6 Pages

B. Implementation of Maximum Likelihood Estimation of Examinee's Ability

The second building block of IRT test analysis is the estimation of an examinee's ability. This process assumes that the parameters of the test items and the examinee's item responses, scored right-wrong, are available. Thus, the BASIC program given below assumes that these quantities exist. The program can estimate an examinee's ability using either a one-, two-, or three-parameter logistic item characteristic curve model. Again, the program is structured with a short driver to control the overall flow of the program. The actual ability estimation is implemented in a subroutine.