[Rasch] CAT ability estimation
garybuck at umich.edu
Wed Jun 4 04:45:40 EST 2008
Thank you for your helpful reply.
I would iike to ask what you are suggesting about dealing with the ³vagaries
of the internet and ability estimation process and subsequent administration
Myself and some colleague have just developed a Java-based internet
test-delivery engine (not browser based) called TGOD, which currently
delivers linear tests over the internet. We find that items are delivered
fairly quickly over a broad band connection, with little or no noticeable
delay to the test taker. The addition of a CAT function may change that of
course, so any suggestions would be much appreciated.
To the Whole List
Although TGOD (Test Generation and Online Delivery System) will also be
available for commercial use, for a fee of course, we are making it
available free to scholars, researchers and educators for unlimited,
non-commercial use. So if any of you are interested in delivering linear
tests online, or maybe your students are, send an email to
pacia at lidgetgreen.org, and she will set you up with an account. This offer
applies to the whole measurement community. There is no obligation at all in
using TGOD, and we hope to continue this policy into the long-term future.
And while I am on the topic, once the CAT version of TGOD is developed,
hopefully later this year if I can figure out the design, the same offer
will stand. It will be free to scholars, researchers and educators for
unlimited, non-commercial use. So if any of you think you will be wanting to
use such a piece of software, you are cordially invited to send me your wish
list, and if any of you know how it should be designed, your ideas, demands,
recommendations and general help would be very welcome.
Lidget Green, Inc
gary at lidgetgreen.org
PS: I am sending this from my University of Michigan account, since that is
the account that is signed up to the list, but Lidget Green, and TGOD, have
no connection with the University of Michigan.
On 6/2/08 9:49 PM, "Nathaniel Lewis" <N.Lewis at eaa.unsw.edu.au> wrote:
> Hi Gary,
> The references below provide may provide good starting points for your
> evaluation of the merits of available approaches. As Mike alluded to a lot of
> the speed/timing problems are now a non-issue with respect to computational
> May I suggest that you pay some attention on how you deal with the vagaries of
> the internets (for a web-based CAT) in relation to the ability estimation
> process and subsequent administration of items.....
> Properties of Ability Estimation Methods in Computerized Adaptive Testing
> Tianyou Wang and Walter P. Vispoel
> Journal of Educational Measurement, Vol. 35, No. 2 (Summer, 1998), pp. 109-135
> (article consists of 27 pages)
> A Review of Item Exposure Control Strategies for Computerized Adaptive Testing
> Developed from 1983 to 2005
> The Journal of Technology, Learning, and Assessment
> Volume 5, Number 8 · May 2007
> Enjoy, it is great stuff!
>> -----Original Message-----
>> From: rasch-bounces at acer.edu.au [mailto:rasch-bounces at acer.edu.au] On Behalf
>> Of Gary Buck
>> Sent: Tuesday, 3 June 2008 6:46 AM
>> To: Rasch at acer.edu.au
>> Subject: [Rasch] CAT ability estimation
>> I wonder whether anyone can help?
>> I am in the process of designing a web-based CAT. I feel I have a fairly
>> good idea of most of the issues involved, except how to estimate the
>> persons¹ Rasch measure based in their responses to a string of calibrated
>> I know there are different estimation procedures, and some are
>> computationally less demanding and presumably less accurate, and visa versa.
>> A discussion of the relative merits of each would be much appreciated, since
>> the server has to calculate all the ability estimates.
>> Kind regards,
>> U of Michigan
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Rasch