Files in this item



application/pdf3160942.pdf (7MB)Restricted to U of Illinois
(no description provided)PDF


Title:An Empirical Investigation of Several Critical Psychometrical Issues in Neuropsychological Testing of Mild Traumatic Brain Injuries
Author(s):Ragan, Brian George
Doctoral Committee Chair(s):Zhu, Weimo
Department / Program:Kinesiology and Community Health
Discipline:Kinesiology and Community Health
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Psychology, Psychometrics
Abstract:Concussions can cause very serious injuries and with repetitive injuries increase the likelihood of serious consequences. There is an increased awareness of concussions, which has evolved into developing better measurement tools to study the effects including neuropsychological components. Cognitive functions such as, memory, attention, concentration, and reaction time are now commonly assessed for dysfunction in sports related concussions. Although there have been many neuropsychological tests used to identify the effects of concussion, there are several major concerns of the methodological soundness and psychometric quality of commonly used practices. The purpose of this study was to examine the commonly used neuropsychological sports concussion practices used for validity, reliability, and other psychometric characteristics. A windows-based software program was created called Short-term Memory Assessment Recall Tool (SMART) for this study. Item Response Theory (IRT) was employed to address the issues with the common practices, specifically the one-parameter or Rasch model was used. The issues include design, construction, testing paradigm, administration, and scoring or identifying injuries. The results from the IRT calibrations show compliance of the test with the stated test construction and testing paradigm commonly used. There was, however, some administration problems identified relating to the random generated item selection of computerized concussion tests. When the randomness of the items and the difficulty level of the items were controlled for the reliability of SMART was good at R = 0.8. Another finding of the study was using the conditional standard errors from the IRT calibration reduced the range of the 90% confidence interval by over 50%. These issues must be addressed before larger issues of examining whether an individual-centered or criterion-referenced standard should be used. Further investigation into the development and construction of these tests are needed to address the alarming psychometric concerns. It is important for these tests to be constructed and evaluated with sound modern measurement theory. Currently, the commonly used practices are potentially riddled with significant issues and problems.
Issue Date:2004
Description:197 p.
Thesis (Ph.D.)--University of Illinois at Urbana-Champaign, 2004.
Other Identifier(s):(MiAaPQ)AAI3160942
Date Available in IDEALS:2015-09-28
Date Deposited:2004

This item appears in the following Collection(s)

Item Statistics