New Statistics for Design Researchers: A Bayesian Workflow in Tidy R (Human–Computer Interaction Series)

Contents
Part I Preparations
1 Introduction
	1.1 Whom This Book Is For
	1.2 Quantitative Design Research
	1.3 What Is New Statistics?
	1.4 How to Use This Book
		1.4.1 Routes
		1.4.2 In the Classroom
		1.4.3 The Stone of Rosetta
	1.5 Thank You and Supplementary Readings
	References
2 Getting Started with R
	2.1 Setting up the R Environment
		2.1.1 Installing CRAN Packages
		2.1.2 Installing Packages from GitHub
		2.1.3 A First Statistical Program
		2.1.4 Bibliographic Notes
	2.2 Learning R: A Primer
		2.2.1 Assigning and Calling Objects
		2.2.2 Vectors
		2.2.3 Basic Object Types
		2.2.4 Operators and Functions
		2.2.5 Storing Data in Data Frames
		2.2.6 Import, Export and Archiving
		2.2.7 Case Environments
		2.2.8 Structuring Data
		2.2.9 Data Transformation
		2.2.10 Plotting Data
		2.2.11 Fitting Regression Models
		2.2.12 Knitting Statistical Reports
	2.3 Further Reading
	Reference
3 Elements of Bayesian Statistics
	3.1 Rational Decision-Making in Design Research
		3.1.1 Measuring Uncertainty
		3.1.2 Benchmarking Designs
		3.1.3 Comparison of Designs
		3.1.4 Prior Knowledge
	3.2 Observations and Measures
		3.2.1 Interaction Sequences
		3.2.2 Performance Measures
		3.2.3 Satisfaction and Other Feelings
	3.3 Descriptive Statistics
		3.3.1 Frequencies
		3.3.2 Central Tendency
		3.3.3 Dispersion
		3.3.4 Associations
	3.4 Bayesian Probability Theory
		3.4.1 Some Set Theory
		3.4.2 Probability
		3.4.3 Likelihood
		3.4.4 Bayesian and Frequentist Probability
		3.4.5 Bayes\' Theorem
		3.4.6 Bayesian Dynamics of Belief
	3.5 Statistical Models
		3.5.1 The Structural Part
		3.5.2 Distributions: Shapes of Randomness
	3.6 Toward Bayesian Estimation
	3.7 On Priors and Defaults
	3.8 Further Readings
	References
Part II Models
4 Basic Linear Models
	4.1 Quantification at Work: Grand Mean Models
		4.1.1 Do the Random Walk: Markov Chain Monte Carlo Sampling
		4.1.2 Likelihood and Random Term
		4.1.3 Working with the Posterior Distribution
		4.1.4 Center and Interval Estimates
	4.2 Walk the Line: Linear Regression
		4.2.1 Transforming Measures
		4.2.2 Correlations
		4.2.3 Endlessly Linear
	4.3 Factorial Models
		4.3.1 A Versus B: Comparison of Groups
		4.3.2 Not Stupid: Dummy Variables
		4.3.3 Treatment Contrast Coding
		4.3.4 Absolute Means Model
		4.3.5 Ordered Factorial Models
5 Multi-predictor Models
	5.1 On Surface: Multiple Regression Models
	5.2 Crossover: Multifactorial Models
	5.3 Line-by-Line: Grouped Regression Models
	5.4 Conditional Effects Models
		5.4.1 Conditional Multiple Regression
		5.4.2 Conditional Multifactorial Models
		5.4.3 Saturation: Hitting the Boundaries
		5.4.4 Amplification: More than the Sum
		5.4.5 Conditional Effects and Design Theory
	5.5 Doing the Rollercoaster: Polynomial Regression Models
		5.5.1 Make Yourself a Test Statistic
	5.6 Further Readings
	References
6 Multilevel Models
	6.1 The Human Factor: Intercept Random Effects
	6.2 Multi-level Linear Regression: Variance in Change
	6.3 Thinking Multi-level
	6.4 Testing Universality of Theories
	6.5 Non-human Populations and Cross-Overs
	6.6 Nested Random Effects
	6.7 What Are Random Effects? On Pooling and Shrinkage
	6.8 Psychometrics and Design-o-Metric Models
		6.8.1 Coverage
		6.8.2 Reliability
		6.8.3 Validity
		6.8.4 Towards Design-o-Metrix
	6.9 Further Readings
	References
7 Generalized Linear Models
	7.1 Elements of Generalized Linear Models
		7.1.1 Re-linking Linearity
		7.1.2 Choosing Patterns of Randomness
		7.1.3 Mean-Variance Relationship
	7.2 Count Data
		7.2.1 Poisson Regression
		7.2.2 Logistic (aka Binomial) Regression
		7.2.3 Modeling Overdispersion
	7.3 Duration Measures
		7.3.1 Exponential and Gamma Regression
		7.3.2 ExGaussian Regression
	7.4 Rating Scales
		7.4.1 Ordered Logistic Regression
		7.4.2 Beta Regression
	7.5 Beyond Mean: Distributional Models
		7.5.1 Item-Level Anchoring in Rating Scales
		7.5.2 Participant-Level Employment of Scale
		7.5.3 Participant-Level Skew in Reaction Times
	7.6 Further Readings
	References
8 Working with Models
	8.1 Model Criticism
		8.1.1 Residual Analysis
		8.1.2 Fitted Responses Analysis
	8.2 Model Comparison
		8.2.1 The Problem of Overfitting
		8.2.2 Cross Validation and LOO
		8.2.3 Information Criteria
		8.2.4 Model Selection
		8.2.5 Comparing Response Distributions
		8.2.6 Testing Hypotheses
		8.2.7 A Note on Bayes Factor
	8.3 Further Readings
	References
Appendix  Cases
.1  Real Cases
A.1  Hugme
A.1.1  Measures
A.1.2  Stimuli
A.1.3  Data Modeling in a Nutshell
A.2  CUE8
A.2.1  Measures
A.3  Uncanny Valley
A.3.1  Experimental Design
A.3.2  Stimuli
A.3.3  Measures
A.4  IPump
A.4.1  Experimental Design
A.4.2  Measures
A.5  Case Sleepstudy
A.6  Egan
A.6.1  Research Design
A.6.2  Measures
A.7  Case: Millers Magic Number
A.7.1  Research Design
A.7.2  Measures
A.8  AUP
A.8.1  Measures
A.2  Synthetic Data Sets
A.1  Rainfall
A.2  99 seconds
A.3  Rational
A.4  BrowsingAB
A.5  Headache
A.6  Reading Time
A.7  AR_game
A.8  Sleep
	References