Student Skill Models in Adaptive Testing

Martin Plajner, Jiří Vomlel
; Proceedings of the Eighth International Conference on Probabilistic Graphical Models, PMLR 52:403-414, 2016.

Abstract

This paper provides a common framework, a generic model, for Computerized Adaptive Testing (CAT) for different model types. We present question selection methods for CAT for this generic model. We use three different types of models, Item Response Theory, Bayesian Networks, and Neural Networks, that instantiate the generic model. We illustrate the usefulness of a special model condition – the monotonicity – and discuss its inclusion in these model types. With Bayesian networks we use specific type of learning using generalized linear models to ensure the monotonicity. We conducted simulated CAT tests on empirical data. Behavior of individual models was assessed based on these tests. The best performing model was the BN model constructed by a domain expert its parameters were learned from data under the monotonicity condition.

Cite this Paper


BibTeX
@InProceedings{pmlr-v52-plajner16, title = {Student Skill Models in Adaptive Testing}, author = {Martin Plajner and Jiří Vomlel}, booktitle = {Proceedings of the Eighth International Conference on Probabilistic Graphical Models}, pages = {403--414}, year = {2016}, editor = {Alessandro Antonucci and Giorgio Corani and Cassio Polpo Campos}}, volume = {52}, series = {Proceedings of Machine Learning Research}, address = {Lugano, Switzerland}, month = {06--09 Sep}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v52/plajner16.pdf}, url = {http://proceedings.mlr.press/v52/plajner16.html}, abstract = {This paper provides a common framework, a generic model, for Computerized Adaptive Testing (CAT) for different model types. We present question selection methods for CAT for this generic model. We use three different types of models, Item Response Theory, Bayesian Networks, and Neural Networks, that instantiate the generic model. We illustrate the usefulness of a special model condition – the monotonicity – and discuss its inclusion in these model types. With Bayesian networks we use specific type of learning using generalized linear models to ensure the monotonicity. We conducted simulated CAT tests on empirical data. Behavior of individual models was assessed based on these tests. The best performing model was the BN model constructed by a domain expert its parameters were learned from data under the monotonicity condition.} }
Endnote
%0 Conference Paper %T Student Skill Models in Adaptive Testing %A Martin Plajner %A Jiří Vomlel %B Proceedings of the Eighth International Conference on Probabilistic Graphical Models %C Proceedings of Machine Learning Research %D 2016 %E Alessandro Antonucci %E Giorgio Corani %E Cassio Polpo Campos} %F pmlr-v52-plajner16 %I PMLR %J Proceedings of Machine Learning Research %P 403--414 %U http://proceedings.mlr.press %V 52 %W PMLR %X This paper provides a common framework, a generic model, for Computerized Adaptive Testing (CAT) for different model types. We present question selection methods for CAT for this generic model. We use three different types of models, Item Response Theory, Bayesian Networks, and Neural Networks, that instantiate the generic model. We illustrate the usefulness of a special model condition – the monotonicity – and discuss its inclusion in these model types. With Bayesian networks we use specific type of learning using generalized linear models to ensure the monotonicity. We conducted simulated CAT tests on empirical data. Behavior of individual models was assessed based on these tests. The best performing model was the BN model constructed by a domain expert its parameters were learned from data under the monotonicity condition.
RIS
TY - CPAPER TI - Student Skill Models in Adaptive Testing AU - Martin Plajner AU - Jiří Vomlel BT - Proceedings of the Eighth International Conference on Probabilistic Graphical Models PY - 2016/08/15 DA - 2016/08/15 ED - Alessandro Antonucci ED - Giorgio Corani ED - Cassio Polpo Campos} ID - pmlr-v52-plajner16 PB - PMLR SP - 403 DP - PMLR EP - 414 L1 - http://proceedings.mlr.press/v52/plajner16.pdf UR - http://proceedings.mlr.press/v52/plajner16.html AB - This paper provides a common framework, a generic model, for Computerized Adaptive Testing (CAT) for different model types. We present question selection methods for CAT for this generic model. We use three different types of models, Item Response Theory, Bayesian Networks, and Neural Networks, that instantiate the generic model. We illustrate the usefulness of a special model condition – the monotonicity – and discuss its inclusion in these model types. With Bayesian networks we use specific type of learning using generalized linear models to ensure the monotonicity. We conducted simulated CAT tests on empirical data. Behavior of individual models was assessed based on these tests. The best performing model was the BN model constructed by a domain expert its parameters were learned from data under the monotonicity condition. ER -
APA
Plajner, M. & Vomlel, J.. (2016). Student Skill Models in Adaptive Testing. Proceedings of the Eighth International Conference on Probabilistic Graphical Models, in PMLR 52:403-414

Related Material