Suppose an education researcher wants to test the impact of a high school drop-out prevention intervention in which at-risk students attend classes to receive intensive summer school instruction. The district will allow the researcher to randomly assign students to the treatment classes or to the control group. Half of the students (the treatment…

The U.S. Department of Education (ED) expects all Teacher Incentive Fund (TIF) grantees to conduct an evaluation of their programs. Experience with earlier rounds of TIF grants has shown that evaluations can provide valuable information for managing and improving TIF-supported activities, as well as evidence that these activities have had a…

The Florida Center for Reading Research (FCRR) Reading Assessment (FRA) consists of computer-adaptive reading comprehension and oral language screening tasks that provide measures to track growth over time, as well as a Probability of Literacy Success (PLS) linked to grade-level performance (i.e., the 50th percentile) on the reading comprehension…

One of the most revolutionary advances in psychometric research during the last decades has been the systematic development of statistical models that allow for cognitive psychometric research (CPR) to be conducted. Many of the models currently available for such purposes are extensions of basic latent variable models in item response theory…

Illustrates the basic concepts of the loglinear approach to research analysis. Explains what type of research questions are answerable and what type of data will lend itself to a loglinear analysis approach. Three educational research examples illustrate this concept. (TLJ)

Presents an approach to document numbering, document titling, and process measurement which, when used with fundamental techniques of statistical process control, reveals meaningful process-element variation as well as nominal productivity models. (SR)

Describes a statistical model that can be used to compute an institution's expected enrollment yield and an average discount to obtain a net tuition and fees forecast for the coming year; using this information, resource allocation decisions can be made earlier and with confidence. (EV)

The paper reviews a series of models designed to reduce bias in the assessment process and describes the intent of each model and their interactive effect in the process of nonbiased educational evaluation. (Author/CL)

The ultimate goal of the institutional researcher is not always to test a research hypothesis, but more often simply to find an appropriate model to gain an understanding of the underlying characteristics and interrelationships of the data. Exploratory data analysis provides a means of accomplishing this. (Author)

This article describes soft modeling with partial least squares in a nontechnical manner and suggests its greater use in special education research. This approach allows model building and evaluation in situations with high complexity but without well-articulated theories. (DB)

A discussion of multifacet Rasch model analysis describes the Rasch model and its assumptions, then presents an extension of the model to include a facet for the influence of examiner severity. The model is illustrated with an application to an oral examination administered by a medical specialty board. (Author/MSE)

Statistical modeling provides a method for colleges to use alumni-development information to create an equation predicting who is likely to respond positively to fund-raising appeals. This can make fund raising more cost-effective, provide higher returns on minimal investment, provide results quickly, and improve competitiveness. Resources…

This paper presents five designs (three one-group and two two-group) which can be effectively employed to study computer-based methods in non-laboratory settings, including actual classroom implementation, where a non-treatment control design may not be feasible. These designs are sequential analysis, value-added analysis, non-equivalent dependent…

Using data describing entry into preschool among children with special needs, this paper introduces the fundamental statistical quantities of event occurrence--hazard and survivor functions. The paper discusses construction of formal statistical models of event occurrence over time, predictors that can be included in hazard models, and importance…

The data collected in higher education research are not always quantitative or continuous. Statistical methods using the log-linear model provide the institutional researcher with a powerful set of tools for addressing research questions when data are categorical. (Author/MSE)