Use this URL to cite or link to this record in EThOS:
Title: Teacher and school effects on student achievement
Author: Villegas, Mario F. S.
ISNI:       0000 0004 5992 3520
Awarding Body: University of Warwick
Current Institution: University of Warwick
Date of Award: 2016
Availability of Full Text:
Access from EThOS:
Access from Institution:
This thesis investigates the topics related to the estimation of Teacher Effectiveness (TEs) and School Effectiveness (SEs). With regard to TEs, the focus is on the predicted teacher impact on improving academic achievement when a student from the median (or the mean) of the achievement distribution is exposed to a more e�ective teacher. Regarding SE estimates, we explore the evolution of SEs over a specific period of time. In both cases, we base our estimations on the Chilean education system, from which we have exclusive access to very rich datasets. Our main objectives are summarised as follows: (i) to consistently estimate TEs and SEs using Value Added Models (VAMs), studying the most common estimation approaches used in the literature, and the required assumptions on which they are founded; (ii) to provide the �rst TE estimates for the Chilean educational context; and (iii) to investigate the evolution of SEs, identifying what factors are associated. The thesis is organised into �ve chapters. In Chapter 2, we present a detailed review of TE estimations based on typical Value Added Models, which are derived from a general achievement function (GAF).We then discuss some estimation methodologies and the validations of the estimations found in the literature. In Chapter 3, we present the data and describe how it is organised, placing special emphasis on the selected sample cohorts and the performance measures used through the thesis. In Chapter 4, we test for evidence of non-random assignment of pupils to classrooms (or teachers) in the Chilean context, in order to examine the random assignment assumption imposed in most of the VAMs. For Chapter 5 and Chapter 6, we choose the VAM that enable us to estimate TEs and SEs simultaneously. We employ the Maximum Likelihood estimation (MLE) methodology and obtain predictions of teacher and school effects from the estimated empirical Bayes (EB) distributions. In both chapters, we discuss the assumptions required to consistently estimate our TE and SE measures using this method. We usually conduct the estimations under two VAM specifications, one with a preset value of the persistence parameter �, and another with an unrestricted value of �. The results suggest that teachers are more able to generate a larger impact on Maths than on Language scores. If a pupil from the median of the standardised examination scores distribution were exposed to 1 standard deviation (SD) more e�ective teacher, she will move up around 9 percentile positions in Language and 12 percentile positions in Maths, in terms of the pupils' ranking by subject. Regarding the SE estimates in the long run, we find that neither downward nor upward trajectories of SEs are explained by differences in observed characteristics, apart from pupil academic performance. We find evidence that trajectories of school effectiveness are associated with the proportion of High (or Low) quality teachers, based on our estimated TEs. We conclude that teachers are important in improving pupil academic performance, and that the level of teacher quality within schools is related to the stability and trajectories of school e�ectiveness in the long run.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: HC Economic History and Conditions