Vertaile menetelmiä

	Testlet Response Theory	Bayesiläinen hierarkkinen malli
Tieteenala	Education	Bayesilainen tilastotiede
Menetelmäperhe	Latent structure	Bayesian methods
Syntyvuosi	2007	2006
Kehittäjä	Howard Wainer, Eric Bradlow & Xiaohui Wang	Gelman & Hill (2006); Bayesian multilevel tradition
Tyyppi	Item response model accommodating local dependence within item bundles (testlets)	hierarchical probabilistic model
Alkuperäislähde	Wainer, H., Bradlow, E. T., & Wang, X. (2007). Testlet Response Theory and Its Applications. Cambridge University Press. ISBN: 9780521681261	Gelman, A. & Hill, J. (2006). Data Analysis Using Regression and Multilevel/Hierarchical Models. Cambridge University Press.
Rinnakkaisnimet	TRT, Testlet Models, Random-Effects Testlet Model, Item-Bundle IRT	multilevel Bayes, Bayesian multilevel model, Bayesian HLM, partial pooling model
Liittyvät	4	4
Tiivistelmä	Testlet response theory (TRT) extends item response theory to tests built from testlets — bundles of items sharing a common stimulus, such as several questions about one reading passage. Standard IRT assumes items are conditionally independent given ability, but items within a testlet violate this because they draw on the same passage. TRT adds a testlet-specific random effect that absorbs this local dependence, preventing the overstated precision and biased parameters that result from ignoring it. Developed by Wainer, Bradlow, and Wang, it is widely used wherever passage-based or scenario-based items appear.	Bayesian hierarchical modelling, popularised by Gelman and Hill (2006), is a Bayesian approach to nested data structures — such as students within schools within districts — that estimates separate parameters at each level while allowing those levels to share statistical strength through a mechanism called partial pooling. Where a classical hierarchical linear model treats group means as fixed unknown quantities, the Bayesian version places hyperprior distributions on those group means so that information flows freely across levels, producing more reliable group-level estimates whenever any individual group has few observations.