>> Home >> Bookstore >> Biostatistics and epidemiology >> Statistical Analysis of Human Growth and Development

Statistical Analysis of Human Growth and Development

Yin Bun Cheung
Publisher: Chapman & Hall/CRC
Copyright: 2014
ISBN-13: 978-1-43-987154-6
Pages: 356; hardcover
Price: $82.50

Comment from the Stata technical group

This book covers a broad array of statistical techniques useful to researchers in the fields of human growth and development. The author defines human growth as changes to anthropometric measurements, including body mass index and height. On the other hand, development is defined as changes in abilities, such as cognitive abilities, and changes in functioning. Noting that growth and development are strongly related, Chueng discusses statistical analysis of both types of data.

Statistical Analysis of Human Growth and Development describes basic and advanced statistical methods, demonstrating how each one may be used to address common research problems. Examples show Stata code and output.

Topics include an overview of basic statistical concepts, causality, linear regression, quantile regression, logistic regression, survival analysis, mixed models, fractional polynomials, item–response theory, validity, reliability, and multiple imputation.

Table of contents

1 Introduction
1.1 Overview
1.2 Human Growth
1.2.1 Fetal Period
1.2.2 At Birth
1.2.3 Infancy to Puberty
1.2.4 Adulthood
1.3 Human Development
1.3.1 Fetal Period
1.3.2 Infancy
1.3.3 Childhood to Adolescence
1.3.4 Adulthood
1.4 Statistical Considerations
1.4.1 Issues Arising from Nature and Nurture
1.4.2 Issues Arising from Measurement
1.4.3 Complex Relationship
1.4.4 Normal versus Abnormal States
2 Causal Reasoning and Study Designs
2.1 Causality
2.1.1 Association
2.1.2 Temporal Sequence
2.1.3 Nonspurious Association
2.1.4 Observed Association Not Due to Chance
2.2 Study Designs
2.2.1 Cross-Sectional versus Longitudinal Studies
2.2.2 Randomized Trials versus Observational Studies
2.2.3 Further Topics in Study Designs
3 Basic Statistical Concepts and Tools
3.1 Normal Distribution
3.2 Statistical Inference and Significance
3.2.1 Population and Sample
3.2.2 Statistical Inference and Statistical Significance
3.2.3 Clinical Significance and Effect Size
3.2.4 Parameters
3.2.5 Maximum Likelihood
3.2.6 Derivatives
3.2.7 Normality and Robustness
3.3 Standardized Scores
3.3.1 Transformation
3.3.2 Calculation and Application
3.4 Statistical Programming
3.4.1 Ensuring Reproducibility
3.4.2 Date and Age
3.4.3 Long and Wide Formats
3.4.4 Saved Results
3.4.5 Computer Software
4 Quantifying Growth and Development: Use of Existing Tools
4.1 Growth
4.1.1 Ratio Index
4.1.2 Transformation: Two- and Three-Parameter Models
4.1.3 Use of Existing References
4.2 Development
4.2.1 Standard Use of Multi-Item Inventories
4.2.2 Use of Selected Items
4.3 Change Scores
5 Regression Analysis of Quantitative Outcomes
5.1 Least-Squares Regression
5.1.1 Model, Estimation, and Inference
5.1.2 Model Diagnostics
5.1.3 Robustness
5.1.4 Multivariable Least-Squares Regression
5.2 Quantile Regression
5.2.1 Single Quantile
5.2.2 Simultaneous Quantiles
5.2.3 Multivariable Quantile Regression
5.3 Covariate-Adjusted Variables
6 Regression Analysis of Binary Outcomes
6.1 Basic Concepts
6.1.1 Risk and Odds
6.1.2 Sensitivity and Specificity
6.2 Introduction to Generalized Linear Models
6.3 Logistic Regression
6.3.1 Model and Estimation
6.3.2 Statistcal Inference
6.3.3 Model Diagnostics
6.4 Log-Binomial and Binomial Regression Models
7 Regression Analysis of Censored Outcomes
7.1 Fundamentals
7.1.1 Censoring
7.1.2 Hazard and Hazard Ratio
7.2 Regression Analysis of Right-Censored Data
7.2.1 Cox Regression Model
7.2.2 Parametric Models
7.3 Analysis of Interval-Censored Data
7.3.1 Midpoint Aprroximation
7.3.2 Parametrics Modeling
8 Analysis of Repeated Measurements and Clustered Data
8.1 Introduction
8.1.1 Why Analyze Repeated Measures?
8.1.2 Naíve Analysis
8.1.3 Population- versus Subject-Level Impact
8.1.4 Panel- and Observation-Level Variables
8.1.5 Weighting
8.2 Robust Variance Estimator
8.3 Analysis of Subject-Level Summary Statistics
8.3.1 Mean and Rate of Change
8.3.2 Spline Models
8.4 Mixed Models
8.4.1 Random Intercept Models
8.4.2 Random Coefficient Models
8.4.3 Growth Curves
8.4.4 Generalized Linear Mixed Models
8.4.5 Further Remarks
9 Quantifying Growth: Development of New Tools
9.1 Capturing Nonlinear Relationships
9.1.1 Fractional Polynomials
9.1.2 Cubic Splines
9.1.3 Transformation of X-Axis Variable
9.2 Modeling
9.2.1 Quantile Regression
9.2.2 Parametric Method
9.2.3 LMS Method
9.2.4 Model Diagnostics
10 Quantifying Development: Development of New Tools
10.1 Summary Index Based on Binary Items
10.1.1 Estimation of Ability Age Using Item Characteristic Curves
10.1.2 Estimation of Latent Trait Using Item-Response Theory
10.2 Summary Index Based on Quantitative Variables
10.2.1 Inferential Norming
10.2.2 Factor Analysis
11 Defining Growth and Development: Longitudinal Measurements
11.1 Expected Change and Unexplained Residuals
11.1.1 More on Change Score
11.1.2 Expected Change
11.1.3 Unexplained Residuals
11.2 Reference Intervals for Longitudinal Monitoring
11.2.1 Fitting Mixed Models
11.2.2 Unconditional References
11.2.3 Conditional References
11.3 Conditional Scores by Quantile Regression
11.4 Trajectory Characteristics
11.4.1 Fractional Polynomials
11.4.2 Penalized Cubic Spline Model
11.4.3 Graphical Assessment
12 Validity and Reliability
12.1 Concepts
12.1.1 Aspects of Validity
12.1.2 Aspects of Reliability
12.2 Statistical Methods
12.2.1 Correlation and Regression
12.2.2 ANOVA and Related Methods
12.2.3 Bland-Altman Method
12.2.4 Kappa
12.3 Further Topics
12.3.1 Cronbach's Alpha
12.3.2 Acceptable Levels
13 Missing Values and Imputation
13.1 Introduction
13.1.1 Patterns of Missing Data
13.1.2 Mechanisms of Missing Data
13.2 When Is Missing Data (Not) a Problem?
13.3 Interpolation and Extrapolation
13.3.1 Interpolation
13.3.2 Extrapolation
13.4 Mixed Models
13.5 Multiple Imputation
13.5.1 Univariate Imputation
13.5.2 Combinations of Results
13.5.3 Choice of m
13.5.4 Multivariate Imputation
13.5.5 Imputation and Analytic Models
13.5.6 Further Remarks
13.6 Imputing Censored Data
14 Multiple Comparisons
14.1 The Problem
14.2 When Not to Do Multiplicity Adjustment
14.2.1 Different Families of Hypothesis
14.2.2 Reject All Null Hypotheses
14.2.3 To Err on the Side of False Positive
14.3 Strategies of Analysis to Prevent Multiplicity
14.3.1 Statistical Analysis Plan
14.3.2 Defining a Primary Analysis
14.4 P-Value Adjustments
14.4.1 Bonferroni Adjustment
14.4.2 Holm Adjustment
14.4.3 False Discovery Rate
14.5 Close Testing Procedure
14.5.1 Multiple Groups
14.5.2 Multiple Endpoints
15 Regression Analysis Strategy
15.1 Introduction
15.2 Rationale of Using Multivariable Regression
15.2.1 Positive and Negative Confounding
15.2.2 Baseline Characteristics in Randomized Trials
15.2.3 Reflection of Study Design
15.2.4 Prediction and Prognostic Index
15.2.5 Answering Different Questions
15.3 Point Measures, Change Scores, and Unexplained Residuals
15.4 Issues in Variable Selection
15.4.1 Forced-Entry
15.4.2 Improper Forward Selection
15.4.3 Futher Remarks
15.5 Interaction
15.5.1 Concepts
15.5.2 Analysis
15.5.3 Model Dependency
15.6 Role of Prior Knowledge
Appendix A: Stata Codes to Generate Simulated Clinical Trial (SCT) Dataset
Appendix B: Stata Codes to Generate Simulated Longitudinal Study (SLS) Dataset
Appendix C: Stata Program for Detrended Q-Q Plot
Appendix D: Stata Program for Weighted Maximum Likelihood Estimation for Binary Items
The Stata Blog: Not Elsewhere Classified Find us on Facebook Follow us on Twitter LinkedIn Google+ Watch us on YouTube