大学英语四级考试成绩预测模型构建与实证分析

Model Construction and Empirical Analysis of College English Test Band 4 Achievement Prediction

通过选取东南大学120名本科生英语成绩,以全国英语四级考试成绩为因变量,以学生大学英语课程期末考试成绩为自变量,构建多元线性回归数学模型,并得到了回归系数的数值。对回归模型以及回归系数的显著性检验表明本科生大学英语课程学习成绩与CET-4成绩具有高度的相关性。最后,根据已建立的预测模型方程预测出120个样本的CET-4成绩与其实际成绩进行对比发现,两项成绩基本吻合,这也进一步表明了预测模型的准确性。

Based on the English examination scores of 120 undergraduate students of Southeast University, a regression model of test achievement evaluation was established after the index was quantified. Used the students＇ scores for College English Test Band 4 （CET-4） as the dependent variable and students＇ final exam performance of college English course as the independent variable, the mathematical model of multiple linear regression was established, and the regression coefficient values were obtained. The regression model and the coefficients significance test showed that the final exam scores of college English course and CET-4 scores have a significant correlation. At last, we predicted 120 samples＇ CET-4 scores based on the established prediction model equation, and by comparing the prediction scores with the actual scores we found that those two scores are basically in agreement, which further showed that our prediction model was very accurate.