Login

Publications  •  Project Statistics

Glossary  •  Schools  •  Disciplines
People Search: 
   
Title/Abstract Search: 

Dissertation Information for Dan Bolt

NAME:
- Dan Bolt
- (Alias) Daniel Marc Bolt

DEGREE:
- Ph.D.

DISCIPLINE:
- [No Discipline Recorded]

SCHOOL:
- University of Illinois, Urbana-Champaign (USA) (1999)

ADVISORS:
- None

COMMITTEE MEMBERS:
- None

MPACT Status: Incomplete - Not_Inspected

Title: Psychometric methods for diagnostic assessment and dimensionality representation

Abstract: "Educational and cognitive psychologists increasingly suggest that the ""low-to-high"" ability models commonly used in test theory are neither consistent with contemporary beliefs about knowledge structure and its relation to test performance nor the diagnostic goals towards which tests are frequently directed. A constrained mixture Rasch model is shown to provide a useful framework in which both considerations can be developed. Through the systematic imposition of ordinal and equality constraints on item parameters, the mixture Rasch model is shown to be a useful extension of several restricted latent class models in which ability is simultaneously characterized by both discrete and continuous components, and the definitions of classes can be made consistent with prior expectations concerning examinee knowledge structure or the diagnostic goals of the test.

Several simulation studies are conducted to demonstrate the potential of the constrained model under varying test conditions. In particular, the ability of the model to approximate test structures corresponding to more cognitively complex models used for diagnosis is evaluated. Data from the quantitative sections of a Graduate Record Examination (GRE-Q) as well as a college algebra test intentionally designed to be diagnostic are also analyzed. The advantages of using a psychometric model in the context of diagnostic assessment are discussed.

A second part of the thesis investigates a new methodology for the representation of multidimensional test structure based on item-pair conditional covariances. The multidimensional ability composite best measured (in terms of information) for an item is sometimes regarded as the measurement direction of the item. We demonstrate how a directional representation of an item can be reconstructed on the basis of the item's pattern of conditional covariances with all of the other items on the test. A proximity measure for item pairs is derived, and circular and spherical scaling techniques are shown to provide a method by which the unique measurement directions of items might be recovered and represented. Several real and simulated data analyses suggest that the approach may be a useful tool for investigating and describing the dimensional structure of tests."

MPACT Scores for Dan Bolt

A = 0
C = 1
A+C = 1
T = 0
G = 0
W = 0
TD = 0
TA = 0
calculated 2008-01-31 06:32:54

Advisors and Advisees Graph