Computerized adaptive test method based on cognitive diagnosis

Abstract

The invention discloses a computerized adaptive test method based on cognitive diagnosis. The method comprises the following steps: S1, establishing a question selection model which requires a large amount of answering data and test question examination knowledge point data; S2, logging in the system by a tester; S3, selecting a first test question, and selecting a test question for testing; S4, after each question is finished, calculating a current knowledge mastering state of the tester; S5, judging whether answering is finished or not through conditions; and S6, ending the test, and outputting a result. The technology provided by the invention is very novel, and compared with the current online test adopting a fixed test paper mode, the technology provided by the invention has the advantages that the test length can be shortened, the time of a testee is saved, and the knowledge mastering level of the testee can be accurately measured; the length of the test can be reduced and testedwith different abilities can be diagnosed by using the self-adaptive test; compared with the traditional test, the self-adaptive test can be used to greatly reduce the test length; and by adopting aShannon entropy question selection method and a GDINA model, the measurement precision can be improved.

Description

technical field [0001] The invention relates to the field of computer application technology, in particular to a computerized self-adaptive test method based on cognitive diagnosis. Background technique [0002] In the traditional test, the length of the test is fixed, and the difficulty and differentiation of the whole test are also uncertain. Only experienced teachers can choose the test questions based on their own experience, and the subjectivity is too strong, resulting in the difficulty of the questions. First, it is impossible to effectively distinguish the real ability level of the subjects; the current online tests use fixed test papers, and the selection of the entire test paper can only be done by the teacher. Each teacher has different experience and may not be able to effectively select the appropriate test paper. The subjects' questions lead to low reliability and discrimination of the whole test, so that the knowledge level of the subjects cannot be effectivel...

AI Technical Summary

Problems solved by technology

[0002] In the traditional test, the length of the test is fixed, and the difficulty and differentiation of the whole test are also uncertain. Only experienced teachers can choose the test questions based on their own experience, and the subjectivity is too strong, resulting in the difficulty of the questions. First, it is impossible to effectively distinguish the real ability level of the subjects; the current online tests use fixed test papers, and the selection of the entire test paper can only be done by the teacher. Each teacher has different experience and may not be able to effectively select the appropriate test paper. The subjects' questions lead to low reliability and discrimination of the whole test, so that the knowledge level of the subjects cannot be effectively tested, and the length of the test is fixed, so some subjects may not need to do so many questions , the level of mastery can be measured, which causes a waste of resources;

Images