[IPAC-List] MH DIF statistic

[Shekerjian, Rene]

Can anyone tell me why -2.35 is used to adjust the natural log of the MH common odds ratio estimate? I understand how the natural log makes the MH statistic more easily interpreted, but I haven't found an explanation for multiplying it by -2.35. Does it "enlarge" the result so that what would be small differences are magnified and become bigger differences (more different from one another)?

Thanks in advance!


René

René Shekerjian | Testing Services Division | NYS Department of Civil Service | 518-402-2660
======================================================================

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://nine.pairlist.net/pipermail/ipac-list/attachments/20140221/88ddda13/attachment.htm