (RMT)" rmt at rasch.org
> Invariance of item difficulties across different person groups can be
> investigated by means of DIF analysis. ETS have developed rules-of-thumb
> for identifying when their items exhibit enough DIF for remedial action to
> be taken. You can see the ETS criteria at
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Rasch