A hierarchical framework for modeling speed and accuracy on test items

Research output: Book/ReportReportOther research output

110 Downloads (Pure)

Abstract

Current modeling of response times on test items has been influenced by the experimental paradigm of reaction-time research in psychology. For instance, some of the models have a parameter structure that was chosen to represent a speed-accuracy tradeoff, while others equate speed directly with response time. Other response-time models seem to be unclear as to the level of parameterization they represent. A hierarchical framework of modeling is proposed to better represent the nature of speed and accuracy on test items as well as the different levels of dependency between them. The framework allows a “plug-and-play approach” with alternative choices of response and response-time models to deal with different types of test items as well as population and item-domain models to represent key relations between their parameters. Bayesian treatment of the framework with Markov chain Monte Carlo (MCMC) computation facilitates the approach. Use of the framework is illustrated for the choice of a normal-ogive response model, a lognormal model for the response times, and multivariate normal models for the population and item domain with Gibbs sampling from the joint posterior distribution.
Original languageEnglish
Place of PublicationNewton, PA, USA
PublisherLaw School Admission Council
Number of pages16
Publication statusPublished - 2005

Publication series

NameLSAC research report series
PublisherLaw School Admission Council
No.05-02

Fingerprint Dive into the research topics of 'A hierarchical framework for modeling speed and accuracy on test items'. Together they form a unique fingerprint.

Cite this