Publication: EĞİTİM BİLİMLERİ ALANINDAKİ LİSANSÜSTÜ TEZLERDE VARSAYIM KULLANIMI
Abstract
Bu araştırmanın amacı, Eğitim Bilimleri alanında yayınlanmış, lisansüstü tezlerin, ilgili istatistikselanalize dayalı olarak, sahip oldukları varsayımların test edilme durumlarının değerlendirilmesidir.Betimsel araştırma modelinde olan bu çalışmada, Türkiye’de 2015, 2016 ve 2017 (Eylül ayınakadar) yıllarında Eğitim Bilimleri Enstitüleri tarafından yayınlanmış ve Ulusal Tez Merkeziüzerinden erişimine izin verilmiş 95 lisansüstü tez incelenmiştir. Araştırmacılar tarafındanhazırlanan tez inceleme formu ile toplanan veriler frekans ve yüzde değerlerinden yararlanarakanaliz edilmiştir. Elde edilen bulgular doğrultusunda istatistiksel analizlerin gereği olanvarsayımların yeterli derecede incelenmediği görülmüştür. Özellikle normallik, varyanshomojenliği ve doğrusal ilişki varsayımlarının daha az sayıda eksiksiz ve doğru şekilde testedildiği, diğer taraftan örneklem büyüklüğü varsayımının diğer varsayımlara göre daha fazla sayıdaeksiksiz ve doğru şekilde test edildiği sonucuna ulaşılmıştır.
The purpose of this research is to assess the testing of assumptions, which have been published in the field of Educational Sciences, based on the related statistical analysis of postgraduate dissertations. The study is designed descriptive research model and reviews 95 postgraduate dissertations which are published in Department of Educational Sciences in 2015, 2016 and 2017 (September) and allows full access in the National Thesis Center in Turkey. A thesis review form is prepared by the researchers and the data collected by this form are analyzed based on frequency and percentage values. It has been observed that the assumptions required for the statistical analyzes are not tested sufficiently. In particular, normality, variance homogeneity and linear relationship assumptions are tested in fewer complete and in the correct way, and on the other hand, the assumption of sample size is tested more accurately and complete than other assumptions.
The purpose of this research is to assess the testing of assumptions, which have been published in the field of Educational Sciences, based on the related statistical analysis of postgraduate dissertations. The study is designed descriptive research model and reviews 95 postgraduate dissertations which are published in Department of Educational Sciences in 2015, 2016 and 2017 (September) and allows full access in the National Thesis Center in Turkey. A thesis review form is prepared by the researchers and the data collected by this form are analyzed based on frequency and percentage values. It has been observed that the assumptions required for the statistical analyzes are not tested sufficiently. In particular, normality, variance homogeneity and linear relationship assumptions are tested in fewer complete and in the correct way, and on the other hand, the assumption of sample size is tested more accurately and complete than other assumptions.