作者: Yeşim Özer Özkan
关键词: Classical test theory 、 Statistics 、 Psychology 、 Item response theory 、 Test battery 、 Student achievement 、 Test (assessment)
摘要: The focus of this research is to test the estimation achievement measurements in battery and empirically compare results after applying classical theory, unidimensional multidimensional item response theory models Student Achievement Assessment Test (OBBS-2008) subtests Turkish Mathematics. It also tries put forward best model that estimates students’ with less error as comparison being made. From analysis test's data results, it identified ability parameters estimated obtained from whole under IRT, have partially scores reached more precise measurement than unidimensional IRT on basis sub dimensions CTT. Similar were mathematics results. Finally, found parameters, within scope scores. Ozet Bu arastirmada, bir bataryasindaki basari olculeri kestiriminin dogrulugunun belirlenmesi ve ampirik olarak Klasik Kurami (KTK), tek cok boyutlu Madde Tepki (MTK) modellerinin Ogrenci Basarilarinin Belirlenmesi Sinavi’nin Turkce matematik alt testi verilerine uygulanarak elde edilen olculerinin karsilastirilmasi amaclanmistir. karsilastirmalar yapilirken olculerini daha az hata ile kestiren en iyi ortaya konulmaya calisilmistir. verilerinin analizi sonucunda tum testten MTK kestirilen yetenek parametrelerinin boyutlar bazinda MTK’ye gore parametreleri KTK’ye puanlarina kiyasla kismen dusuk standart hataya sahip oldugu belirlenmistir. Matematik sonucunda, kestiriminde hatanin gore; yuksek ise testinin boyutlarindan belirlenen puanlardan edildigi