Advanced Search

Journal Navigation

Journal Home

Subscriptions

Archive

Contact Us

Table of Contents

Click here to sign up for SAGE Journal Email Alerts today!

Sign In to gain access to subscriptions and/or personal tools.
Journal of Educational and Behavioral Statistics
This Article
Right arrow Full Text (PDF)
Right arrow References
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Add to Saved Citations
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Request Reprints
Right arrow Add to My Marked Citations
Citing Articles
Right arrow Citing Articles via Google Scholar
Right arrow Citing Articles via Scopus
Google Scholar
Right arrow Articles by Wu, I.-L.
Right arrow Search for Related Content
Social Bookmarking
 Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati   Add to Twitter  
What's this?

Articles

A New Computer Algorithm for Simultaneous Test Construction of Two-Stage and Multistage Testing

Ing-Long Wu

National Chung Cheng University

This paper discusses the simultaneous approach for test generation of two-stage and multistage tests from an item bank. However, most previous test generation problems involved binary programming models, and the efficiency of available solution algorithms is the major concern for these problems. Therefore, this study considers two important concepts on the solution process, i.e., alternative ways to formulate mathematical models and alternative solution algorithms. Based on these two concepts, this paper is two-fold. First, two binary programming models with a special network structure that would be explored computationally are presented for modeling these problems. The first model maximizes test information function at one specified ability point and the second model matches the target test information function at several specified ability points as closely as possible. Second, an efficient special-purpose network algorithm is then used to solve these two models. Hence, the test construction process tends to be improved in terms of both computational efforts and quality of tests. An empirical study shows the results in line with the above two criteria.

Key Words: two-stage and multistage testing • binary programming • network algorithm • item response theory • item bank • test construction

Journal of Educational and Behavioral Statistics, Vol. 26, No. 2, 180-198 (2001)
DOI: 10.3102/10769986026002180


Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati   Add to Twitter Twitter    What's this?




AER home page RER home page JEB home page EPA home page RRE home page