Mr Ismail Aslantas
Ismail is a PhD student at the School of Education, Durham University. His research is on the evaluation of teachers’ performance based on student test scores using Value-added Models (VAMs).
After his BA in Education (2003-2007) in Turkey, Ismail worked in various positions in the Ministry of Education in Turkey, as a primary school teacher, deputy principal and principal. Then, with a scholarship from the Turkish government in 2016, he completed his MA at the University of East Anglia, England investigating teachers’ perceptions of formative assessment in the classroom.
In May 2017, he started his PhD at Durham University. The aim of his doctorate research is to examine the stability of results obtained under different data analysis methods in estimating teacher effectiveness using contextual predictors such as observable teacher characteristics, school characteristics and the number of test scores. This study consists of two research approaches; a systematic review to summarise and synthesise evidence from previous research, and a primary study to analyse the stability of teacher effectiveness in Turkish schools measured using Value Added Models (VAMs). The systematic review evaluates primary research to examine the stability of VAMs in measuring teacher effectiveness. After completing the screening process of 1439 records, 50 relevant studies were retained for data extraction.
The nature of the primary research design is retrospective correlational study which is intended to predict the impact at teacher- and school-level of contextual factors on students’ prior test score(s) in five subjects (Mathematics, Turkish, Science, Revolution History and English) using longitudinal data from Turkish schools. The study will also test whether the choice of data analysis methods used on teacher effectiveness affects the stability of the results. Data from 17,005 Grade 8 students taken from 282 elementary schools in (which part of Turkey) has been obtained. Due to missing data at teacher-level (which is necessary for matching pupil performance against teacher effectiveness indicators), the restricted sample size of the estimations used is around 7,000 linked with roughly 200 teachers in each teaching subject.
The main questions of this study are:
1. How much fluctuation in teacher effectiveness can be explained by the observable teacher characteristics?
2. How much fluctuation in teacher effectiveness can be explained by the conceptual school characteristics?
3. Is there a relationship between the fluctuation in teacher effectiveness and the test subjects selected (Mathematics, Turkish, Science, Revolution History and English)?
4. How stable are the teacher effectiveness rankings using one previous year test score, and two previous years test scores?
5. How stable are the teacher effectiveness rankings across running various statistical analyses methods? (the one-step VAM, two- and third-level Hierarchical Linear Regression (HLR), Multiple Logistic Regression)?