Application of genetic algorithm-based automatic test in cardiovascular physiology

AIM and METHODSDatabase-based automatic test, a potential application of computer-aided instruction, generates customer-defined examination paper to evaluate students’ ability and knowledge and avoids subjective factors involved in examination. However, database-based automatic test is not widely used related to lack of well-established algorithms. Genetic algorithm has been reported as one of the most effective algorithms in automatic test. Thus, the present investigation established a genetic algorithm-based automatic test system and used this system in generating examination papers in evaluation of students’ ability and knowledge in cardiovascular physiology. RESULTS The results demonstrate that the system could well control the quality of examination papers. The result variation of different examination papers generated at the same level used for the same examinees was less than 10% and no significant difference was found when different papers were used for different examinees, while randomized algorithm produced significant difference. CONCLUSION Genetic algorithm-based automatic test system can be used in evaluation of students' ability and knowledge, which is independent of the paper itself.