Muutke küpsiste eelistusi

E-raamat: Applied Multiple Imputation: Advantages, Pitfalls, New Developments and Applications in R

3.00/5 (2 hinnangut Goodreads-ist)
, , ,
Teised raamatud teemal:
  • Formaat - PDF+DRM
  • Hind: 67,91 €*
  • * hind on lõplik, st. muud allahindlused enam ei rakendu
  • Lisa ostukorvi
  • Lisa soovinimekirja
  • See e-raamat on mõeldud ainult isiklikuks kasutamiseks. E-raamatuid ei saa tagastada.
Applied Multiple Imputation: Advantages, Pitfalls, New Developments and Applications in RSuurem pilt
Teised raamatud teemal:

DRM piirangud

  • Kopeerimine (copy/paste):

    ei ole lubatud

  • Printimine:

    ei ole lubatud

  • Kasutamine:

    Digitaalõiguste kaitse (DRM)
    Kirjastus on väljastanud selle e-raamatu krüpteeritud kujul, mis tähendab, et selle lugemiseks peate installeerima spetsiaalse tarkvara. Samuti peate looma endale  Adobe ID Rohkem infot siin. E-raamatut saab lugeda 1 kasutaja ning alla laadida kuni 6'de seadmesse (kõik autoriseeritud sama Adobe ID-ga).

    Vajalik tarkvara
    Mobiilsetes seadmetes (telefon või tahvelarvuti) lugemiseks peate installeerima selle tasuta rakenduse: PocketBook Reader (iOS / Android)

    PC või Mac seadmes lugemiseks peate installima Adobe Digital Editionsi (Seeon tasuta rakendus spetsiaalselt e-raamatute lugemiseks. Seda ei tohi segamini ajada Adober Reader'iga, mis tõenäoliselt on juba teie arvutisse installeeritud )

    Seda e-raamatut ei saa lugeda Amazon Kindle's. 

This book explores missing data techniques and provides a detailed and easy-to-read introduction to multiple imputation, covering the theoretical aspects of the topic and offering hands-on help with the implementation. It discusses the pros and cons of various techniques and concepts, including multiple imputation quality diagnostics, an important topic for practitioners. It also presents current research and new, practically relevant developments in the field, and demonstrates the use of recent multiple imputation techniques designed for situations where distributional assumptions of the classical multiple imputation solutions are violated. In addition, the book features numerous practical tutorials for widely used R software packages to generate multiple imputations (norm, pan and mice). The provided R code and data sets allow readers to reproduce all the examples and enhance their understanding of the procedures. This book is intended for social and health scientists and other quantitative researchers who analyze incompletely observed data sets, as well as master’s and PhD students with a sound basic knowledge of statistics. 

Arvustused

This is an interesting book encouraging the application of the content presented. (Maria de Ridder, ISCB News, iscb.info, Issue 70, December, 2020)

1 Introduction and Basic Concepts
1(22)
1.1 Overview
1(1)
1.2 Introduction
1(1)
1.3 Unit and Item Nonresponse
2(1)
1.4 Patterns of Observed and Missing Values
3(1)
1.5 Inferential Approaches and Concepts
4(4)
1.5.1 Design-Based Approach
5(1)
1.5.2 Model-Based Approaches
6(2)
1.6 Statistical Inference and Missing Data
8(1)
1.7 The Method of Multiple Imputation
9(1)
1.8 Illustration of Concepts and Techniques
10(10)
1.8.1 Unit and Item Nonresponse, (Non)response Pattern
10(1)
1.8.2 Design-Based Approach
10(4)
1.8.3 Model-Based Approaches
14(6)
1.9 Outline
20(3)
References
21(2)
2 Missing Data Mechanism and Ignorability
23(30)
2.1 Overview
23(1)
2.2 Introduction
23(2)
2.2.1 Some Notation
24(1)
2.3 The Missing Data Mechanism
25(6)
2.3.1 Missing Completely at Random (MCAR)
26(2)
2.3.2 Missing at Random (MAR)
28(2)
2.3.3 Missing Not at Random (MNAR)
30(1)
2.4 Ignorability of the Missing Data Mechanism
31(10)
2.4.1 Frequentist Approach
33(2)
2.4.2 Likelihood Approach
35(5)
2.4.3 Bayesian Approach
40(1)
2.5 Supplementary Notes
41(4)
2.6 Diagnostic Tools
45(5)
2.7 Summary
50(3)
References
51(2)
3 Missing Data Methods
53(32)
3.1 Overview
53(1)
3.2 Ad-hoc Methods
53(9)
3.2.1 Complete Case and Available Case Analysis
54(1)
3.2.2 Completing Data Sets
55(7)
3.3 Maximum Likelihood Estimation
62(13)
3.3.1 Introduction
62(3)
3.3.2 The Ignorable Case
65(5)
3.3.3 The Nonignorable Case
70(5)
3.4 Weighting
75(3)
3.5 Imputation Methods
78(3)
3.5.1 Single Imputation
79(1)
3.5.2 Multiple Imputation
80(1)
3.6 Summary
81(4)
References
82(3)
4 Multiple Imputation: Theory
85(48)
4.1 Overview
85(1)
4.2 Multiple Imputation: Introduction
85(1)
4.3 Theoretical Background
86(14)
4.3.1 Valid Inferences
86(2)
4.3.2 Bayesian Motivation
88(1)
4.3.3 The Combining Rules
89(1)
4.3.4 Frequentist Evaluation
90(8)
4.3.5 Inferences Based on Finite M
98(2)
4.4 How to Generate Multiple Imputations
100(11)
4.4.1 General
100(2)
4.4.2 Monotone-Distinct Structure
102(4)
4.4.3 Beyond Monotone-Distinct Structures
106(1)
4.4.4 Arbitrary Missing Data Patterns: A General Procedure
107(2)
4.4.5 Generating Imputations by Joint Modeling
109(1)
4.4.6 Generating Imputations by Fully Conditional Specification
110(1)
4.5 Further Topics and Open Problems
111(17)
4.5.1 Transforming, Rounding and Trimming
111(4)
4.5.2 Imputation of Functions of Variables
115(3)
4.5.3 Imputing Subgroups
118(4)
4.5.4 The Imputation Models: (Non-)ignorability
122(1)
4.5.5 Reproducible Results
122(1)
4.5.6 Some Additional Notes
123(1)
4.5.7 Choosing from a Large Number of Imputation Techniques
124(4)
4.6 Summary
128(5)
References
129(4)
5 Multiple Imputation: Application
133(86)
5.1 Overview
133(2)
5.2 Data and Substantive Model
135(7)
5.2.1 The CrimoC data
135(2)
5.2.2 Growth Curve Models
137(5)
5.3 Multiple Imputation of Missing Data
142(8)
5.3.1 Reading in the CrimoC data
142(1)
5.3.2 Missing Data Inspection
143(6)
5.3.3 Choosing the Multiple Imputation Framework
149(1)
5.3.4 Building the Imputation Model
149(1)
5.4 Multiple Imputation with mice
150(8)
5.4.1 Argument method
151(1)
5.4.2 Argument predictorMatrix
151(1)
5.4.3 Multiple Imputation by Predictive Mean Matching
152(5)
5.4.4 Summary
157(1)
5.5 Other R Packages and Functions for Multivariate Data Imputation
158(16)
5.5.1 Norm2
158(7)
5.5.2 Amelia
165(4)
5.5.3 Mi
169(2)
5.5.4 Areglmpute
171(1)
5.5.5 Comparison of Results
172(2)
5.6 Multiple Imputation of Clustered or Panel Data Using Multilevel Models
174(9)
5.6.1 A Brief Introduction to Multilevel Regression Models
175(1)
5.6.2 Strategies to Impute Clustered or Panel Data
176(1)
5.6.3 Application
177(6)
5.7 Repeated Data Analysis and Pooling of Results in R
183(21)
5.7.1 T-Test
184(3)
5.7.2 Linear Model
187(1)
5.7.3 Growth Model
188(2)
5.7.4 Generalized Linear Mixed-Effects Model
190(3)
5.7.5 Two-Level Negative Binomial Hurdle Model
193(2)
5.7.6 Correlation Coefficients
195(1)
5.7.7 Further Useful Tools for Multiple Imputation Inference
196(8)
5.8 Diagnostic Tools
204(15)
5.8.1 Inspection of Imputed Values Using the boxplot() Function
204(2)
5.8.2 Comparison of the Distributions of Observed and Imputed Values Using the hist() Function
206(2)
5.8.3 Comparison of Observed and Imputed Values Using the stripplot() Function
208(1)
5.8.4 Comparison of Observed and Imputed Values Using compare.percent.count()
209(1)
5.8.5 Comparison of Observed and Imputed Values Using compare.obs.imp()
210(1)
5.8.6 Assessing the Suitability of the Imputation Method by Means of Monte Carlo Simulation
211(3)
References
214(5)
6 Multiple Imputation: New Developments
219(38)
6.1 Overview
219(1)
6.2 Parametric Multiple Imputation of Incomplete Count Data
220(10)
6.2.1 Count Data Modeling in a Nutshell
221(1)
6.2.2 The count imp Package
222(8)
6.2.3 Discussion
230(1)
6.3 ImputeRobust--Imputation Based on Generalized Additive Models for Location, Scale and Shape
230(5)
6.3.1 ImputeRobust in a Nutshell
231(2)
6.3.2 Application
233(2)
6.3.3 Discussion
235(1)
6.4 Quantile Regression Based Multiple Imputation
235(4)
6.4.1 Application
237(1)
6.4.2 Discussion
238(1)
6.5 Two-Level Predictive Mean Matching
239(2)
6.5.1 The General Idea
239(1)
6.5.2 Application
240(1)
6.5.3 Discussion
240(1)
6.6 The Growth Curve ZIP Model Revisited
241(10)
6.7 Future Research--Multiple Imputation After More Than Three Decades
251(6)
References
254(3)
A Matrix Algebra, Random Variables and Some Technical Results
257(30)
A.1 Matrix Algebra
257(8)
A.1.1 Vectors and Matrices
257(3)
A.1.2 Basic Operations
260(2)
A.1.3 Properties of Matrices
262(3)
A.2 Random Variables and Statistics
265(11)
A.2.1 Sample Moments
266(1)
A.2.2 Random Variables
267(2)
A.2.3 Some Results on Expectations, Variances and Covariances
269(1)
A.2.4 Normal Distribution
270(1)
A.2.5 Conditional Normal Distribution
271(1)
A.2.6 Truncated Normal Distribution and Normal Sample Selection Model
272(4)
A.3 Estimation
276(7)
A.3.1 Least Squares
276(2)
A.3.2 Maximum Likelihood
278(5)
A.3.3 Information Criteria
283(1)
A.4 Multiple Imputation: Some Technical Results
283(4)
A.4.1 Derivation of (4.6), (4.8) and (4.9)
284(1)
A.4.2 Simplification of f(u(0) | u(i), 6) and p(θ | u(i))
284(1)
References
285(2)
Glossary 287(2)
Index 289
Kristian Kleinke received his PhD from the University of Bielefeld and is currently an interim Professor of Psychological Methods and General Psychology at the University of Siegen, Germany. His primary research interests include missing data and multiple imputation. His methodological research focuses on multiple imputation solutions for complex data structures like panel data and non-normal missing data problems, i.e. when convenient distributional assumptions of the standard MI procedures are violated.





Jost Reinecke is a Professor of Quantitative Methods of Empirical Social Research at the University of Bielefeld, Germany. His current methodological research focuses on growth curve and growth mixture models and the development of techniques related to multiple imputation in complex survey designs. His substantive research focuses on the development of adolescents' delinquent behavior and relationships between group-focused enmity and individual and contextual variables.







Daniel Salfrán was a member of the Applied Mathematics Department and the Cryptography Group at the University of Havana, Cuba, where he worked on a spatial stochastic model for Dengue epidemics. He received his PhD from the University of Hamburg, Germany and is currently lecturer at the Institute for Psychology, University of Hamburg. His research focuses on robust methods to generate multiple imputations.







Martin Spiess is a Professor of Psychological Methods and Statistics at the University of Hamburg, Germany. He studied Psychology, received his PhD in Statistics on the estimation of categorical panel models and was a Research Assistant at the German Institute for Economic Research (DIW). His current research focuses on the estimation of regression and panel data models and techniques to compensate for missing units and missing items.
Lisainfo e-raamatute kohta Lisainfo e-raamatute kohta
Püsilink: https://www.kriso.ee/db/97830303816462e.html
Märksõnad: