This study investigated the feasibility of using the three-parameter model in Florida's minimum competency testing program. LOGIST 4 was used to analyze 1984 Statewide Student Assessment Tests (SSAT)-II data, exploring possibilities that easiness of the test would cause problems in the estimation of the a and c parameters. These problems and results of the trial analysis are discussed in chapter 2 of this 7-chapter report. An elaborate investigation of the assumption of unidimensionality of SSAT scores is reported on in chapter 3. In the next two chapters, goodness to fit of the one-, two-, and three-parameter models to the SSAT-II data is discussed, and the results of a parameter estimation investigation are described, along with a consideration of the appropriateness of different models. Chapter 6 reports on studies of the optimum number of examinees to be used in analyzing the data available from Florida's tests, including an attempt to improve parameter estimates by oversampling the lower end of the achievement distribution. A brief chapter of conclusions completes the report. (LMO)