| Abstract: | Abstract
This research take fifth grade "quantitative relations" nine-year mathematics proficiency targets 5-a-03 as an example, the purposes of the research are to design supplementary teaching curricula containing activities applied to teachings, evaluations and remedial instructions, and to investigate the teaching effects after instructional experiments. In this study, the computerized adaptive diagnostic test based on the Bayesian network and knowledge structure theories is combined with the teaching design. Besides, the error patterns usually occurred in this unit is adopted as the diagnostic implement and the Bayesian Network based on probabilistic reasoning is taken as the tool for analysis.
The conclusions resulted from the study are listed bellowed:
1. In addiction to multimedia-based curriculums like PPT, the digital presentations used in the mathematical teachings and multimedia materials adopted to the remedial instructions, manuals for teachers, lecture notes and handbooks for students are also included in the digital and remedial instructions.
2. By using the digital curricula developed by the researcher for the remedial instructions, it is found that students of the experimental group obtain higher average score in the pretest than students of the control group (91.95>87.14). The remarkable difference after the examination shows that the self-designed digital curricula are effective for students’ learning.
3. By integrating information technology into remedial instructions, the average scores for students of the experimental group in the pretest is 91.95,the average scores in the post-test is 96.25.Comparing with these scores , the difference attained by the examination shows that the remedial teaching materials designed by the researcher are truly helpful for students’ learning.
4. The average grade which the experimental group students get in the postpone test is 93.52. On an average, the control group students score 83.78 points in the delayed test. The fact indicates that students of the experimental group get higher grades than students of the control group. The notable difference attained by the examination reaches as high as 9.74. It shows that the continuous effects resulted from the digital curriculum designed by the researcher on the experimental group is superior to the continual efficiency on the control group.
5. Experimental group, the average error patterns Incidence of post-test reduce than the pre-test by 9.5%; control group, the average error patterns Incidence of post-test measured less than the pre-test by 5.2%. Experimental group, the average error magnitude of less big than the control group.
Experimental group, the average post-test sub skills reached a rate counted was 94.05% in the control group's post-test sub skills reached a rate counted for the 76.53 percent average of the experimental group, the average increase reached a magnitude of rate counted obviously much larger than the control grou Great many.
6. It is estimated that over 58% of questions can be saved by using the Bayesian Network based Computerized Adaptive Test (BNAT) and the accuracy rate of predication is above 93%.
7. When the computerized adaptive diagnostic test deduced from the Bayesian network is done to the full and well adapted, the uniformity rate of the error patterns and sub-skills can reach 93.78% in the pretest, 96.61% in the post test and 95.87% in the postpone test respectively.
Keyword：quantitative relationships , knowledge structure , Bayesian networks, Integration of Information into Teaching , the computerized adaptive diagnostic test |
