Use this URL to cite or link to this record in EThOS:
Title: School and teacher effectiveness of senior high schools in western China
Author: Zhang, Lei
ISNI:       0000 0004 5991 4296
Awarding Body: University of Bristol
Current Institution: University of Bristol
Date of Award: 2016
Availability of Full Text:
Access from EThOS:
Access from Institution:
This research uses secondary data from the Improving Teacher Development and Educational Quality in China (ITDEQC) project to conduct new school and teacher effectiveness research for senior high schools in one Local Education Authority (LEA) in western China. The significant substantive findings include: 1) There are remarkable differences in raw estimates of school and teacher effects on student's achievement in Entrance Examination for Higher Education (EEHE) English as well as adjusted estimates when controlling Entrance Examination for senior high schools (EESHS) English score regarding the different teachers and different senior high schools in one LEA of western China. 2)The raw, unadjusted estimates without controlling the prior attainment and five student background characteristics (gender, citizenship status, place of household registration , the number of siblings and tuition fee status) may underestimate the school and teacher effects on student's EEHE English score. 3) Some student background characteristics (e.g. place of household registration and tuition fee status) also have significant effects on student's achievements and progress in English. Moreover, after aggregating these five student background characteristics to the teacher level and school level, the additional contextual effects of some background variables also tend to influence the individual, teacher and school performance in EEHE English. 4) The results also suggest there is marked differential school effectiveness by prior attainment in terms of student's achievement in English. Schools and teachers are not equally effective in promoting the progress of all initial achievers. When compared with top and bottom initial achievers, the average students seem to make the least progress. 5) The effects of prior attainment on student's achievement also differ for students with different citizenship status or their number of siblings. All these findings can make contributions to current education policies in China. For example, due to the lack of the database in China, it is crucial to create a larger electronic database which contains more of student's background characteristics and practical family background information for relatively valid and fair evaluation of teachers and schools' performances in EEHE in western China's context. Also, more policies with respect to equal opportunity for senior secondary education, the optimization of students' allocations across different schools and teachers are needed for the improvement of education quality. Likewise, from the methodological perspective, this research also revealed that the three-level fixed and random effects model can give different estimates and standard error for school and teacher effects because of the different estimation methods used. The bias of the effects of the individual-level covariates in three-level random intercept model can be adjusted by including its cluster-level means. And regarding using weighting to compensate the possible bias for unit non-response in variables from the questionnaire, the standard error of the coefficients of individual-level covariates tend to be overestimated in three-level random intercept and random slope model in our case. Also, the three approaches (deleting all observations with missing teacher ID, the missing category method and multiple membership models method) tend to perform differently when dealing with the missing identifiers in our ITDEQC dataset.
Supervisor: Thomas, Sally ; Leckie, George Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: Multilevel Modelling ; Causal inference ; school effectiveness ; teacher effectiveness ; missing data ; endogeneity ; data harmonisation ; data cleaning ; data management