|
The study of fifth grade to "cross-denominator fraction addition and subtraction" unit for the scope of the study, according to the teaching objectives, sub-skills, the type of error, so as to Bayesian inference network as a tool to establish a set of Bayesian network to computer-based diagnostic tests national minority subjects, the diagnosis of students' error types, provide teaching reference. The results are as follows: First analysis of "cross-denominator fraction addition and subtraction, "the teaching materials, the establishment of the concept of proposition from proposition based on the concept of a computerized diagnostic test, sampling by using Bayesian network as a post-analysis tools, inference and sub-types of errors the students skills. The results are as follows: 1. The self of the fifth grade of elementary school mathematics "different denominator fraction addition and subtraction,"unit test questions, the test of Cronbach's α value of 0.737, with questions of this study showed good reliability. 2. Students of computer-based diagnostic expert identification of the type of error and the correct type of error rates were 95.3333%, 95.3333%, 93%, 89%, construction problems of diagnosis the average correct response rate of 93.1667%. 3 Four models of sub-skills to identify the type of error rate and pattern of the overall average recognition rate were 90%, display mode helps to enhance the recognition rate.
|