BILOG-MG Operational

BILOG-MG is an extension of BILOG that is designed for the efficient analysis of binary items, including multiple-choice or short-answer items scored right, wrong, omitted, or not-presented. BILOG-MG is capable of large-scale production applications with unlimited numbers of items or respondents. It can perform item analysis and scoring of any number of subtests or subscales in a single program run. All the program output may be directed to text files for purposes of selecting items or preparing reports of test scores.

BILOG-MG implements an extension of Item Response Theory (IRT) to multiple groups of respondents. It has many applications in test development and maintenance. Applications of multiple-group item response theory in education assessment and other large-scale testing programs include the following features.

• Nonequivalent groups equating for maintaining the comparability of scale scores as new forms of the test are developed.
• Vertical equating of test forms across school grades or age groups.
• Analysis of Differential Item Functioning (DIF) associated with demographic or other group differences.
• Detecting and correcting for item parameter trends over time (DRIFT).
• Calibrating and scoring tests in two-stage testing procedures designed to reduce total testing time.
• Estimating latent ability or proficiency distributions of students in schools, communities, or other aggregations.

In addition, BILOG-MG provides for variant items that are inserted in tests for purpose of estimating item statistics, but that are not included in the scores of the examinees.

BILOG-MG for operational purposes includes, but is not limited to, test/assessment calibration, scoring, linking or standard setting for a client, product, or student/employee evaluation.