Research On Data-driven Mathematical Literacy Evaluation Method For Primary School

Mathematical literacy is a key factor that accompanies students’ lifetime growth,and it is an important guarantee for students to grow into talents who are good at thinking and caring about society.The current mainstream mathematical literacy evaluation methods,such as PISA,are all result-based tests.The test scores are used to evaluate the development of students’ mathematical literacy,which lacks the evaluation of students’ mathematical literacy development in the learning process.At the same time,at the elementary school stage,students already have preliminary mathematical consciousness.It is of great significance to lay a solid foundation for the development of students’ mathematical literacy at the source and lay a solid foundation for students’ lifelong learning.For these purposes,based on related research,this study designed the selection method for the evaluation indicators of primary school mathematics literacy from the data-driven perspective,discussed the applicability of data mining algorithms based on the characteristics of the indicators,and proposed elementary students mathematical literacy evaluation model based on data mining.This method establishes a student mathematical literacy model by analyzing students’ daily academic and behavioral data.According to the model,real-time evaluation of students’ mathematical literacy can be achieved from both individual and group aspects of students.This method not only allows teachers to keep abreast of the student’s individual ability development but also assists teaching decision-makers in formulating corresponding teaching plans.This study combs the relevant research theories,methods,and techniques from three aspects:education evaluation,mathematical literacy,and data mining.On this basis,based on Taylor’s behavioral goal theory and related research,the thoughts of student behavior observation and analysis are determined.Mathematical literacy-related research discusses the definition,connotation,and evaluation methods of mathematical literacy.Data mining-related technologies and methods are the technical means to support this research.Based on the data mining features of students,the evaluation is more objective and has better evaluation standards.Based on this,the research carried out the following work:Based on the analysis of the current mainstream mathematical literacy evaluation framework and indicators and the actual situation of primary school education in China,from the data-driven perspective,combined with relevant mathematical literacy evaluation indicators and data mining technology,this study proposes the data-driven mathematical literacy evaluation model of primary school students.Focusing on the concept of mathematical literacy,primary mathematical literacy modeling methods are discussed from three aspects:clustering modeling,data fusion rules,and mathematical literacy prediction.Based on the data-driven primary school mathematical literacy evaluation model,the method of combining Markov chain and two-step clustering was used to evaluate the students’ mathematical communication competency.In this study,based on the authoritative test frameworks,combined with China’s mathematics curriculum standards and students’ daily behavior and academic data,the appropriate evaluation indicators were selected,and the classroom behaviors and assignments of two students during a semester were collected through tracking observation.By analyzing students’mathematics daily homework,classroom behavior,extracurricular behavior,and other factors,a student evaluation data set is formed.The data collected each time is regarded as a moment,forming a time series of students’ mathematical communication competency development in the semester.The data of this series is used to determine the student’s mathematical communication competency development level through two-step clustering,and then the Markov chain is used to analyze the probability that the student is at each level and the conversion probability between each level to evaluate the student’s current ability development status.Aiming at the problem-solving competency requirements of primary school,this study refers to multiple problem-solving evaluation frameworks,discusses the connotation of mathematical problem-solving competency of primary school students,and conducts the actual solution process of students’ problem-solving part in the simulation test,and according to the corresponding evaluation indicators,the actual answering process of the problem-solving part of the students in the simulation test is analyzed and quantified.The PAM algorithm was used to analyze the development of students’mathematical problem-solving from the overall perspective of the class.The results show that the algorithm has a good result on the evaluation of students’ mathematical problem-solving competency.The students are differentiated and can explain their specific characteristics,which can help teachers to understand the development of students’ mathematical problem-solving competency on time.To verify the effectiveness of the data-driven primary school mathematical literacy evaluation method,this study designed a comparative experiment,by analyzing the daily learning behavior and academic data of the two classes in a semester,the students’ literacy development is evaluated from the perspective of mathematical literacy-related competencies,and providing data support for teachers’decision-making,and finally,according to the evaluation of students’ mathematical literacy development and teachers’ evaluation.The experimental conclusion shows that when there is no significant difference between the pre-test data and the two classes,the post-test data shows that the overall development status of the mathematical literacy of the experimental class is better than the control class,and it explained that the data-driven mathematical literacy evaluation method can be used in actual evaluation activities and can achieve certain results.Based on the above series of work,this study draws the following conclusions:(1)Data-driven mathematical literacy evaluation method is feasible,(2)The evaluation model proposed in this study can achieve a rapid,objective and detailed evaluation of mathematical literacy,(3)The data-driven mathematical literacy evaluation method can mine students’ academic characteristics and provide data support for teachers’ decision making.In summary,this study proposes a data-driven mathematical literacy evaluation method based on educational evaluation related theories and mathematical literacy evaluation framework.This method combines data mining technology to achieve the processivity,precision,and long-term evaluation of primary school mathematical literacy.It provides a new perspective for the evaluation of primary school mathematical literacy,and also develops corresponding evaluation tools.Finally,the research is compeleted from the Theory-Method-Technology-Tool thinking.In this study,there is still space for improvement in the data dimension of mathematical literacy evaluation.In the following research,if we can combine more efficient data collection technology in an information-based teaching and learning environment to obtain more types of students’ data.It is believed that this study can achieve a more comprehensive and complete evaluation of mathematical literacy.