Madde Tepki Modellemesinde Genellenebilirlik İle İki Yüzeyli Desenlerin İncelenmesi
Özet
An approach called generalizability in item response modeling (GIRM) is investigated with two facets bx(m:t) design and results are compared with results of generalizability theory in this study.In this study simulated data is used. Three different models are employed to generate data for having unbiased results. This models are named respectively generalizability theory (GT), uncorrelated testlet response theory (TRT) and correlated testlet response theory. In GT linear model random facets bx(m:h) design are used for generating data. The second model is generated as the multidimensional TRT and referred to as uncorrelated MTRT. The third model adopted is the same MTRT model but there are correlation between general ability and testlet effect. Therefore this model is referred to as correlated MTRT.Generated data are differed by factors. This factors are testlet effect, correlations of latent traits, testlet length and number of testlet.