A Live Tailored Testing Comparison Study of the One- and Three-Parameter Logistic Models.

Abstract

Although both the one-parameter and the three-parameter logistic models have frequently been applied to computerized tailored tests, no direct empirical comparisons of the two models have been reported in the literature. The major research problem, therefore, was to compare the models based on a live tailored testing study conducted with two identical item pools derived from a vocabulary test. A total of 128 undergraduate and graduate students enrolled in educational psychology and measurement courses at the University of Missouri - Columbia served as examinees for the study. A counterbalanced test-retest design was employed in which there were two separate test sessions one week apart for each examinee, with both the one- and three-parameter tests administered at each session. The tailored tests were administered on Applied Digital Data Systems Consul 980 cathode ray tube terminals which were connected to an IBM 370/168 computer through a timesharing system. Relative efficiency curves, test-retest reliability coefficients, goodness of fit of the models, convergence rates, descriptive statistics, and the correlation of ability estimates with an outside criterion were used to compare the two models.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jun 01, 1978
Accession Number
ADA058528

Entities

People

  • Mark D. Reckase
  • William R. Koch

Tags

Communities of Interest

  • Biomedical
  • Human Systems

DTIC Thesaurus Topics

  • Birds
  • Cognition
  • Computational Science
  • Computer Programs
  • Correlation Analysis
  • Data Analysis
  • Data Science
  • Descriptive Analytics
  • Educational Psychology
  • Information Science
  • Military Research
  • Probability
  • Psychology
  • Statistics
  • Students
  • Surveys
  • Test Methods

Readers

  • Computer Science/Computer Engineering/Data Science/Digital Signal Processing.
  • Exercise and Sports Science.
  • Statistical inference.