This study utilizes an analysis of diminishing returns to examine the relationship between sample size and item parameter estimation precision when utilizing the Masters' Partial Credit Model for polytomous items. Item data from the standardization of the Batelle Developmental Inventory, 3rd Edition were used. Each item was scored with a…

Educational assessment data are often collected from a set of test centers across various geographic regions, and therefore the data samples contain clusters. Such cluster-based data may result in clustering effects in variance estimation. However, in many grouped jackknife variance estimation applications, jackknife groups are often formed by a…

Knowledge tracing is a fundamental area of educational data modeling that aims at gaining a better understanding of the learning occurring in tutoring systems. Knowledge tracing models fit various parameters on observed student performance and are evaluated through several goodness of fit metrics. Fitted parameter values are of crucial interest in…

Multiple-baseline studies provide meta-analysts the opportunity to compute effect sizes based on either within-series comparisons of treatment phase to baseline phase observations, or time specific between-series comparisons of observations from those that have started treatment to observations of those that are still in baseline. The advantage of…

This study examines the relationship between sample size and item parameter estimation precision when utilizing the one-parameter model. Item parameter estimates are examined relative to "true" values by evaluating the decline in root mean squared deviation (RMSD) and the number of outliers as sample size increases. This occurs across…

Location estimates calculated from heuristic data were examined using traditional and robust statistical methods. The current paper demonstrates the impact outliers have on the sample mean and proposes robust methods to control for outliers in sample data. Traditional methods fail because they rely on the statistical assumptions of normality and…

Nonparametric, or kernel, estimation of item response curve (IRC) is a concern theoretically and operationally. Accuracy of this estimation, often used in item analysis in testing programs, is biased when the observed scores are used as the regressor because the observed scores are contaminated by measurement error. In this study, we investigate…

The focus of this presidential address is on the contemporary treatment of reliability and validity in educational assessment. Highlights on reliability are provided under the classical true-score model using tools from latent trait modeling to clarify important assumptions and procedures for reliability estimation. In addition to reliability,…

An extensive review of the research literature and a study comparing over 36,000 survey responses with archival true scores indicated that one should expect a minimum of at least three percent random error for the least ambiguous of self-report measures. The Gulliver Effect occurs when a small proportion of error in a sizable subpopulation exerts…

The present study compared item and ability invariance as well as model-data fit between the one-parameter (1PL) and three-parameter (3PL) Item Response Theory (IRT) models utilizing real data across five grades; second through sixth as well as simulated data at second, fourth and sixth grade. At each grade, the 1PL and 3PL IRT models were run…