The paper provides a survey of 18 years' progress that my colleagues, students (both former and current) and I made in a prominent research area in Psychometrics-Computerized Adaptive Testing (CAT). We start with a historical review of the establishment of a large sample foundation for CAT. It is worth noting that the asymptotic results were derived under the framework of Martingale Theory, a very theoretical perspective of Probability Theory, which may seem unrelated to educational and psychological testing. In addition, we address a number of issues that emerged from large scale implementation and show that how theoretical works can be helpful to solve the problems. Finally, we propose that CAT technology can be very useful to support individualized instruction on a mass scale. We show that even paper and pencil based tests can be made adaptive to support classroom teaching.
Recently, Shealy and Stout (1993) proposed a DIF detecting procedure SIBTEST, which is 1) IRT model based, 2) non‐parametric, 3) does not require IRF estimation, 4) provides a test of significance, and 5) estimates the amount of DIF. Current versions of SIBTEST can only be used for dichotomously scored items. However, in this paper an extension to handle polytomous items is developed. This paper presents: (1) a discussion of an appropriate definition of DIF for polytomously scored items, (2) a modified SIBTEST procedure for detecting DIF for polytomous items, and (3) the results of two simulation studies comparing the modified SIBTEST with the Mantel and SMD procedures, one study with data constrained by the Rasch‐like partial credit model (same discrimination across polytomous items), and the other study with data having distinctly discrimations across items. These simulation studies indicate that the methodology of including the studied item in matching subtest for controling impact induced (group ability differences existing) Type I error tends to yield Type‐I/Type II error inflation rates that are highly unacceptable when the equal discrimination condition is violated. These simulation studies provide compelling evidence that the modified SIBTEST procedure is much more robust with regard to controlling impact‐induced Type I error rate inflation than the other procedures.
A recent ''third wave'' of neural network (NN) approaches now delivers state-ofthe-art performance in many machine learning tasks, spanning speech recognition, computer vision, and natural language processing. Because these modern NNs often comprise multiple interconnected layers, work in this area is often referred to as deep learning. Recent years have witnessed an explosive growth of research into NN-based approaches to information retrieval (IR). A significant body of work has now been created. In this paper,
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.