Further standardization of some NES3 tests

Richard Letz, Colleen K. DiIorio, Patricia O. Shafer, Katherine A. Yeager, Donald L. Schomer, Thomas R. Henry

Research output: Contribution to journalArticlepeer-review

22 Scopus citations

Abstract

NES3 is a computer-based neurobehavioral testing system designed for use in investigating potential cognitive-impairment. Data from NES3 tests employed in a study of epilepsy self-management were analyzed to estimate test-retest reliability for NES3 tests on a large sample and to estimate the effect of some common covariates of test performance. A total of 319 participants in an epilepsy self-management study were examined on three occasions (baseline, 3-month follow-up, and 6-month follow-up) with a set of psychological questionnaires and seven neuropsychological tests from NES3. Test-retest correlations were calculated between measures obtained at baseline and at 3 months. Principal components analysis was performed on the baseline data. The potential effects of covariates (age, education, reading test scores, depression status, and examiner) were investigated in regression models. Three-month test-retest correlations were excellent for Adult Reading Test (ART) (r = 0.95), strong for Digit-Symbol (r = 0.82), Sequence B (r = 0.79), and Sequence A (r = 0.76); and modest (r's between 0.56 and 0.67) for Digit Span Forward and Backward, Visual Span Forward and Backward, and Pattern Memory. Alternate-forms correlations were strong for HVLT (r's between 0.71 and 0.82). Principal components analysis yielded four interpretable components. Age and reading score were significant covariates of virtually all of the test summary measures, while education, gender, race, and depression were not generally significant covariates. Changes to the method of calculation of some summary measures, changes to the initial instructions to the subjects, and addition of correctional feedback to subjects during the tests appeared to improve the reliability of some NES3 tests. Implementation of the HVLT and ART in computer-assisted format added breadth of coverage to the battery. NES3 tests may provide reliable, efficient data for use in epidemiologic studies of potential cognitive effects of occupational and environmental exposures.

Original languageEnglish (US)
Pages (from-to)491-501
Number of pages11
JournalNeuroToxicology
Volume24
Issue number4-5
DOIs
StatePublished - Aug 2003
Externally publishedYes

Bibliographical note

Funding Information:
This project was supported by the US National Institute for Nursing Research (R01 NR04770).

Keywords

  • Computer-based testing
  • Neuropsychological tests
  • Psychometrics

Fingerprint

Dive into the research topics of 'Further standardization of some NES3 tests'. Together they form a unique fingerprint.

Cite this