Publication: Masuring secondary school education effeciencies of Turkish provinces: an application of data envelopment analysis
Abstract
TÜRK İLLERİNİN İKİNCİ ÖĞRETİM EĞİTİM VERİMLİLİKLERİNİN ÖLÇÜLMESİ: BİR VERİ ZARFLAMA ANALİZİ UYGULAMASI İlleri (Karar Verme Mekanizmaları-KVM) ÖSS sonuçlarına göre hangi ilin daha iyi eğitim sunduğunu bulmak adına sıralamak sıkça gerçekleştirilen uygulamalardandır. Gazeteler, TV kanalları dikkat çekmek amacıyla bu tip sıralamalara yer vermektedirler. Fakat eğitim planlamacıları ve yöneticileri için daha önemli ve değerli olan KVM'lerin çevirim işlemlerinin verimliliğine göre performans değerlendirilmesinin yapılmasıdır. Yani bu, girdilerin çıktılara ne kadar verimli şekilde dönüştürüldüğüdür ki bu da bu tezin esas amacıdır. Bu tezde, belirtilen amaç için Türk illerinin (gerçekte Milli Eğitim Bakanlığı ve il eğitim müdürlüklerinin) ikinci öğretim okullarının 2003-2004 eğitim dönemindeki teknik ve ölçek verimliliği ölçüldü ve değerlendirildi. Okul/ Öğrenci, Sınıf/ Öğrenci ve Öğretmen/ Öğrenci oranları girdi olarak, ÖSS (Öğrenci Seçme Sınavı) puanları çıktı olarak kullanıldı. 5 değişik VZA (Veri Zarflama Analizi) modeli; 1 aşamalı CCR ve BCC modelleri, 2 aşamalı Ray ve Tobit modelleri ve Ruggiero'nun 3 aşamalı modeli olmak üzere verimlilik ölçümü için kullanıldı. 2 ve 3 aşamalı modeller ortalama hane-halkı büyüklüğü, dershaneye giden öğrenci sayısı gibi kontrol edilemeyen değişkenlerin hesaba katılması için kullanıldı. Bu değişkenlere sosyo-ekonomik faktörlerin verimlilik ölçümünün etkilememesi ve saf okul verimliliğin ölçülmesi amacıyla yer verildi. 1 aşamalı model sonuçları, girdileri verimli kullanma anlamında doğu illerinin batı illerinden daha verimli olduklarını gösterdi. Kontrol edilemeyen değişkenleri hesaba kattığımızda iki bölge arasındaki verimlilik farkları daha da arttı. Ayrıca, korelasyon değerlerinden çıkarım yapıldığında birçok ildeki okulların özellikle de doğudakilerin yeteri kadar öğretmen sayısına sahip olmadıkları, bunu sonucunda daha başarısız sonuçlar verdikleri gözlemlendi. Verimlilik, VZA, Eğitim, Kontrol edilemeyen, Regresyon, Tobit, Çok-aşamalı 07.2005 M. Eray Kıranoğlu
MEASURING SECONDARY SCHOOL EDUCATION EFFICIENCIES OF TURKISH PROVINCES: AN APPLICATION OF DATA ENVELOPMENT ANALYSIS It is a common practice to rank provinces (Decision Making Units-DMUs) according to their average ÖSS scores for showing which province is better in terms of educational achievement. Newspapers, TV channels do that kind of comparisons to attract people's attention. But for educational managements and policy-makers, it is more important and valuable to evaluate performances of DMUs by looking at effectiveness of their conversion process. That is, how efficiently inputs are converted to outputs, which is the main purpose of this thesis. In this thesis, for mentioned purpose, I measured and evaluated technical and scale efficiencies of Turkish secondary schools over the period 2003-2004 by taking provinces (in fact Ministry of National Education and its provincial directories) as DMUs. School/ Student, Class/ Student and Teacher/ Student ratios are used as inputs whereas ÖSS (Student Selection Exam) scores as outputs. Five different DEA (Data Envelopment Analysis) models which were one-stage basic CCR and BCC models, two-stage Ray and Tobit models and Ruggiero's third stage model were used for measuring efficiencies. Two and third-stage models were used for incorporating non-discretionary (non-controllable) variables into the analysis such as average family size, number of students attending to support courses, etc. These variables were necessary for externalizing socio-economic factors and measuring pure school efficiencies. The 1st stage results showed that eastern provinces were more efficient relative to western provinces in terms of efficient input usage. After taking non-controllable variables into account the efficiency gaps between two regions were widened. Also inference from correlations, we found the fact that most of the schools especially at eastern regions are not adequately staffed resulting in poor exam performances. Efficiency, DEA, Education, Non-discretionary, Regression, Tobit, Multi-stage 07.2005 M. Eray Kıranoğlu
MEASURING SECONDARY SCHOOL EDUCATION EFFICIENCIES OF TURKISH PROVINCES: AN APPLICATION OF DATA ENVELOPMENT ANALYSIS It is a common practice to rank provinces (Decision Making Units-DMUs) according to their average ÖSS scores for showing which province is better in terms of educational achievement. Newspapers, TV channels do that kind of comparisons to attract people's attention. But for educational managements and policy-makers, it is more important and valuable to evaluate performances of DMUs by looking at effectiveness of their conversion process. That is, how efficiently inputs are converted to outputs, which is the main purpose of this thesis. In this thesis, for mentioned purpose, I measured and evaluated technical and scale efficiencies of Turkish secondary schools over the period 2003-2004 by taking provinces (in fact Ministry of National Education and its provincial directories) as DMUs. School/ Student, Class/ Student and Teacher/ Student ratios are used as inputs whereas ÖSS (Student Selection Exam) scores as outputs. Five different DEA (Data Envelopment Analysis) models which were one-stage basic CCR and BCC models, two-stage Ray and Tobit models and Ruggiero's third stage model were used for measuring efficiencies. Two and third-stage models were used for incorporating non-discretionary (non-controllable) variables into the analysis such as average family size, number of students attending to support courses, etc. These variables were necessary for externalizing socio-economic factors and measuring pure school efficiencies. The 1st stage results showed that eastern provinces were more efficient relative to western provinces in terms of efficient input usage. After taking non-controllable variables into account the efficiency gaps between two regions were widened. Also inference from correlations, we found the fact that most of the schools especially at eastern regions are not adequately staffed resulting in poor exam performances. Efficiency, DEA, Education, Non-discretionary, Regression, Tobit, Multi-stage 07.2005 M. Eray Kıranoğlu