Impact of Parental Psychological Properties on Children’s Academic Performance Measured by Machine Learning Technology

Parents’ psychological properties influence their children’s academic performance. The internal relationship between these two is complex, and it is difficult to conclude with qualitative research. Therefore, the children’s academic performance and the parents’ psychological properties were investigated in this study again as the dependent and independent variables. Firstly, the problem was defined as a binary classification mathematical problem. Secondly, a questionnaire survey was carried out among the parents of primary school students in grades 4-6 in a region of China to obtain the data. Then, the classification model was constructed by using the decision tree algorithm in machine learning and the Xgboost algorithm in ensemble learning. Rules between the dependent and independent variables were mined. The experimental results showed that the F1-score of the decision tree model and Xgboost model were 81.9 and 89.7%, respectively. The decision tree model was explanatory significantly. Three variables were identified to show the greatest impact on children’s academic performance: Positive emotions, Parent-child relationship, and Educational anxiety. The accuracy of the Xgboost model was higher than that of the decision tree model, which indicated that parents’ psychological properties had complex effects on children’s academic performance in addition to the above three variables.