The Development of an Automatic Test Assembly System for a Formative Assessment in Mastery Learning Instruction: Case of the SQL Mastery Course

The teaching method of mastery learning requires the lecturers to conduct their regular formative assessments so as to monitor each student's learning progress. However, regularly preparing quality test forms for these assessments has caused a tremendous workload on the lecturers and has prevented them from effectively executing the master learning instruction. To resolve the problem, this study has built an automatic test assembly system with the automatic and manual test assembly functions for preparing fixed-length testing forms. Our system is implemented in a Structural Query Language (SQL) mastery course that helps students prepare for the ORACLE SQL Certification (OCE SQL) examination. OCE SQL is a program offered by the Oracle Corporation to validate and recognize the skills of professionals in the SQL programming. By applying the Hambleton and De Gruijter's method, our system can automatically assemble items to a test form with a quality indication of the maximum probability of misclassifying mastery. Furthermore, our system can evaluate the prementioned probability for a manually assembled test form. The present study has also investigated the numeric methods for converting the domain cutoff score to the latent ability scale for items with heterogeneous parameters when applying the method. The empirical computing results showed that the secant method is superior to the bisection method for the conversion. This study contributes an automatic test assembly system to the engineering education practice so as to lighten the lecturers' burden of preparing quality assessments and to further facilitate the execution of the regular formative assessments in the mastery learning instruction.