Various classical approaches to evaluating test bias are reviewed, and their conceptual underpinnings are discussed. Bias is examined as it relates to predictive validity, rather than construct validity. One of the first test bias models was developed by T. A. Cleary (1968). Cleary's work with test bias and regression stimulated others to develop…

This thesis addresses problems from two areas of theoretical computer science. The first area is that of computational learning theory, which is the study of the phenomenon of concept learning using formal mathematical models. The goal of computational learning theory is to investigate learning in a rigorous manner through the use of techniques…

The normal curve has long been important in statistics. Most interval variables yield normal or quasi-normal distributions when data are collected from large samples, and the normal "Z" distribution is also used as a test statistic (e.g., to test differences between two means when sample size is large, since "t" approaches…

New multivariate permutation tests are proposed that may be effectively substituted for Hotelling's T-Square test in situations commonly arising in educational research. The new tests: (1) are distribution-free; (2) provide tests of directional as well as non-directional hypotheses; (3) may be tailored for sensitivity to specific treatment…

Test items are designed to be representative of the subject areas that they measure and to reflect the importance of specific domains or item types within those subject areas. Content validity is achieved by content specification and number of items in each content domain included in the design of the test. However, largely due to the normal…

A study compared a parallel distributed processing (PDP) model with a more traditional symbolic information processing model that accounts for early reading acquisition by human subjects. Two experimental paradigms were simulated. In one paradigm (a "savings" paradigm) subjects were divided into two groups and trained with two sets of…

In 1992 N. S. Raju and others proposed a general procedure for assessing differential test functioning (DTF) and item bias (differential item functioning or DIF) in tests developed with unidimensional, multidimensional, or polytomous item response theory (IRT) models. The purpose of this paper is to assess the adequacy and validity of their…

This report details the data requirements and procedures used to develop new school bus routes for Six Mile Elementary School in South Carolina. The project examined the current routes of the school and applied computer based techniques to develop new routes given the existing bus stops. Bus routes were developed so that distance and travel time…

Designing a test using three-parameter item response theory (IRT) is discussed. A brief review of IRT is followed by a discussion of two types of test design: (1) selecting items using confidence envelopes (confidence envelope method); and (2) using item characteristic curves and their confidence intervals (test envelope method). The confidence…

This paper examines the effect of using unidimensional item response theory (IRT) item parameter estimates of multidimensional items to create weakly parallel test forms using target information curves. To date, all computer-based algorithms that have been devised to create parallel test forms assume that the items are unidimensional. This paper…

To evaluate test speededness and to derive implications for test program activity, this study reviewed the literature on speed and power, identifying four major approaches to the assessment of speed: the Gulliksen approach, the Cronbach and Warrington approach, the Stafford approach, and the approach of the Educational Testing Service (ETS) as…

A study was conducted to evaluate six goodness-of-fit procedures using data simulation techniques. The procedures evaluated included the weighted average absolute deviation (WAAD), the mean absolute deviation (MAD), Bock's chi-square (BCHI), Yen's chi-square (YCHI), Wright and Mead's chi-square (WCHI), and the likelihood ratio chi-square (LCHI)…

Based on sample data representing five years of monthly circulation totals from 50 academic libraries in Illinois, Iowa, Michigan, Minnesota, Missouri, and Ohio, a study was conducted to determine the most efficient smoothing forecasting methods for academic libraries. Smoothing forecasting methods were chosen because they have been characterized…

The overidentification of structural equation models with latent variables is discussed. The use of two- and three-indicator models is not recommended since such models do not allow a testing of the crucial assumption of unidimensionality among indicators in most cases. Models with four or more indicators may be more sensitive to departures from…

The computational methodology illustrated in this chapter provides higher educational planners at the federal, state, and institutional level with several alternative procedures for calculating productivity measures. Illustrations show that better institutional productivity measures can be constructed from existing data with a multidimensional…