[Rasch] Advice needed on analysis

Snider-Lotz, Tom TSnider-Lotz at qwiz.com
Fri Dec 2 07:29:39 EST 2005

I'd appreciate some advice on an analysis I'm planning.  I have a large number of cases (250,000+) of item-level results from a 30-item, dichotomously scored test of skill with a word processing software.  The test simulates the use of the software, and each item involves the completion of a certain task with the software.  The items are arranged so that items 1-10 were SME-judged to represent Basic tasks, items 11-20 are Intermediate, and 21-30 are Advanced.
The test comes in another version, called the "Essentials" version, which is identical to the regular test except it contains only the first 20 (Basic and Intermediate) items.  I have another 50,000+ cases of results from the Essentials version.
All tests were adminstered by employers who use our tests for assessment of job candidates.
There are two things I would like to know:  (1) whether the candidates who are given the full version of the test are different in skill from candidates who are given the Essentials version; and (2) whether scores on the tests have risen over time.
Regarding the first question, the two approaches I have considered are (a) merging the data from the data sets and treating "version" as a person variable that I can use to compare the ability levels of the two groups of candidates, and (b) analyzing the data from the full version, and then equating on the 20 common items when I analyze the data from the Essentials version.
I can envision similar strategies to address the second question.
I'd appreciate comments and suggestions on this analysis.  
 -- Tom Snider-Lotz

 THOMAS G. SNIDER-LOTZ, PHDMANAGERCONTENT RESEARCH & DEVELOPMENTQWIZ, INC.1805 OLD ALABAMA ROADSUITE 150ROSWELL, GEORGIA 30076PH 678.832.0555PH 800.281.9713 X555FX 770.642.6115WWW.QWIZ.COMTSNIDER-LOTZ at QWIZ.COM <https://mail.qwiz.com/signatures/tsnider-lotz.gif> 	 <http://www.qwiz.com/signature> 	
This communication, along with any attachment(s), is intended only for the use of the addressee(s) and may contain proprietary, confidential or privileged information. If you are not the intended recipient, you are hereby notified that any dissemination, distribution or copying of any information contained in or attached to this communication is strictly prohibited. If you have received this message in error, please notify the sender immediately and destroy the original communication and its attachments without reading, printing or saving in any manner. Thank you. 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://mailinglist.acer.edu.au/pipermail/rasch/attachments/20051201/c8fd0131/attachment.html 

More information about the Rasch mailing list