Research Framework - Statistical Analysis

Calibration and Ability Estimation Methods

This area focuses on the statistical methods and accompanying software used to estimate item parameters, ability distribution parameters, and individual examinee parameters. For complex models and accompanying statistical estimation methods, research includes studying the effectiveness of such methods in terms of statistical convergence, numerical efficiency, estimation bias, etc.

---

Research

Recovery of Test Characteristic Curves for Short Dimensionally Distinct Subtests: Comparison of Unidimensional and Multidimensional Calibrations, by Louis Roussos and Abdullah Ferdous (2007), Paper presented at the American Education Research Association Annual Meeting, Chicago, IL

Missing Data Treatment Methods in Parameter Recovery for a Mixed-Format Test, by Thakur Karkee and Matt Finkelman (2007), Paper presented at the American Education Research Association Annual Meeting, Chicago, IL 

Robustness of Unidimensional Hierarchical Modeling of Discrete Attribute Association in Cognitive Diagnosis Models, by Jonathan L. Templin, Robert A. Henson, Sara E. Templin, and Louis Roussos (2004), Unpublished ETS Project Report, Princeton, NJ

Estimation of Attribute Association, by Robert Henson, Louis Roussos, William Stout
Jeff Douglas, Xuming He, and Jon Templin (2003), Unpublished ETS Project Report, Princeton, NJ