Measurement Methods for the Social Science – serie

"This manuscript addresses an important and complex topic in test development in a manner that is precise and accurate, yet very accessible to students and practitioners with a modest background in classical test theory and item response theory. It also provides an excellent introduction to nonparametric IRT models for the more mathematically sophisticated student or faculty member who will welcome the extensive additional reading lists that are found at the conclusion of each chapter."

—LINDA F. WIGHTMAN, School of Education, University of N. Carolina, Greensboro

"I thoroughly enjoyed this book, and liked the clear way the authors have worked through the chapters and examples. There are rich examples with plenty of exercises that encouraged me to try these methods with my own data. The quality of the interpretation is rich, particularly in the polytomous item domain. It is well worth having on the shelf as a reference tool and as available for graduate students who wish to know more."

—JOHN HATTIE, Head of the School of Education, University of Auckland, NZ

This book introduces social and behavioral science students and researchers to the theory and practice of the highly powerful methods of nonparametric item response theory (IRT). Anyone who uses or constructs tests or questionnaires for measuring abilities, achievements, personality traits, attitudes, or opinions will find nonparametric IRT useful for designing and improving such measurements. The authors show how the broadness of the nonparametric item response models allows them to fit many data sets and remain powerful enough for implying useful measurement properties, such as the ordering of persons using the simple total score (number-correct for dichotomous item tests and sum of rating scale score for polytomous item tests) and the ordering of the items using the item means. Many data analysis examples are given in the book, and a user-friendly computer program used throughout the book supports data analysis using nonparametric IRT. Given the importance of school admissions, certification, personnel selection, marketing, social policy evaluation, quality-of-life measurements, and assessments of deviant behavior, this book is a must read for students or researchers engaged in this work.

"This manuscript addresses an important and complex topic in test development in a manner that is precise and accurate, yet very accessible to students and practitioners with a modest background in classical test theory and item response theory. It also provides an excellent introduction to nonparametric IRT models for the more mathematically sophisticated student or faculty member who will welcome the extensive additional reading lists that are found at the conclusion of each chapter."

—LINDA F. WIGHTMAN, School of Education, University of N. Carolina, Greensboro

"I thoroughly enjoyed this book, and liked the clear way the authors have worked through the chapters and examples. There are rich examples with plenty of exercises that encouraged me to try these methods with my own data. The quality of the interpretation is rich, particularly in the polytomous item domain. It is well worth having on the shelf as a reference tool and as available for graduate students who wish to know more."

—JOHN HATTIE, Head of the School of Education, University of Auckland, NZ

This book introduces social and behavioral science students and researchers to the theory and practice of the highly powerful methods of nonparametric item response theory (IRT). Anyone who uses or constructs tests or questionnaires for measuring abilities, achievements, personality traits, attitudes, or opinions will find nonparametric IRT useful for designing and improving such measurements. The authors show how the broadness of the nonparametric item response models allows them to fit many data sets and remain powerful enough for implying useful measurement properties, such as the ordering of persons using the simple total score (number-correct for dichotomous item tests and sum of rating scale score for polytomous item tests) and the ordering of the items using the item means. Many data analysis examples are given in the book, and a user-friendly computer program used throughout the book supports data analysis using nonparametric IRT. Given the importance of school admissions, certification, personnel selection, marketing, social policy evaluation, quality-of-life measurements, and assessments of deviant behavior, this book is a must read for students or researchers engaged in this work.

Accessible to any professional or researcher who has a basic understanding of analysis of variance, Shavelson and Webb offer an intuitive development of generalizability theory, a technique for estimating the relative magnitudes of various components of error variation and for indicating the most efficient strategy for achieving desired measurement precision. Covering a variety of topics such as generalizability studies with nested facets and with fixed facets, measurement error and generalizability coefficients, and decision studies with same and with different designs, the text includes exercises so the reader may practice the application of each chapter's material. By using detailed illustrations and examples, Shavelson and Webb clearly describe the logic underlying major concepts in generalizability theory to enable readers to apply these methods when investigating the consistency of their own measurements.