Muutke küpsiste eelistusi

Theory and Practice of Item Response Theory, Second Edition [Kõva köide]

  • Formaat: Hardback, 643 pages, kõrgus x laius: 254x178 mm, kaal: 1320 g
  • Ilmumisaeg: 27-May-2022
  • Kirjastus: Guilford Press
  • ISBN-10: 1462547753
  • ISBN-13: 9781462547753
Teised raamatud teemal:
  • Formaat: Hardback, 643 pages, kõrgus x laius: 254x178 mm, kaal: 1320 g
  • Ilmumisaeg: 27-May-2022
  • Kirjastus: Guilford Press
  • ISBN-10: 1462547753
  • ISBN-13: 9781462547753
Teised raamatud teemal:
"Simple to more complex models are covered in consistently formatted chapters that build sequentially. The book takes the reader from model development through the fit analysis and interpretation phases that would be performed in practice. To facilitate understanding, common datasets are used across chapters, with the examples worked through for increasingly complex models. Exemplary model applications include free (BIGSTEPS, NOHARM, Facets, R packages) and commercial (BILOG-MG, flexMIRT, SAS, WINMIRA, SPSS, SYSTAT) software packages. The companion website provides data files and online-only appendices. New to This Edition: Chapter on multilevel models. New material on loglinear models, mixed models, the linear logistic trait model, and fit statistics. Many additional worked-through examples. Updated guidance on software; now includes R, SAS, and flexMIRT. Keywords: introduction to IRT, Rasch models, psychometrics, tests and measurement, intermediate, advanced psychometrics texts, modern measurement, latent variable analysis, graduate courses, classes, applied statistics, psychometricians, flexMIRT, testing"--

Noted for addressing both the "hows" and "whys" of item response theory (IRT), this text has been revised and updated with the latest techniques (multilevel models, mixed models, and more) and software packages. Simple to more complex models are covered in consistently formatted chapters that build sequentially. The book takes the reader from model development through the fit analysis and interpretation phases that would be performed in practice. To facilitate understanding, common datasets are used across chapters, with the examples worked through for increasingly complex models. Exemplary model applications include free (BIGSTEPS, NOHARM, Facets, R packages) and commercial (BILOG-MG, flexMIRT, SAS, WINMIRA, SPSS, SYSTAT) software packages. The companion website provides data files and online-only appendices.
 
 New to This Edition
*Chapter on multilevel models.
*New material on loglinear models, mixed models, the linear logistic trait model, and fit statistics.
*Many additional worked-through examples.
*Updated guidance on software; now includes R, SAS, and flexMIRT.
 

Arvustused

"The second edition of the IRT 'bible' is now even more accessible and useful for psychometricians and educational measurement specialists who are new to IRT or want to upgrade their knowledge of the field. It expands on the first edition in meaningful ways, primarily with regard to the implementation of IRT. Virtually every chapter has been expanded with examples of IRT analyses using R, SAS, and/or flexMIRT. A very helpful new chapter covers multilevel IRT models, and new appendices cover the LLTM and mixture Rasch models and discuss contemporary model fit indices and other recent topics. I have been using the first edition in my advanced measurement class since it was first published and it has been well received by my advanced undergraduates and graduate students; the new material in the second edition makes the book even better. In addition, this book will be very informative to measurement specialists--especially those who are not mathematically sophisticated--so that they can produce instruments that benefit from the enhanced measurement power of IRT.--David J. Weiss, PhD, Department of Psychology, University of Minnesota

"This is the most comprehensive and accessible text on IRT. De Ayala does a remarkable job of clearly describing fundamental IRT concepts, basic models, and even advanced models. The text's explanations do not heavily rely on equations; instead, de Ayala takes a conceptual approach and often utilizes graphs to illustrate key ideas. The second edition is up to date on the most frequently applied models and estimation procedures. It includes applied examples using popular IRT software, including R. I highly recommend this book for graduate-level courses focusing on measurement, psychometrics, and IRT, and as a guide for researchers using IRT."--Ojmarrh Mitchell, PhD, School of Criminology and Criminal Justice, Arizona State University

"I love this book, and find it quite readable. What sets this text apart are its extensive exposition of technical details related to models and estimation and its detailed explanations of concepts. For example, I had never seen an author decompose the partial credit model and show piece-by-piece computation of the probabilities, which de Ayala does very well. This text is a great contribution to the field of IRT that will be invaluable for both class and personal use."--Karen M. Schmidt, PhD, Department of Psychology, University of Virginia

"An excellent treatment of IRT that combines a clear exposition of theoretical concepts with applied examples that are relevant and useful."--Larry R. Price, PhD, College of Education, Texas State University-A must read for practitioners who use item response theory to calibrate test data. It also would serve as a tremendous resource for measurement researchers who daily navigate the circuitous paths of various IRT estimation software programs to analyze and understand their assessment data....Each of the 12 chapters is packed with annotated examples of how to use IRT estimation software and the subsequent output....The author does an excellent job of supplementing explanations of various models with calibration examples and output of multiple data sets using several different IRT calibration software programs including BILOG, MULTILOG, BIGSTEPS, and NOHARM....The book is more practitioner-oriented and applied than previous classic books that provide foundational understanding of IRT models and applications....Would be an excellent text for a graduate level IRT class in which the goal of the course would be to review dichotomous, polytomous, and multidimensional IRT models an how to estimate parameters in the various models using a variety of commercially available software....I would encourage all testing practitioners who work with various IRT models, as well as graduate students who plan to go into the measurement field, to seriously consider this book. It is an excellent resource.I applaud Dr. de Ayala for all the time and effort he has put into this book. He has clearly done the measurement field a great service. (on the first edition)--Journal of Educational Measurement, 12/21/2010The main strength of the text is in the descriptions and elaborations of the common IRT models....De Ayala also covers fundamental relationships that exist between models, such as the relationships between the parameters of the nominal response model and the partial credit model. In addition, the chapters contain practical advice for sample sizes commonly used with each model and how to interpret the parameters. De Ayala also presents results as statistical indices and graphics for various examples across different contexts, which allows readers the ability to see how the models work from several different perspectives....Does a good job of introducing common estimation strategies employed in IRT software packages. Especially helpful are the illustrations de Ayala includes with the code from IRT software packages. (on the first edition)--Psychometrika, 12/1/2010 "The second edition of the IRT 'bible' is now even more accessible and useful for psychometricians and educational measurement specialists who are new to IRT or want to upgrade their knowledge of the field. It expands on the first edition in meaningful ways, primarily with regard to the implementation of IRT. Virtually every chapter has been expanded with examples of IRT analyses using R, SAS, and/or flexMIRT. A very helpful new chapter covers multilevel IRT models, and new appendices cover the LLTM and mixture Rasch models and discuss contemporary model fit indices and other recent topics. I have been using the first edition in my advanced measurement class since it was first published and it has been well received by my advanced undergraduates and graduate students; the new material in the second edition makes the book even better. In addition, this book will be very informative to measurement specialists--especially those who are not mathematically sophisticated--so that they can produce instruments that benefit from the enhanced measurement power of IRT.--David J. Weiss, PhD, Department of Psychology, University of Minnesota

"This is the most comprehensive and accessible text on IRT. De Ayala does a remarkable job of clearly describing fundamental IRT concepts, basic models, and even advanced models. The text's explanations do not heavily rely on equations; instead, de Ayala takes a conceptual approach and often utilizes graphs to illustrate key ideas. The second edition is up to date on the most frequently applied models and estimation procedures. It includes applied examples using popular IRT software, including R. I highly recommend this book for graduate-level courses focusing on measurement, psychometrics, and IRT, and as a guide for researchers using IRT."--Ojmarrh Mitchell, PhD, School of Criminology and Criminal Justice, Arizona State University

"I love this book, and find it quite readable. What sets this text apart are its extensive exposition of technical details related to models and estimation and its detailed explanations of concepts. For example, I had never seen an author decompose the partial credit model and show piece-by-piece computation of the probabilities, which de Ayala does very well. This text is a great contribution to the field of IRT that will be invaluable for both class and personal use."--Karen M. Schmidt, PhD, Department of Psychology, University of Virginia

"An excellent treatment of IRT that combines a clear exposition of theoretical concepts with applied examples that are relevant and useful."--Larry R. Price, PhD, College of Education, Texas State University-A must read for practitioners who use item response theory to calibrate test data. It also would serve as a tremendous resource for measurement researchers who daily navigate the circuitous paths of various IRT estimation software programs to analyze and understand their assessment data....Each of the 12 chapters is packed with annotated examples of how to use IRT estimation software and the subsequent output....The author does an excellent job of supplementing explanations of various models with calibration examples and output of multiple data sets using several different IRT calibration software programs including BILOG, MULTILOG, BIGSTEPS, and NOHARM....The book is more practitioner-oriented and applied than previous classic books that provide foundational understanding of IRT models and applications....Would be an excellent text for a graduate level IRT class in which the goal of the course would be to review dichotomous, polytomous, and multidimensional IRT models an how to estimate parameters in the various models using a variety of commercially available software....I would encourage all testing practitioners who work with various IRT models, as well as graduate students who plan to go into the measurement field, to seriously consider this book. It is an excellent resourceâ¦.I applaud Dr. de Ayala for all the time and effort he has put into this book. He has clearly done the measurement field a great service. (on the first edition)--Journal of Educational Measurement, 12/21/2010ÆÆThe main strength of the text is in the descriptions and elaborations of the common IRT models....De Ayala also covers fundamental relationships that exist between models, such as the relationships between the parameters of the nominal response model and the partial credit model. In addition, the chapters contain practical advice for sample sizes commonly used with each model and how to interpret the parameters. De Ayala also presents results as statistical indices and graphics for various examples across different contexts, which allows readers the ability to see how the models work from several different perspectives....Does a good job of introducing common estimation strategies employed in IRT software packages. Especially helpful are the illustrations de Ayala includes with the code from IRT software packages. (on the first edition)--Psychometrika, 12/1/2010

Symbols and Acronyms xxi
1 Introduction to Measurement
1(11)
Measurement
1(2)
Some Measurement Issues
3(2)
Item Response Theory
5(1)
Classical Test Theory
5(3)
Latent Class Analysis
8(1)
Summary
9(3)
2 The One-Parameter Model
12(30)
Conceptual Development of the Rasch Model
12(5)
The One-Parameter Model
17(3)
The One-Parameter Logistic Model and the Rasch Model
20(1)
Assumptions Underlying the Model
21(2)
An Empirical Data Set: The Mathematics Data Set
23(1)
Conceptually Estimating an Individual's Location
23(5)
Some Pragmatic Characteristics of Maximum Likelihood Estimates
28(1)
The Standard Error of Estimate and Information
29(3)
An Instrument's Estimation Capacity
32(3)
Summary
35(7)
3 Joint Maximum Likelihood Parameter Estimation
42(44)
Joint Maximum Likelihood Estimation
42(2)
Indeterminacy of Parameter Estimates
44(1)
How Large a Calibration Sample?
45(1)
Example: Application of the Rasch Model to the Mathematics Data, JMLE, BIGSTEPS
46(22)
Example: Application of the Rasch Model to the Mathematics Data, JMLE, mixRasch
68(7)
Validity Evidence
75(1)
Summary of the Application of the Rasch Model
76(1)
Summary
77(9)
4 Marginal Maximum Likelihood Parameter Estimation
86(49)
Marginal Maximum Likelihood Estimation
86(7)
Estimating an Individual's Location: Expected A Posteriori
93(5)
Example: Application of the Rasch Model to the Mathematics Data, MMLE, BILOG-MG
98(13)
Metric Transformation and the Total Characteristic Function
111(4)
Example: Application of the Rasch Model to the Mathematics Data, MMLE, mirt
115(10)
Summary
125(10)
5 The Two-Parameter Model
135(44)
Conceptual Development of the Two-Parameter Model
135(2)
Information for the Two-Parameter Model
137(2)
Conceptual Parameter Estimation for the 2PL Model
139(1)
How Large a Calibration Sample?
140(2)
Metric Transformation, 2PL Model
142(1)
Example: Application of the 2PL Model to the Mathematics Data, MMLE, BILOG-MG I
143(3)
Fit Assessment: An Alternative Approach for Assessing Invariance
146(6)
Example: Application of the 2PL Model to the Mathematics Data, MMLE, mirt
152(10)
Information and Relative Efficiency
162(3)
Summary
165(14)
6 The Three-Parameter Model
179(58)
Conceptual Development of the Three-Parameter Model
179(3)
Additional Comments about the Pseudo-Guessing Parameter, %j I
182(1)
Conceptual Parameter Estimation for the 3PL Model
183(4)
How Large a Calibration Sample?
187(1)
Assessing Conditional Independence
188(4)
Example: Application of the 3PL Model to the Mathematics Data, MMLE, BILOG-MG I
192(3)
Fit Assessment: Conditional Independence Assessment
195(3)
Fit Assessment: Model Comparison
198(2)
Example: Application of the 3PL Model to the Mathematics Data, MMLE, mirt
200(9)
Assessing Person Fit: Appropriateness Measurement
209(7)
Information for the Three-Parameter Model
216(4)
Metric Transformation, 3PL Model
220(1)
Handling Missing Responses
220(4)
Issues to Consider in Selecting among the 1PL, 2PL, and 3PL Models
224(2)
Summary
226(11)
7 Rasch Models for Ordered Polytomous Data
237(76)
Conceptual Development of the Partial Credit Model
238(5)
Conceptual Parameter Estimation of the PC Model
243(1)
Example: Application of the PC Model to a Reasoning Ability Instrument, MMLE, flexMIRT I
244(12)
Example: Application of the PC Model to a Reasoning Ability Instrument, MMLE, mirt
256(11)
The Rating Scale Model
267(5)
Conceptual Parameter Estimation of the RS Model
272(1)
Example: Application of the RS Model to an Attitudes Toward Condoms Scale, JMLE, BIGSTEPS I
272(15)
Example: Application of the PC Model to an Attitudes Toward Condoms Scale, JMLE, mixRasch
287(5)
How Large a Calibration Sample?
292(2)
Information for the PC and RS Models
294(2)
Metric Transformation, PC and RS Models
296(1)
Summary
296(17)
8 Non-Rasch Models for Ordered Polytomous Data
313(43)
The Generalized Partial Credit Model
313(5)
Example: Application of the GPC Model to a Reasoning Ability Instrument, MMLE, flexMIRT
318(3)
Example: Application of the GPC Model to a Reasoning Ability Instrument, MMLE, mirt
321(3)
Conceptual Development of the Graded Response Model
324(9)
How Large a Calibration Sample?
333(1)
Information for Graded Data
334(2)
Metric Transformation, GPC and GR Models
336(1)
Example: Application of the GR Model to an Attitudes Toward Condoms Scale, MMLE, flexMIRT I
337(3)
Example: Application of the GR Model to an Attitudes Toward Condoms Scale, MMLE, mirt
340(3)
Conceptual Development of the Continuous Response Model
343(8)
Summary
351(5)
9 Models for Nominal Polytomous Data
356(35)
Conceptual Development of the Nominal Response Model
357(8)
Information for the NR Model
365(1)
Metric Transformation, NR Model
366(1)
Conceptual Development of the Multiple-Choice Model
366(2)
How Large a Calibration Sample?
368(2)
Example: Application of the NR Model to a General Science Test, MMLE, mirt
370(13)
Summary
383(8)
10 Models for Multidimensional Data
391(52)
Conceptual Development of a Multidimensional IRT Model
391(6)
Multidimensional Item Location and Discrimination
397(4)
Item Vectors and Vector Graphs
401(3)
The Multidimensional Three-Parameter Logistic Model
404(1)
Assumptions of the MIRT Model
404(1)
Estimation of the M2PL Model
405(1)
Information for the M2PL Model
406(2)
Indeterminacy in MIRT
408(2)
Metric Transformation, M2PL Model
410(1)
Example: Calibration of Interpersonal Engagement Instrument, M2PL Model, sirt.noharm
411(10)
Obtaining Person Location Estimates
421(1)
Example: Calibration of Interpersonal Engagement Instrument, M2PL Model, mirt
422(7)
Example: Calibration of Interpersonal Engagement Instrument, M2PL Model, flexMIRT I
429(2)
Summary
431(12)
11 Linking and Equating
443(35)
Equating Defined
443(2)
Equating: Data Collection Phase
445(1)
Equating: Transformation Phase
446(8)
Example: Application of the Total Characteristic Function Equating Method, EQUATE
454(9)
Example: Application of the Total Characteristic Function Equating Method, SNSequate
463(2)
Example: Fixed-Item and Concurrent Calibration Equating
465(6)
Summary
471(7)
12 Differential Item Functioning
478(47)
Differential Item Functioning and Item Bias
479(4)
Mantel-Haenszel Chi-Square
483(3)
The TSW Likelihood Ratio Test
486(1)
Logistic Regression
487(4)
Example: DIF Analysis of Vocabulary Test, SAS CMH
491(3)
Example: DIF Analysis of Vocabulary Test, mantelhaen. test and difR
494(7)
Example: DIF Analysis of Vocabulary Test, SAS proc logistic
501(7)
Example: DIF Analysis of Vocabulary Test, glm and difR
508(10)
Summary
518(7)
13 Multilevel IRT Models
525(72)
Multilevel IRT---Two Levels
525(5)
Example: Estimating the Rasch Model from a Multilevel Perspective, proc glimmix
530(11)
Example: Rasch Model Estimation, lme4
541(4)
Person-Level Predictors for Items
545(2)
Example: Person-Level Predictors for Items---DIF Analysis, proc glimmix
547(4)
Example: Person-Level Predictors for Items---DJF Analysis, lme4
551(5)
Person-Level Predictors for Respondents
556(2)
Example: Person-Level Predictors for Respondents---Nutrition Literacy, proc glimmix
558(4)
Example: Person-Level Predictors for Respondents, lme4
562(5)
Item-Level Predictors for Items
567(2)
Example: Item-Level Predictors for Items---Nutrition Literacy, proc glimmix
569(2)
Example: Item-Level Predictors for Items---Nutrition Literacy, lme4
571(3)
Multilevel IRT---Three Levels
574(5)
Example: Three-Level Model Analysis---Nutrition Literacy, proc glimmix
579(3)
Example: Three-Level Analysis of Nutrition Literacy Data, lme4
582(5)
Summary
587(10)
Appendices A-G Can be accessed online at the book's companion website (www. guilford.com/deayala-materials), which also provides links to data, syntax, and output files in different software packages for the book's examples
Appendix A Maximum Likelihood Estimation of Person Locations
Estimating an Individual's Location: Empirical Maximum Likelihood
Estimation
Estimating an Individual's Location: Newton's Method for MLE
R Function for MLE of with the Rasch Model
Revisiting Zero Variance Binary Response Patterns
Appendix B Maximum Likelihood Estimation of Item Locations
R function for MLE of 5 with the Rasch Model
Appendix C The Normal Ogive Models
Conceptual Development of the Normal Ogive Model
The Relationship between IRT Statistics and Traditional Item Analysis
Indices
Relationship of the Two-Parameter Normal Ogive and Logistic Models
Extending the Two-Parameter Normal Ogive Model to a Multidimensional
Space
Appendix D Computerized Adaptive Testing
A Brief History
Fixed-Branching Techniques
Variable-Branching Techniques
Advantages of Variable-Branching over Fixed-Branching Methods
IRT-Based Variable-Branching Adaptive Testing Algorithm
Appendix E. Linear Logistic Test Model (LLTM)
Example of LLTM Calibration Using eRm
Appendix F Mixture Models
Latent Class Analysis
Mixture Rasch Model
Example: Application of the Mixture Rasch Model to Writing Problem Data, CMLE, WINMIRA
Example: Application of the Mixture Rasch Model to Writing Problem Data, CMLE, psychomix
Appendix G Miscellanea
Using Principal Axis for Estimating Item Discrimination
Infinite Item Discrimination Parameter Estimates
Example: NOFIARM Unidimensional Calibration
An Approximate Chi-Square Statistic for NOHARM
Relative Efficiency, Monotonicity, and Information
FORTRAN Formats
Odds, Odds Ratios, and Logits
The Person Response Function
Linking: A Temperature Analogy Example
Should DIF Analyses Be Based on Latent Classes?
The Separation and Reliability Indices
Dependency in Traditional Item Statistics and Observed Scores
Conditional Independence Using 3
Standalone NOHARM Calibration of Interpersonal Engagement Instrument, M2PL Model
CFI, GFI, M2, RMSEA, TLI, and SRMR
An Introduction to Kernel Equating
Correspondence between the Rasch Model and a Loglinear Model
R Introduction
References 597(28)
Author Index 625(6)
Subject Index 631(12)
About the Author 643
R. J. de Ayala, PhD, is Professor of Quantitative, Qualitative, and Psychometric Methods and Director of the Institutional Research Master's Program in the College of Educational and Human Sciences at the University of NebraskaLincoln (UNL). His research interests include psychometrics, item response theory, computerized adaptive testing, applied statistics, and multilevel models. His work has appeared in Applied Psychological Measurement, Applied Measurement in Education, the British Journal of Mathematical and Statistical Psychology, Educational and Psychological Measurement, the Journal of Applied Measurement, and the Journal of Educational Measurement. He is a Fellow of the American Psychological Associations Division 5: Evaluation, Measurement, and Statistics and of the American Educational Research Association. He is a recipient of a Big 12 Faculty Fellowship and holds a Gallup Research Professorship at UNL.