We are a leading provider of software for adaptive test development and psychometric analysis. Our psychometric software is used to drive big-picture K–12 assessment programs, develop and to deliver psychometric based assessments.

The FASTCAT IRT solution is a web-based computer-adaptive system that education providers use to quickly and accurately place students as to their appropriate instructional level at grades k thru12. Used to pinpoint the students initial level of proficiency, it then further reports growth, and can be used to predict performance in forthcoming tests. The FASTCAT IRT system includes a set of powerful IRT test development and analysis tools.

The FASTCAT IRT System includes:

• User Interface

  • Student Test Centre

• Content Management

  • Create IRT-Items, build and publish IRT exams

• IRT-Item Bank

  • IRT-Item Banking System, Bulk Uploading of Items

• IRT Report Generator

  • Build IRT graphs including item characteristic curves

  • Create exam-associated IRT information graphs

• IRT 3PL engine

  • Uses Item Response Theory algorithms to select items

• IRT 3PL calibration

  • Weighting of questions (difficulty, discrimination, guessing)

  • Identification/weeding of poor items (negative discrimination)

  • Differential item functioning

• IRT Toolkit

  • Student and item parameter estimation

  • Introduction and calibration of new questions to item pool

• IRT Database

  • IRT-Items, Student Responses, Item Parameters, Test Instances

HOW IT WORKS

The FASTCAT IRT test is a computer adaptive test, where each student receives a unique test. The test engine works by targeting item selections to the abilities of students, generating more difficult questions if the student is answering correctly and vice versa, presenting easier items if the student is answering incorrectly. Thus alleviating continued rounds of questions at levels too difficult or too easy. The engine uses the calibrated difficulty, discrimination and guess factor of each question to determine the next item. Tests can be built to ensure that all content areas strands are covered.