BILOG 3.04 implements modern item response theoretic methods of item analysis and test scoring in a fast, user-oriented computer program. It is designed for a wide range of IRT applications to practical testing problems with long or short tests, multiple subtests, and multiple test forms. Item parameter estimation is by constrained marginal maximum likelihood, with provision for concurrent estimation of the latent distribution. BILOG-3.04 is available for IBM PCs and compatibles with 640K RAM. Features include:
* choice of 1-, 2-, or 3-parameter logistic item-response models;
* marginal maximum likelihood or marginal maximum a posteriori estimation of item intercepts, slopes, and lower asymptotes;
* standardized posterior residuals for individual items;
* tests of fit for individual items;
* standard errors for all item parameter estimates and scale scores of respondents;
* analysis of multiple subtests in one pass;
* handling of not-presented items and linking of test forms through common items;
* several options for handling omitted responses;
* provision for constrained parameter estimation to insure identifiability;
* analysis of case-weighted probability samples;
* analysis of multiple-form tests;
* maximum likelihood and Bayes modal , or Bayes , estimation of scale scores;
* test and item information analysis;
* plots of item-response functions, item information curves, and test information curves;
* rescaling of test scores;
* estimation of latent distributions; and
* analysis of group-level educational assessment data.
BILOG 3.04 implements modern item response theoretic methods of item analysis and test scoring in a fast, user-oriented computer program. It is designed for a wide range of IRT applications to practical testing problems with long or short tests, multiple subtests, and multiple test forms. Item parameter estimation is by constrained marginal maximum likelihood, with provision for concurrent estimation of the latent distribution. BILOG-3.04 is available for IBM PCs and compatibles with 640K RAM. Features include:
* choice of 1-, 2-, or 3-parameter logistic item-response models;
* marginal maximum likelihood or marginal maximum a posteriori estimation of item intercepts, slopes, and lower asymptotes;
* standardized posterior residuals for individual items;
* tests of fit for individual items;
* standard errors for all item parameter estimates and scale scores of respondents;
* analysis of multiple subtests in one pass;
* handling of not-presented items and linking of test forms through common items;
* several options for handling omitted responses;
* provision for constrained parameter estimation to insure identifiability;
* analysis of case-weighted probability samples;
* analysis of multiple-form tests;
* maximum likelihood and Bayes modal , or Bayes , estimation of scale scores;
* test and item information analysis;
* plots of item-response functions, item information curves, and test information curves;
* rescaling of test scores;
* estimation of latent distributions; and
* analysis of group-level educational assessment data.