Educational systems typically contain a large pool of items (questions, problems). Using data mining techniques we can group these items into knowledge components, detect duplicated items and outliers, and identify missing items. To these ends, it is useful to analyze item similarities, which can be used as input to clustering or visualization…

Modeling student knowledge is a fundamental task of an intelligent tutoring system. A popular approach for modeling the acquisition of knowledge is Bayesian Knowledge Tracing (BKT). Various extensions to the original BKT model have been proposed, among them two novel models that unify BKT and Item Response Theory (IRT). Latent Factor Knowledge…

The rise of digital game and simulation-based learning applications has led to new approaches in educational measurement that take account of patterns in time, high resolution paths of action, and clusters of virtual performance artifacts. The new approaches, which depart from traditional statistical analyses, include data mining, machine…

When validating assessment models built with data mining, generalization is typically tested at the student-level, where models are tested on new students. This approach, though, may fail to find cases where model performance suffers if other aspects of those cases relevant to prediction are not well represented. We explore this here by testing if…

A fundamental goal of educational research is identifying students' current stage of skill mastery (complete/partial/none). In recent years a number of cognitive diagnosis models have become a popular means of estimating student skill knowledge. However, these models become difficult to estimate as the number of students, items, and skills grows.…

Researchers who make tutoring systems would like to know which sequences of educational content lead to the most effective learning by their students. The majority of data collected in many ITS systems consist of answers to a group of questions of a given skill often presented in a random sequence. Following work that identifies which items…

SNAP/SHOT (System Network Analysis Program-Simulated Host Overview Technique) is a discrete simulation of a network and/or host model available through IBM at the Raleigh System Center. The simulator provides an analysis of a total IBM Communications System. Input data must be obtained from RMF, SMF, and the CICS Analyzer to determine the existing…

Describes an interactive software system (implemented in IBM interactive FORTRAN) which carries out data analyses, parameter estimation, and process generation. Includes discussion of the four system modules (data preparation; selecting process for goodness-of-fit testing; goodness-of-fit calculations/reports; process generation) and educational…

A 2-to-the-kth-power mirror-image design is defined as one in which repeated observations of subjects occurs among the levels of a usual 2-to-the-kth-power design, but there is the restriction that no subject may receive a given level of any factor more than once. Such a restriction might arise, for example, if a subject's response is expected to…

Various methods have been suggested for the analysis of data collected in research settings where random assignment of subjects to groups has not occurred. For the purposes of this paper the set of allowable nonrandomized designs is made up of those research designs where data are collected for one or more groups of subjects at two or more time…

A way to support the educational ethnographer in developing a perspective on the art of qualitative research during an introductory course on qualitative research methods is explored through a study of how novice researchers begin to learn the elements and processes of qualitative research. A second purpose of the study is to investigate the use…

Multiple group factor analysis is described and illustrated through a simulation involving 5,000 examinees. The estimation process of the group factors were implemented using the TESTFACT program of Wilson and others (1987). Group factor analysis is described as a special case of confirmatory factor analysis. Group factors can be computed based on…

The purpose of this research is to demonstrate that a systematic approach to the graphical analysis of Rasch model residuals can lead to an increased understanding of ordered response data, and that residual patterns do change in predictable ways, and that summary statistics need not be the only piece of evidence for assuring the fit between model…

Real test data of unknown structure were analyzed using both a unidimensional and a multidimensional latent trait model in an attempt to determine the underlying components of the test. The models used were the three-parameter logistic model and a multidimensional extension of the two-parameter logistic model. The basic design for the analysis of…

This paper describes the first phase of a study to investigate students' evaluations of computer laboratory modules in a university-level, non-majors biology course. The National Science Foundation-funded project has two primary goals: (1) to develop programmable, multifunctional Bio LabStations for data collection and analysis, lab extensions,…