%0 Journal Article %J Journal of Educational and Behavioral Statistics %D 2009 %T A Knowledge-Based Approach for Item Exposure Control in Computerized Adaptive Testing %A Doong, S. H. %X

The purpose of this study is to investigate a functional relation between item exposure parameters (IEPs) and item parameters (IPs) over parallel pools. This functional relation is approximated by a well-known tool in machine learning. Let P and Q be parallel item pools and suppose IEPs for P have been obtained via a Sympson and Hetter–type simulation. Based on these simulated parameters, a functional relation k = fP (a, b, c) relating IPs to IEPs of P is obtained by an artificial neural network and used to estimate IEPs of Q without tedious simulation. Extensive experiments using real and synthetic pools showed that this approach worked pretty well for many variants of the Sympson and Hetter procedure. It worked excellently for the conditional Stocking and Lewis multinomial selection procedure and the Chen and Lei item exposure and test overlap control procedure. This study provides the first step in an alternative means to estimate IEPs without iterative simulation.

%B Journal of Educational and Behavioral Statistics %V 34 %P 530-558 %U http://jeb.sagepub.com/cgi/content/abstract/34/4/530 %R 10.3102/1076998609336667 %0 Book Section %D 2009 %T Kullback-Leibler information in multidimensional adaptive testing: theory and application %A Wang, C. %A Chang, Hua-Hua %X Built on multidimensional item response theory (MIRT), multidimensional adaptive testing (MAT) can, in principle, provide a promising choice to ensuring efficient estimation of each ability dimension in a multidimensional vector. Currently, two item selection procedures have been developed for MAT, one based on Fisher information embedded within a Bayesian framework, and the other powered by Kullback-Leibler (KL) information. It is well-known that in unidimensional IRT that the second derivative of KL information (also termed “global information”) is Fisher information evaluated atθ 0. This paper first generalizes the relationship between these two types of information in two ways—the analytical result is given as well as the graphical representation, to enhance interpretation and understanding. Second, a KL information index is constructed for MAT, which represents the integration of KL nformation over all of the ability dimensions. This paper further discusses how this index correlates with the item discrimination parameters. The analytical results would lay foundation for future development of item selection methods in MAT which can help equalize the item exposure rate. Finally, a simulation study is conducted to verify the above results. The connection between the item parameters, item KL information, and item exposure rate is demonstrated for empirical MAT delivered by an item bank calibrated under two-dimensional IRT. %C D. J. Weiss (Ed.), Proceedings of the 2009 GMAC Conference on Computerized Adaptive Testing. %G eng %0 Report %D 2006 %T Kernel-smoothed DIF detection procedure for computerized adaptive tests (Computerized testing report 00-08) %A Nandakumar, R. %A Banks, J. C. %A Roussos, L. A. %I Law School Admission Council %C Newton, PA %G eng %0 Journal Article %J Zeitschrift für Differentielle und Diagnostische Psychologie %D 2004 %T Kann die Konfundierung von Konzentrationsleistung und Aktivierung durch adaptives Testen mit dern FAKT vermieden werden? [Avoiding the confounding of concentration performance and activation by adaptive testing with the FACT] %A Frey, A. %A Moosbrugger, H. %K Adaptive Testing %K Computer Assisted Testing %K Concentration %K Performance %K Testing computerized adaptive testing %X The study investigates the effect of computerized adaptive testing strategies on the confounding of concentration performance with activation. A sample of 54 participants was administered 1 out of 3 versions (2 adaptive, 1 non-adaptive) of the computerized Frankfurt Adaptive Concentration Test FACT (Moosbrugger & Heyden, 1997) at three subsequent points in time. During the test administration changes in activation (electrodermal activity) were recorded. The results pinpoint a confounding of concentration performance with activation for the non-adaptive test version, but not for the adaptive test versions (p = .01). Thus, adaptive FACT testing strategies can remove the confounding of concentration performance with activation, thereby increasing the discriminant validity. In conclusion, an attention-focusing-hypothesis is formulated to explain the observed effect. (PsycINFO Database Record (c) 2005 APA ) (journal abstract) %B Zeitschrift für Differentielle und Diagnostische Psychologie %V 25 %P 1-17 %G eng