Madde tepki kuramına dayalı tek boyutlu ve çok boyutlu modellerin test uzunluğu ve örneklem büyüklüğü açısından karşılaştırılması
Özet
The purpose of this study is; to investigate Turkısh language test data, developed by researcher, under unidimensional and multidimesnional item response theory models and to submit the best model which fits the data. Independent variables that were manipulated in this study, are sample size, test length and both.Eighth grade primary school students (n=1516) in Aksaray have participated in this study. 24 item Turkısh test, developed by researcher, has been used as a research instrument. Every test item contains two latent traits, semantical in language and knowledge of punctuation, which compensates each other. Examinee data, has been analysed under unidimensional and multidimensional IRT models and item and ability parameters have been estimated, model-data fits and error rates have been presented.The findings suggested that item and ability parameters estimated under multidimensional IRT, have less error scores and reached more precise measurement and model data fits are in the favor of multidimensional IRT. Also, these are the other results of this research that sample size variable is the most important variable to affect item parameter estimation and test length is the most important variable to affect person ability parametrer. In addition, it was found that manipulation of sample size and test lenght together has positive effect just on item parameter estimetion. As variables sample size and test lenght together has no effect on person ability astimation and model-data fit in both models.