A case study of regression analysis based on modeling Gilliland's correlation was described for use in a computational methods course. The case study uses a familiar example to train students in nonlinear least squares regression and to use standardized residual plots for model assessment. Previously published equations for Gilliland's correlation…

Percentage of body fat, age, weight, height, and 14 circumference measurements (e.g., waist) are given for 184 women aged 18-25. Body fat, one measure of health, was accurately determined by an underwater weighing technique which requires special equipment and training of the individuals conducting the process. Modeling body fat percentage using…

Although researchers have documented that some data make larger contributions than others to predictions made with least squares models, it is relatively unknown that some data actually make no contribution to the predictions produced by these models. This article explores such noncontributory data. (Contains 1 table and 2 figures.)

Design-based methods have recently been developed as a way to analyze data from impact evaluations of interventions, programs, and policies. The impact estimators are derived using the building blocks of experimental designs with minimal assumptions, and have good statistical properties. The methods apply to randomized controlled trials (RCTs) and…

Individuals vary in their responses to incentives and opportunities. For example, additional education will affect one person differently than another. In recent years, econometricians have given increased attention to such heterogeneous responses and to the consequences of such responses for interpreting regression estimates, especially…

One econometric rule of thumb is that greater dispersion in observations of the independent variable improves estimates of regression coefficients and therefore produces better results, i.e., lower standard errors of the estimates. Nevertheless, students often seem to mistrust precisely the observations that contribute the most to this greater…

University rankings are key drivers in national and institutional strategic planning. The increase in the number of university ranking systems and the diversity of methods and indicators used by these systems necessitate the development of an index that can measure a university's performance in all these systems at once. This article presents…

Business students taking business analytics courses that have significant predictive modeling components, such as marketing research, data mining, forecasting, and advanced financial modeling, are introduced to nonlinear regression using application software that is a "black box" to the students. Thus, although correct models are…

Measurement error significantly biases interaction effects and distorts researchers' inferences regarding interactive hypotheses. This article focuses on the single-indicator case and shows how to accurately estimate group slope differences by disattenuating interaction effects with errors-in-variables (EIV) regression. New analytic findings were…

A well-known approach to linear least squares regression is that which involves minimizing the sum of squared orthogonal projections of data points onto the best fit line. This form of regression is known as orthogonal regression, and the linear model that it yields is known as the major axis. A similar method, reduced major axis regression, is…

The authors propose a general modeling framework called the general monotone model (GeMM), which allows one to model psychological phenomena that manifest as nonlinear relations in behavior data without the need for making (overly) precise assumptions about functional form. Using both simulated and real data, the authors illustrate that GeMM…

Multilevel modeling provides several advantages over traditional ordinary least squares regression analysis; however, reporting results to stakeholders can be challenging. This article suggests some strategies for presenting complex, multilevel data and statistical results to institutional and higher education decision makers. The article is…

The dependent variable in a regular linear regression is a numerical variable, and in a logistic regression it is a binary or categorical variable. In these models the dependent variable has varying values. However, there are problems yielding an identity output of a constant value which can also be modelled in a linear or logistic regression with…

Several data-analysis problems could be addressed in different ways, ranging from a series of related "local" fitting problems to a single comprehensive "global analysis". The approach has become a powerful one for fitting data to moderately complex models by using library functions and the methods are illustrated for the analysis of HCI-IR…

We posit that institutions of higher education attempt to maximize their reputation, and that an institution's reputation, research output, and average student quality are determined simultaneously. Because these outputs are produced jointly, three-stage least squares is used to estimate the parameters of the model. We find that faculty research…