Study On Improving B-learning Performance Based On The Learners’ Statistical Model

Blended learning is indeed the most effective way of learning due to its combination of classroom learning and online learning. But there are still significant performance issues in the actual application process. This kind of the instruction does not gain the expected effect. In the aspect of instructional activities, there is a much larger proportion that classroom instruction accounts for, which easily causes the problem that classroom teaching and online learning are disjointed. In terms of curriculum resources, the percentage of students who make use of the course resources which put a deal of effort into is relatively low. In regard to online learning activities, their enthusiasm for online learning is generally low, while students with high participation online have no significant difference from those with low participation online in academic achievement. In order to promote the integration of information technology and instruction, an increasing number of studies have been devoted to blended learning performance in high university.From the perspective of learning performance factors, this paper analyzed the performance promotion in the context of blended learning according to the deep learning theory. First, this paper conveyed a detailed understanding of the deep learning theory that involves Biggs 3P learning process model and other important views. Then, it integrated these ideas into the in-depth analysis of the relation between the various factors that affect the blended learning performance, and made a blended learning performance model in deep learning process. Based on the factor logical table derived from the B-learning deep learning process model, this paper adopted the interpretative structural modeling method to construct the B-learning instructional analysis model. Further, it took students who participated in ’College Physics’ course in the context of blended learning as research participations. We carried out the correlation analysis of the first batch of data, and it constructed the B-learning holistic performance evaluation model that reflected the overall instructional situation based on the correlation between factors. The second batch of data came from the experimental class under instructional intervention. The combination between data analysis and B-learning model successfully explained the experimental performance structural model construction method. Comparing the B-learning instructional analysis model with the B-learning holistic performance evaluation model could ensure the main problem existing in the present mixture learning, and designed some targeted instructional intervention strategies. Moreover, comparing with the experimental performance model, we could quantitatively evaluate those instructional strategies.To sum up, the data construction of B-learning model can help understand the deep learning process, and develop some effective instructional intervention strategies to accelerate learning, so that learners could go into the in-depth learning and improve their academic achievement in blended learning.