A Study on the Path of Enhancing English Learners’ Discourse Organization Skills Based on Deep Learning and Evaluation Research

In this section, a smart classroom teaching model based on the concept of deep learning will be constructed, and the teaching activities will be designed in four aspects: smart pre-study, smart classroom, smart enhancement, and smart evaluation. This teaching model emphasizes that the teacher adopts the teaching strategies of guidance, inspiration, exploration and contextualization to achieve independent pre-study before class, immersive learning in class, and innovative development after class, with the focus on helping students to get ready before class by guiding, motivating, exploring, and creating a conducive environment, so that the students can have a better understanding of the English knowledge learnt in this class, and complete the teaching tasks better in class to They can fully learn the classroom knowledge points and improve their collaborative learning ability in class, and they can make use of the automatically generated learning report on the intelligent teaching cloud platform to reflect and summarize their learning situation after class. In the course evaluation, the full use of wisdom education cloud platform end, the use of self-assessment, mutual assessment, interactive exchanges, feedback and other ways to promote the improvement of students’ self-reflection ability, to promote the improvement of the cognitive system, the application of knowledge and migration to achieve the development of innovation, and in the process of English language teaching, the students are always the main body of learning, to achieve the synergistic development of teachers and students and the growth of teaching and learning.

There are two components of the pre-laboratory test, one is learning interest and the other is achievement. For the first task, the author will use the Questionnaire on Students’ Learning Interests to assess the learning interests of students in Class A and Class B. The questionnaire has good reliability and validity, including three dimensions of logical, critical and creative thinking, each dimension contains three indicators and a total of 21 questions, the specific preparation of which has been described in the previous section and will not be repeated here. For the test of reading achievement, for the consideration of focusing on the quality of students’ thinking, the author widely read the reading test questions and books, and after screening, selected the test questions focusing on the quality of students’ thinking, and pre-tested the students’ reading achievement, and the teacher strictly supervised the test to ensure the fairness and objectivity of the test.

The second stage is experimental teaching and reflective improvement, in which teachers engage in instructional design, instructional implementation, and instructional reflection. Under the deep learning perspective, the instructional model has three phases-before, during, and after class, with six segments-pre-reading preparation, classroom introduction, information sorting, deep reading, post-reading generation, and summarizing and reflecting. Before class, teachers need to pre-assess the students and the content of the unit. Pre-assessment of students’ basic level of English, learning characteristics and prior knowledge, as well as reading the content of the whole unit, sorting out the logical relationship between each lesson, and designing the teaching from the perspective of the unit as a whole. At the same time, the author got along well with the students in the experimental class during the internship, creating a positive classroom atmosphere and a good teacher-student relationship, laying the foundation for the next experimental teaching. The students need to consolidate their basic knowledge of discourse organization before class in order to participate more efficiently in the English classroom.

The in-class stage includes four parts: classroom introduction, information sorting, deep reading and post-reading generation. Classroom introduction can take the form of quick questions and answers, brainstorming, and group discussion to mobilize students’ thinking, stimulate their learning motivation, and activate the existing knowledge and experience in their brains. After a smooth introduction, students begin to read the content of the text, complete the reading tasks assigned by the teacher, and understand the main points of the text to sort out the information. This part is mainly task-oriented, leading students to think through the task. Teachers can use teaching activities such as judging the correctness or incorrectness, summarizing the main idea, matching information, mind mapping, and so on. After completing the information sorting of the text, it enters the deep reading section. Teachers can guide students to further deconstruct the text through skillful questioning. Activities include reasoning about details, analyzing characters, inferring style and main idea, comparing cultures, and identifying the author’s viewpoints and attitudes. This session is an important part of developing the quality of students’ thinking. In the post-reading generation segment, students can apply what they have learned, criticize, and innovate. Teachers need to design a rich variety of activities, so that students in the experience of revitalizing thinking, can be taken to activities such as title inquiry, reading emotions, article retelling, dialogue creation, group debates, reading followed by writing and so on.

Post-lesson stage is mainly evaluation and reflection, teachers conduct multiple evaluation of students, reflect on the effectiveness of teaching to improve the next teaching, students reflect on their own learning to remedy the shortcomings of the improvement. Teachers need to reflect on the achievement of learning objectives, classroom activities, student participation, teaching highlights and shortcomings, students reflect on the knowledge mastery of the lesson and classroom engagement, specific self-assessment of the learning effect of self-assessment form, written reflection, and feedback from peers and teachers and other reflection methods.

The third stage is the experimental post-test, which mainly includes the learning interest post-test and achievement post-test for the students in both classes, as well as semi-structured interviews with the students in the experimental class. After finishing the experimental teaching, the author utilized the test paper with the same question type and similar difficulty as the pre-test writing test questions to conduct the post-test on the students of the experimental class and the control class at the same time, and imported the reading scores of the two classes into SPSS for independent sample test to verify whether and how much the method proposed in this paper affects the writing scores of the experimental class students from the scientific point of view. In addition, the author used the Learning Interest Questionnaire to post-test the students in the two classes, and analyzed the data with SPSS as well, to detect whether there is any change in the students’ learning interest after the experimental teaching. For the effect of experimental teaching, the author also randomly invited 10 students from the experimental class to conduct semi-structured interviews in order to understand the advantages and shortcomings of the teaching mode, which can be easily improved and perfected.

Counterfactual: In order to test the validity of a “hypothesis”, it is assumed that the “hypothesis” is valid, and then see what kind of results will be derived from it. If there is an unreasonable phenomenon, then it shows that the original assumption is incorrect, that is, the “hypothesis” is not valid. Therefore, we reject the hypothesis. If no unreasonable phenomenon is introduced, then the original “hypothesis” cannot be rejected, and the original hypothesis is said to be compatible.

It is also different from the counterfactual method in pure mathematics, because the so-called “unreasonable” here is not an absolute contradiction in formal logic, but is based on a principle widely used in people’s practice: small probability events can be considered to be basically unlikely to occur in a single experiment.

In multiple regression analysis, when the dependent variable y is interfered by external influences, these influences are defined as independent variable x_i. If there is a certain correlation between y and x_i, a multiple regression model can be established based on the dependent variable y and the various influences x_i, and the mathematical expression of the model is as follows: 1y1=β0+β1xt1+β2xt2+…+βpxtp+εt

The study utilized SPSS 27.0 to analyze the pre- and post-test data of the questionnaire with descriptive statistics. The paired sample statistics are shown in Table 1. It can be seen from the table that before the method proposed in this paper was used for teaching implementation, the scores of students’ emotional dimension were only slightly higher than 3 points, the scores of cognitive dimension were only slightly higher than 3 points for “application”, and the scores of “analysis”, “evaluation” and “creation” were all lower than 3 points. After the instructional implementation using the methodology of this paper, the students were in the range of 3 to 4 on each of the affective and cognitive dimensions and the means under each dimension of the posttest of the questionnaire (4.048, 3.769, 3.732, 3.298, 3.31, 3.278) were higher than the means of the pretest (3.22, 3.145, 3.112, 2.939, 2.873, 2.791). This results show that the method in this paper has a certain effect on improving students’ “interest”, “motivation”, “application”, “analysis”, “evaluation” and “creation” in the organization of English learners’ discourse. Therefore, implementing English teaching based on the method proposed in this paper has a positive impact on learners’ interest in discourse organization ability.

The paired sample test is shown in Table 2. As can be seen from the table, the corresponding p-value of the pre- and post-test data for each of the dimensions of students’ affect and cognition is 0.000 (p<0.05), which indicates that there is a significant difference between the pre- and post-test questionnaires, i.e., English language teaching based on the methodology proposed in this paper is able to produce a positive impact on both students’ affect and cognition.

In order to verify whether there is a facilitating effect on students’ English performance based on the methodology of this paper, the researcher tested the English language of the students in the two classes before and after the action research. In this study, SPSS was used to conduct a paired-sample t-test on the reading pre-test and post-test scores of the students in the class, and the paired-sample statistics are shown in Table 3. As can be seen from the table, the mean value of the students’ pre-test scores is 24.512, while the mean value of the post-test scores is 27.315, which indicates that the students’ scores have been improved after the implementation of the methodology of this paper for teaching. The standard error of the students’ pre-test scores is 1.0237 and the standard deviation of the post-test scores is 0.7951, which also indicates that the differences in English scores among students are being reduced after using the method proposed in this paper for teaching English.

The paired samples test is shown in Table 4. As can be seen from the table, the significant two-tailed Sig value of the pre and post-test English scores of the study participants is 0.000 (p<0.05), which means that there is a significant difference in the students’ English scores between the two times. This paper suggests that teaching English using the proposed method is beneficial for improving students’ English achievement and has a positive impact on their English learning.

In this study, English scores were categorized into three score ranges: high, medium, and low subgroups, with high subgroup scores greater than 30, medium subgroup scores between 22.5 and 30, and low subgroup scores less than 22.5. The mean score of students’ pre-test reading scores was 23.889, with 18 students in the low group, 26 in the middle group, and 10 in the high group. The distribution of pre and post-test English reading scores in the intervals is shown in Table 5. As can be seen from the table, the mean score of students’ English reading achievement increased to 27.547, the number of low grouping decreased significantly by 12, and the number of high grouping increased by 4. This shows that after practicing English teaching by applying the method proposed in this paper, students’ English scores are not only significantly improved, but also the trend of students in the low group is reduced, and the difference in students’ English scores is gradually reduced.

Through Pearson correlation analysis, we can derive the correlation coefficients and significance of each measure of discourse organization with English achievement. In order to further examine the explanatory or predictive power of each measure of discourse features on English achievement, multiple step-by-step regression analysis was used to test it, with each variable of the discourse step features as the independent variable and English achievement as the dependent variable. The regression model summary of the discourse organization measurement index and the regression coefficient of the discourse organization measurement index are shown in Table 6 (a. Prediction conjugation: (constant), the reason for explaining the number of sentence steps. b. Predictor variables: (constant), the number of sentence sentences explained by reason, and the number of sentence sentences evaluated. c. Prediction conjugation: (constant), reason explanation of the number of step sentences, evaluation of the number of step sentences, DM. d. Variants: Grades) and Table 7.

From the R² and the adjusted R² in Table 6, it can be learned that Model 3 has the three variables of the number of small sentences in the causal expository discourse step, the number of small sentences in the evaluative discourse step, and the discourse markers entered into the regression model, and together they were able to explain about 65.8% of the variance of the English grades. From the value of standardized regression coefficients, it can be seen that the influence of these three independent variables on the dependent variable of English achievement degree of explanatory power, their three variables of standardized regression coefficients were 0.711, 0.263 and 0.183, respectively. the last third model of the estimated standard error of 4.81244, respectively, much lower than the standard deviation of the English achievement of 61, which suggests that the three variables of the English achievement of the prediction of the more accurate.

From the above correlation and multiple regression analyses, it can be seen that the greatest degree of correlation with performance is the number of steps, the number of clauses and the discourse markers of the cause-explanatory discourse step. This shows that the relevance, adequacy and organization of core steps are the most important factors of English quality, and because of the strong generality of discourse steps and the high degree of homogeneity of the step structure of the same discourse type, the characteristics of the discourse steps are not significant to the achievement in general. In addition to the greater explanatory power of the number of clauses in expository discourse steps for English performance, the number of clauses in evaluative discourse steps also has some explanatory power for English performance, which suggests that the completeness of the discourse step is also an important factor influencing raters’ scores. Finally, the use of discourse markers enhances the organization of the discourse, and accordingly, the higher the English scores.

In order to explore whether the parameters and measures of discourse organization have discriminatory power on the level of English achievement, the subjects in this study were divided into three groups according to their performance: high and low. The high subgroup of 20 individuals scored between 71.35 and the highest score in English (M=76.21, SD=4.321). The middle subgroup of 20 individuals had scores between 63 and 72.51 (M=65.12, SD=2.412). The low subgroup of 20 individuals scored between 62.51 and the lowest score in English (M=61.24, SD=3.150). The results of descriptive statistics and ANOVA for the discourse organization measures at different achievement levels are shown in Table 8. The results in the table show that the test results for the number of reasoned elaboration discourse steps, the number of reasoned elaboration discourse mini-clauses, and discourse markers reached significance (p < 0.05).

To further test the discourse organization measures at different achievement levels, the multiple comparison analysis of the discourse organization measures is shown in Table 9. As can be seen from the table, specific differences in the high scores for the number of clauses in the explicative discourse step were analyzed in this section with post-hoc multiple comparisons. Significant differences were presented between the groups, the middle group and the low group, indicating that the number of clauses in the explicative discourse step can be used to differentiate between different achievement level groups. The number of steps in the expository discourse and discourse markers was only significantly different between the high and low subgroups. The other two measures failed to show significant differences and were not able to distinguish between achievement level groups. These results can be summarized by saying that the number of clauses in expository discourse steps has a greater discriminatory power for achievement levels. Second language level is the most important factor affecting students’ discourse structure awareness and structural completeness of speech samples; the higher the second language level, the stronger the discourse structure awareness, the more structural completeness of the utterance step, and the more adequate the output of the utterance step. Finally, as a whole, the illustrative discourse produced by English majors is still loose in structure, and there are still some deficiencies in expository adequacy and discourse markers.