Item Scaling Models
The data from NAEP are scaled using Item Response Theory (IRT; e.g., Lord 1980) models. For multiple-choice items (dichotomously scored items), the three-parameter logistic form of the model is used. For constructed-response items that are scored right-wrong (dichotomously scored items), the two-parameter logistic form of the model is used. For other constructed-response items (polytomously scored items), the generalized partial credit model form is used. These models express respondents' tendencies to achieve certain scores (such as correct/incorrect) on the items contributing to a scale as a function of a parameter that is not directly observed. These different types of items and models collectively contribute to NAEP scales. Item parameter estimates on a provisional scale are obtained using the NAEP BILOG (Mislevy and Bock 1982)/PARSCALE (Muraki and Bock 1997) program using marginal maximum likelihood methods.
Last updated 14 July 2008 (DB)
Printer-friendly Version
|