**2. Methods**

#### *2.1. Participants*

The dataset for this study combines the data concerning students and teachers. Student data mainly refer to the standardized test scores of Chinese, mathematics, and chemistry for students who took the college entrance examination in Haidian District, Beijing from 2016 to 2019, including the first simulated test results of the college entrance examination and the results of the senior high school entrance examination. Specifically, the senior high school entrance exam scores represent the knowledge acquired by students before entering high school, i.e., the entrance scores of high schools. The first simulated test scores of the college entrance exam serve as a proxy variable for college entrance exam scores and represent the exit scores of high schools after students have experienced three years of learning and training in a high school.

Teacher data are derived from the Regional Teaching and Research Survey questionnaire conducted between February and March 2019 for schools in Haidian District, Beijing. The survey aims to understand teachers' demands for professional development and their needs for teaching and research in the new era of educational reform. Based on the globally used questionnaire derived from the TALIS (Teaching and Learning International Survey) regarding teachers' professional development, effectiveness, teaching practices, and classroom behaviors, the questionnaire was developed concerning teachers' current professional development in China [32].

Based on student data, this study matched teacher data with student data according to the names and schools of teachers participating in the survey, and the names and schools of teachers in the student data. Thus, the dataset can be deemed as a combination of both secondary and primary data. The matching resulted in the creation of a multi-tier database containing the scores of the high school entrance examination, the first simulated test of the college entrance examination, and the survey data of corresponding teachers and school names. Through matching, the 542 teacher data from 60 ordinary high schools were linked with the 39,894 student data. The number of students with Chinese, mathematics, and chemistry scores was 14,296, 15,662, and 9936, respectively; and the number of Chinese, mathematics, and chemistry teachers was 195, 216, and 131, respectively [32].
