Abstract:
It is observed that examinations held at educational institutions globally are conducted on periodical basis. We, being inhabitants of the Computer Age, have deployed automated examination systems where schools of students can take the same exam on parallel test schedules with instant report generation evaluating their test performance. The major underlying theories for these tests are response theories. This research analyzes and compares the response theories such as “Item Response Theory” (IRT) and “Classical Test Theory” (CTT). Included in this research is literature reviews consisting of CTT and IRT background knowledge and previous works done in this field. An in-depth study was conducted on all “Item Response Theory” Para models (namely IRT-One Para model, IRT-Two Para model, and IRT-Three Para model)along with their parameters for ability, discrimination and probability respectively to present an improved Para model; “Time IRT” Model or TIRT Model. In the proposed fourth model, i.e. TIRT model, the response and the response time calculations are taken as to be associated with distinct parameters.
