Value-added evaluation is an evaluation method that focuses on students’ growth and aims to evaluate school effectiveness more fairly. There are various mathematical models for value-added evaluation. Compared with two models based on average rank and standard deviation, the advantages of models based on student growth percentile and multilevel linear regression are that they can avoid the problem of limited value-added space for excellent students. In the multilevel linear regression model, uncontrollable student factors such as gender and family socio-economic status index can be included to effectively eliminate the influence of student factors on students, which is more in line with the evaluation concept of focusing on the “net effectiveness” of schools.