Applying the Rasch Model: Fundamental Measurement in the Human Sciences / Edition 4

by Trevor Bond, Zi Yan, Moritz Heene

Overview

Recognised as the most influential publication in the field, ARM facilitates deep understanding of the Rasch model and its practical applications. The authors review the crucial properties of the model and demonstrate its use with examples across the human sciences. Readers will be able to understand and critically evaluate Rasch measurement research, perform their own Rasch analyses and interpret their results. The glossary and illustrations support that understanding, and the accessible approach means that it is ideal for readers without a mathematical background.

Highlights of the new edition include:

• More learning tools to strengthen readers’ understanding including chapter introductions, boldfaced key terms, chapter summaries, activities and suggested readings.
• Greater emphasis on the use of R packages; readers can download the R code from the Routledge website.
• Explores the distinction between numerical values, quantity and units, to understand the measurement and the role of the Rasch logit scale (Chapter 4).
• A new four-option data set from the IASQ (Instrumental Attitude towards Self-assessment Questionnaire) for the Rating Scale Model (RSM) analysis exemplar (Chapter 6).
• Clarifies the relationship between Rasch measurement, path analysis and SEM, with a host of new examples of Rasch measurement applied across health sciences, education and psychology (Chapter 10).

Intended as a text for graduate courses in measurement, item response theory, (advanced) research methods or quantitative analysis taught in psychology, education, human development, business, and other social and health sciences. Professionals in these areas will also appreciate the book’s accessible introduction.

Product Details

ISBN-13:	9780367141424
Publisher:	Taylor & Francis
Publication date:	07/20/2020
Edition description:	4th ed.
Pages:	376
Product dimensions:	7.00(w) x 10.00(h) x (d)

About the Author

Trevor G. Bond is currently Adjunct Professor at the College of Arts, Society and Education at James Cook University, Australia.

Zi Yan is Associate Professor in the Department of Curriculum and Instruction at the Education University of Hong Kong.

Moritz Heene is Full Professor of Learning Sciences Research Methodologies (i.e., Quantitative Methods) at the Ludwig-Maximilians-Universität München, Germany.

Table of Contents

List of Figures xii

List of Tables xv

About the Authors xviii

Foreword xx

Preface xxii

Notes on This Volume xxv

Acknowledgments xxviii

1 Why Measurement Is Fundamental 1

Children Can Construct Measures 3

Interval Scales v. Ratio Scales: A Conceptual Explanation 5

Statistics and/or Measurement 6

Why Fundamental Measurement? 7

Derived Measures 7

Conjoint Measurement 9

The Rasch Model for Measurement 11

A More Suitable Analogy for Measurement in the Human Sciences 12

In Conclusion 14

Summary 15

2 Important Principles of Measurement Made Explicit 18

An Example: "By How Much?" 21

Moving from Observations to Measures 26

Summary 28

3 Basic Principles of the Rasch Model 31

The Pathway Analogy 31

A Basic Framework for Measurement 41

The Rasch Model 43

Summary 47

4 Building a Set of Items for Measurement 50

The Nature of the Data 50

Analyzing Dichotomous Data: The BLOT 51

A Simple Rasch Summary: The Item Pathway 52

Item Statistics 54

Item Fit 54

The Wright Map 56

Targeting 58

Comparing Persons and Items 59

Summary 60

Extended Understanding 62

The Problem of Guessing 63

Difficulty, Ability, and Fit 64

The Theory-Practice Dialog 66

Summary 67

5 Invariance: A Crucial Property of Scientific Measurement 69

Person and Item Invariance 72

Common-Item Linking 72

Please Keep in Mind 74

Anchoring Item Values 74

Vertical Scaling 77

Common-Person Linking 78

Invariance of Person Estimates across Tests: Concurrent Validity 80

The PRTIII-Pendulum 81

Common-Person Linking: BLOT & PRTIII 82

The Theory-Practice Dialog 87

Measurement Invariance: Where It Realty Matters 88

Failures of Invariance: DIF 89

Differential Rater Functioning 91

DIF: Not Just a Problem, but an Opportunity 92

Summary 92

6 Measurement Using Likert Scales 96

The Rasch Model for Polytomous Data 97

Analyzing Rating Scale Data: The Instrumental Attitude toward Self-Assessment Questionnaire 100

Summary 105

Extended Understanding 107

Summary 120

7 The Partial Credit Rasch Model 124

Clinical Interview Analysis: A Rasch-Inspired Breakthrough 128

Scoring Interview Transcripts 129

Partial Credit Model Results 132

Interpretation 134

The Theory-Practice Dialog 137

Unidimensionality 137

Summary 138

Extended Understanding 139

Point-Measure Correlations 141

Fit Statistics 142

Dimensionality: Primary Components Factor Analysis of the Rasch Residuals 142

Summary 142

8 Measuring Facets Beyond Ability and Difficulty 145

A Basic Introduction to the Many-Facets Rasch Model 146

Why Not Use Interrater Reliability? 147

Relations among the Rasch Family of Models 148

Data Specifications of the Many-Facets Rasch Model 149

Rating Creativity of Junior Scientists 150

Many-Facets Analysis of Eighth-Grade Writing 152

Summary 158

Extended Understanding 158

Rasch Measurement of Facets Beyond Rater Effects 159

Summary 160

9 Making Measures, Setting Standards, and Rasch Regression 163

Creating a Measure from Existing Data: The RMPFS (Zi Yan, EdUHK) 163

Method: Data 163

Physical Fitness Indicators 164

Data Analysis 164

Seven Criteria to Investigate the Quality of Physical Fitness Indicators 165

Results and Discussion 165

Optimising Response Categories 167

Influence of Underfilling Persons on the RMPFS 167

Properties of the RMPFS with Subsamples 168

Age Dependent or Age Related? 168

The Final Version of RMPFS 168

Objective Standard Setting: The OSS Model (Gregory Stone, U Toledo) 171

Early Definitions 174

Tlie Objective Standard Setting Models 175

Objective Standard Setting for Dichotomous Examinations 175

Objective Standard Setting for Judge-Mediated Examinations 179

Fair Standards, Not Absolute Values 181

Rasch Regression (Svetlana Beltyukova, U Toledo) 182

Predicting Physician Assistant Faculty Intention to Leave Academia 182

Rasch Regression Using the Anchored Formulation 183

Rasch Regression: Alternative Approaches 188

Discussion 189

Summary 190

10 The Rasch Model Applied across the Human Sciences 193

Rasch Measurement in Health Sciences 193

Optimising an Existing Instrument: The NIHSS and a Central Role for PCA 196

Creating a Short Form of an Existing Instrument: The FSQ 197

FSQ-SF 198

Theory Guides Assessment Revisions: The PEP-S8 198

Applications in Education and Psychology 199

Rasch Measures as Grist for the Analytical Mill 201

Rasch Gain Calculations: Racking and Stacking 202

Rasch Learning Gain Calculations: The CCI 203

Racking and Stacking 203

Stacking Can Be Enough: UPAM 204

Sub-Test Structure Informs Scoring Models 205

Applications to Classroom Testing 206

Can Rasch Measurement Help S.S. Stevens? 212

Using Rasch Measures with Path Analysis (SEM Framework) 212

Rasch Person Measures Used in a Partial Least Squares (PLS) Framework 213

AndThose Rasch Measurement SEs? 215

Can We Really Combine SEM and Rasch Models? 216

Conclusion 217

Summary 218

11 Rasch Modeling Applied: Rating Scale Design 222

Rating Scale Design 222

Category Frequencies and Average Measures 224

Thresholds and Category Fit 225

Revising a Rating Scale 228

An Example 228

Guidelines for Collapsing Categories 229

Problems with Negatively Worded Items 232

The Invariance of the Measures across Groups 234

Summary 235

12 Rasch Model Requirements: Model Pit and Unidimensionality 238

Model Fit and Unidimensionality 238

The Data, the Model, and the Residuals 239

Residuals 240

Fit Statistics 241

Expectations of Variation 241

Fit, Misfit, and Interpretation 245

Fit: Issues for Resolution 251

Misfit: A Fundamental Issue 252

Principal Components Analysis of Rasch Residuals: The BLOT as an Exemplar 255

One Dimension, Two Dimensions, Three Dimensions, More? 259

Extended Understanding 261

A Further Investigation: BLOT and PRTIII 262

Summary 264

13 A Synthetic Overview 268

Additive Conjoint Measurement (ACM) 269

True Score Theory, Latent Traits, and Item Response Theory 273

Would You Like an Interval Scale with That? 277

Model Assumptions and Measurement Requirements 279

Construct Validity 281

The Rasch Model and Progress of Science 284

Back to the Beginning and Back to the End 285

Summary 288

Appendix A Getting Started 293

Appendix B Technical Aspects of the Rasch Model 308

Appendix C Going All the Way 318

Glossary 331

Author Index 344

Subject Index 345

Editorial Reviews

From a previous edition:

"The tiresome debate about Rasch vs. IRT is over — if you want to construct valid measurements from uncertain observations you need to understand and learn how to use Rasch measurement. Bond and Fox is your huckleberry — read it and get to work!" – Robert W. Massof, Johns Hopkins University School of Medicine, USA

"Bond & Fox's book is a must read for anyone interested in measurement. This book is my go-to for introducing graduate students to the Rasch model." – Kelly D. Bradley, University of Kentucky, USA

"The authors have successfully made sophisticated measurement theory into feasible practice for practitioners by providing clear and intuitive explanations, numerous examples, and nice computer outputs. It is a textbook that I have used and will continue to use in the future." – Wen Chung Wang, Hong Kong Institute of Education, Hong Kong

"Bond and Fox provide a clear and accessible introduction to the Rasch model… I always recommend their book when a student or colleagues asks me to explain Rasch measurement theory." – George Engelhard, Jr., The University of Georgia, USA

"The Rasch model represents modern measurement theory at its best … Rasch models are used around the world to create psychometrically defensible scales and tests. Bond and Fox provide an accessible introduction to the Rasch model that describes the logic and essential importance of fundamental measurement in the human sciences." – George Engelhard, Jr., The University of Georgia, USA

From a previous edition:

"The tiresome debate about Rasch vs. IRT is over – if you want to construct valid measurements from uncertain observations you need to understand and learn how to use Rasch measurement. Bond and Fox is your huckleberry – read it and get to work!" – Robert W. Massof, Johns Hopkins University School of Medicine, USA

"Bond & Fox's book is a must read for anyone interested in measurement. This book is my go-to for introducing graduate students to the Rasch model." – Kelly D. Bradley, University of Kentucky, USA

"The authors have successfully made sophisticated measurement theory into feasible practice for practitioners by providing clear and intuitive explanations, numerous examples, and nice computer outputs. It is a textbook that I have used and will continue to use in the future." – Wen Chung Wang, Hong Kong Institute of Education, Hong Kong

"The Rasch model represents modern measurement theory at its best … Rasch models are used around the world to create psychometrically defensible scales and tests. Bond and Fox provide an accessible introduction to the Rasch model that describes the logic and essential importance of fundamental measurement in the human sciences." – George Engelhard, Jr., The University of Georgia, USA

From the Publisher