Contemporary Psychometrics: A Festschrift for Roderick P. McDonald

Contemporary Psychometrics: A Festschrift for Roderick P. McDonald

Contemporary Psychometrics: A Festschrift for Roderick P. McDonald

Contemporary Psychometrics: A Festschrift for Roderick P. McDonald

Synopsis

Contemporary Psychometricsfeatures cutting edge chapters organized in four sections: test theory, factor analysis, structural equation modeling, and multivariate analysis. The section on test theory includes topics such as multidimensional item response theory (IRT), the relationship between IRT and factor analysis, estimation and testing of these models, and basic measurement issues that are often neglected. The factor analysis section reviews the history and development of the model, factorial invariance and factor analysis indeterminacy, and Bayesian inference for factor scores and parameter estimates. The section on structural equation modeling (SEM) includes the general algebraic-graphic rules for latent variable SEM, a survey of goodness of fit assessment, SEM resampling methods, a discussion of how to compare correlations between and within independent samples, dynamic factor models based on ARMA time series models, and multi-level factor analysis models for continuous and discrete data. The final section on multivariate analysis includes topics such as dual scaling of ordinal data, model specification and missing data problems in time series models, and a discussion of the themes that run through all multivariate methods. Thistour de forcethrough contemporary psychometrics will appeal to advanced students and researchers in the social and behavioral sciences and education, as well as methodologists from other disciplines.

Excerpt

With the approach of Roderick P. McDonald’s retirement, a group of former students, colleagues, and friends gathered to honor him celebrating his contributions to psychometrics in the form of this Festschrift volume and a reunion. As Rod had been elected president of the Society of Multivariate Experimental Psychology in 2002, we used this occasion as the date and location of our celebration. “Advances in Psychometrics. a Day to Honor Roderick P. McDonald” took place at the Rotunda of the University of Virginia, Charlottsville, on October 20, 2002. Many of us gathered at a day-long meeting to praise him and present the chapters that now make up this volume. Some of us, like his long-time friend and colleague Colin Fraser, showed up as a surprise.

In many cases, these talks were impromptu, but in all cases, they were inspired. It was clear that each person had a slightly different story to tell, some as colleagues and some as students. However, a common theme emerged clearly—Rod McDonald personally affected both our work and our life. For all of us, this became an inspiring occasion, and, as a collective, we realized we have been trying to follow the strong principles advocated by Rod since we first encountered him. To Rod, it was a very emotion-filled occasion, and in his presentation he cited from classical verse. “I feel,” he said, “like an ancient Roman being driven in a chariot for his triumph, and need someone standing behind me whispering, ‘Remember, Caesar, that you are mortal.’”

This book is intended as a celebration of the seminal work of Rod McDonald over a career that has spanned more than 40 years and three continents. This work has covered a wide range of topics in psychometrics, and has inspired us in different areas of the field. the chapters included here cover most of the areas Rod has worked on over these years. the chapters are organized in four diverse sections: test theory, factor analysis, structural equation modeling, and multivariate analysis. Yet, a common theme can be found in Rod McDonald’s work—the search for a unified framework of psychometric theory. We hope that this volume makes a small contribution toward this goal.

The first section of the book, test theory, includes topics such as multidimensional item response theory, the relationship between item response theory and factor analysis, estimation and testing of these models (limited information vs. full information), and a reflection on basic measurement issues that are often . . .

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.