NotesFAQContact Us
Search Tips
ERIC Number: ED318790
Record Type: Non-Journal
Publication Date: 1990-Apr
Pages: 18
Abstractor: N/A
Reference Count: N/A
Validity of Using Two Numerical Analysis Techniques To Estimate Item and Ability Parameters via MMLE: Gauss-Hermite Quadrature Formula and Mislevy's Histogram Solution.
Seong, Tae-Je
The similarity of item and ability parameter estimations was investigated using two numerical analysis techniques via marginal maximum likelihood estimation (MMLE) with a large simulated data set (n=1,000 examinees) and changing the number of quadrature points. MMLE estimation uses a numerical analysis technique to integrate examinees' abilities over the ability distribution for item and ability parameter estimations because of the difficulty of direct integration with a digital computer. For integrating ability, the values of quadrature points and the weights corresponding to each quadrature point are specified. The Gauss-Hermite quadrature formula and R. J. Mislevy's histogram solution (1984) have been used for numerical integration over the normal density function. It was determined that the Gauss-Hermite quadrature formula and Mislevy's graphical solution via MMLE estimated item and ability parameters equally when a large number of quadrature points was specified. When a small number of quadrature points was specified, Mislevy's histogram solution via the MMLE approach estimated item and ability parameters more accurately than did the Gauss-Hermite quadrature formula. Seven tables summarize the study. (Author/SLD)
Publication Type: Reports - Research; Speeches/Meeting Papers
Education Level: N/A
Audience: N/A
Language: English
Sponsor: N/A
Authoring Institution: N/A