# [Rasch] Question

Anthony James luckyantonio2003 at yahoo.com
Thu Jul 20 18:40:52 EST 2006

```Dear list,
I have a statistical question. I hope you don’t mind if I post it here.
I have two ability measures for a sample of 542 students based on two parallel tests. Each measure has its own associated standard error. In order to study whether the differences between the measures on the two tests for each student is significantly different or not I
1-Computed the difference between the measures (contrast)
2-Computed the joint standard error, i.e., the square root of (first SE squared +second SE squared)
3-Computed t. t= contrast/joint SE
The df should be 542*2-2=1082
The critical value of t for df of infinity is 1.64 (alph=0.05).
If the t for a student exceeds this value then I can claim that the two measures are significantly different for that student. Right?
How can I compute the probability for the t?

Is the procedure sound?
Am I on the right path?
Thanks a lot
Anthony

---------------------------------
Do you Yahoo!?
Next-gen email? Have it all with the  all-new Yahoo! Mail Beta.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20060720/fdc0225e/attachment.html
```