Robert Rosenthal is DistinguishedProfessor at the University of
Californiaat Riverside and Edgar Pierce Professorof Psychology,
Emeritus, Harvard University.His research has centered for some50
years on the role of the self-fulfi llingprophecy in everyday life
and in laboratorysituations. Special interests include theeffects
of teachers’ expectations on students’performance, the effects of
experimenters’expectations on the results oftheir research, and
the effects of clinicians’expectations on their patients’mental and
physical health. He also hasstrong interests in sources of
artifact inbehavioral research and in various quantitative
procedures. In the realm of data analysis,his special interests
are in experimental design and analysis, contrast analysis,and
meta-analysis. His most recent books and articles are about these
areas of dataanalysis and about the nature of nonverbal
communication in teacher-student, doctorpatient,manager-employee,
judge-jury, and psychotherapist-client interaction. He hasbeen
Co-Chair of the Task Force on Statistical Inference of the American
PsychologicalAssociation and has served as Chair of the Research
Committee of the BayerInstitute for Health Care Communication. He
was a co-recipient of two behavioralscience awards of the American
Association for the Advancement of Science (1960,1993) and
recipient of the James McKeen Cattell Award of the American
PsychologicalSociety, the Distinguished Scientist Award of the
Society of ExperimentalSocial Psychology, the Samuel J. Messick
Distinguished Scientifi c ContributionsAward of the APA’s Division
5—Evaluation, Measurement, and Statistics, and theAPA’s
Distinguished Scientifi c Award for Applications of Psychology.

Ralph L. Rosnow is Thaddeus Bolton Professor Emeritus at Temple
University, where

he taught for 34 years and directed the graduate program in social
and organizational

psychology. He has also taught research methods at Boston
University and Harvard

University and does consulting on research and data analysis. The
overarching theme

of his scholarly work concerns how people make sense of, and impose
meaning on,

their experiential world, called the “will to meaning” by Viktor
Frankl. Rosnow has

explored aspects of this construct in research and theory within
the framework of

contextualism, the psychology of rumor and gossip, attitude and
social cognition, the

structure of interpersonal acumen, artifacts and ethical dilemmas
in human research,

and the statistical justifi cation of scientifi c conclusions. He
has authored and coauthored

many articles and books on these topics and, with Mimi Rosnow,
coauthored Writing

Papers in Psychology, a popular writing manual now in its seventh
edition (published

by Thomson Wadsworth, 2006). He has served on the editorial boards
of journals and

encyclopedias, was coeditor (with R. E. Lana) of the Reconstruction
of Society Series

published by Oxford University Press, and chaired the APA’s
Committee on Standards

in Research. He is a fellow of the American Association for the
Advancement of

Science, the APA, and the Association for Psychological Science,
received the Society

of General Psychology’s George A. Miller Award, and was recently
honored with a

Festschrift book edited by D. A. Hantula, Advances in Social and
Organizational Psychology

Rosenthal and Rosnow have also collaborated on other books on
research methods

and data analysis, including Artifact in Behavioral Research
(Academic Press, 1969);

The Volunteer Subject (Wiley, 1975); Primer of Methods for the
Behavioral Sciences

(Wiley, 1975); Understanding Behavioral Science: Research Methods
for Research

Consumers (McGraw-Hill, 1984); Contrast Analysis: Focused
Comparisons in the

Analysis of Variance (Cambridge University Press, 1985); People
Studying People:

Artifacts and Ethics in Behavioral Research (W. H. Freeman, 1997);
(with D. B. Rubin)

Contrasts and Effect Sizes in Behavioral Research: A Correlational
Approach

(Cambridge University Press, 2000); and Beginning Behavioral
Research: A Conceptual

Primer (6 th ed., Pearson/PrenticeHall, 2008).

編 概念與倫理基礎

第1章行為研究的精神

第2章探索與辯護的環境

第3章倫理問題、道德兩難問題與指導方針

第二編 因變量的可操作化和可測量化

第4章測量信度和效度

第5章觀察、判斷和復合變量

第6章問卷法、訪談法和日志法

第三編 研究設計的邏輯

第7章隨機控制的實驗和因果推論

第8章非隨機化研究與函數關繫

第9章隨機與非隨機抽取樣本

第四編 數據分析的基本原理

第10章描述、展示以及探究數據

第11章相關關繫

第12章統計檢驗力與效應值

第五編 單因素設計

第13章用t 檢驗比較均值

第14章方差分析與F 檢驗

第15章單因素對比分析

第六編 因素設計

第16章多因素方差分析

第17章方差分析中的交互作用

第18章重復測量方差分析

第七編 數據分析的補充問題

第19章顯著性檢驗和列聯表分析

第20數據分析

第2分析：研究結果的比較和綜合

第八編 附錄

詳細目錄

PART I CONCEPTUAL AND ETHICALFOUNDATIONS 1

Chapter 1 The Spirit of Behavioral Research 3

Science and the Search for Knowledge 3

What Do Behavioral Researchers Really Know? 6

Social Constructionism 7

Contextualism/Perspectivism 9

Evolutionary Epistemology 12

Peirce’s Four Ways of Knowing 13

Rhetoric, Perceptibility, and Aesthetics 15

Limitations of the Four Supports of Conviction 18

Behavioral Research Defi ned 18

Three Broad Research Orientations 21

The Descriptive Research Orientation 23

The Relational Research Orientation 25

The Experimental Research Orientation 29

Empirical Principles as Probabilistic Assertions 32

Orienting Habits of Good Scientifi c Practice 34

Chapter 2 Contexts of Discovery and Justifi cation 37

Inspiration and Explanation 37

Theories and Hypotheses 38

Sources of Inspiration and Insight 39

Serendipity in Behavioral Research 42

Molding Ideas Into Working Hypotheses 43

Positivism, Falsifi cationism, and Conventionalism 48

Type I and Type II Decision Errors 53

Statistical Signifi cance and the Effect Size 55

Two Families of Effect Sizes 56

Interval Estimates Around Effect Sizes 58

Summing Up 59

Chapter 3 Ethical Considerations, Dilemmas,and Guidelines 61

Puzzles and Problems 61

A Delicate Balancing Act 63

Historical Context of the American Psychological Association Code
64

The Belmont Report, Federal Regulations, and the
Institutional

Review Board 67

Principle I: Respect for Persons and Their Autonomy 69

Principle II: Benefi cence and Nonmalefi cence 71

Principle III: Justice 74

Principle IV: Trust 76

Principle V: Fidelity and Scientifi c Integrity 77

Costs, Utilities, and Institutional Review Boards 79

Scientifi c and Societal Responsibilities 82

PART II OPERATIONALIZATION AND MEASUREMENTOF DEPENDENT VARIABLES
85

Chapter 4 Reliability and Validity of Measurements 87

Random and Systematic Error 87

Assessing Stability and Equivalence 89

Internal-Consistency Reliability and Spearman-Brown 92

KR20 and Cronbach’s Alpha 94

Effective Reliability of Judges 98

Effective Cost of Judges 100

Effective Cost of Items 102

Interrater Agreement and Reliability 103

Cohen’s Kappa 105

Replication in Research 111

Validity Criteria in Assessment 113

Convergent and Discriminant Validity 115

Test Validity, Practical Utility, and the Taylor-Russell Tables
117

Relationship of Validity to Reliability 119

Chapter 5 Observations, Judgments, andComposite Variables
123

Observing, Classifying, and Evaluating 123

Observing While Participating 124

 Maximizing Credibility and Serendipity 125

Organizing and Sense-Making in Ethnographic Research 127

Interpreter and Observer Biases 128

Unobtrusive Observations and Nonreactive Measurements 130

Selecting the Most Appropriate Judges 134

Choosing the Number of Response Alternatives 137

Effects of Guessing and Omissions on Accuracy 138

Intrinsic Factors and the Level of Accuracy 140

Applications of Categorical Judgments 141

Category Scales and Rating Scales 145

Numerical, Graphic, and Magnitude Ratings 146

Rating Biases and Their Control 150

Bipolar Versus Unipolar Scales 151

Forming Composite Variables 151

Forming Multiple Composites 154

Quantifying the Clarity of Composites 156

Chapter 6 Questionnaires, Interviews, and Diaries 160

Concerns About Self-Report Data 160

Open-Ended Versus Structured Items 163

Critical Incident Technique 165

Stages in Developing Interview Protocols 167

Research Interviews by Telephone 171

Developing Research Questionnaires 172

Defensiveness, Inconsistency, and Yea-Saying 174

Cross-Cultural Questionnaire and Interview Research 176

One-Dimensional and Multidimensional Attitude Scales 177

Semantic Differentials for Attitudinal Meaning 178

Q-Sorts for Subjectivity Ratings 179

Likert Method of Item Analysis 181

Thurstone Equal-Appearing Intervals Method 182

Memory and the Use of Self-Recorded Diaries 185

PART III THE LOGIC OF RESEARCH DESIGNS 187

Chapter 7 Randomized Controlled Experiments andCausal Inference
189

Experimentation in Science 189

Randomized Experimental Designs 190

Characteristics of Randomization 193

The Philosophical Puzzle of Causality 196

Contiguity, Priority, and Constant Conjunction 198

Four Types of Experimental Control 200

Mill’s Methods of Agreement and Difference 201

Between-Group Designs and Mill’s Joint Method 203

Independent, Dependent, and Moderator Variables 204

Solomon’s Extended Control Group Design 206

Threats to Internal Validity 209

Threats to External Validity 212

Statistical Conclusion and Construct Validity 215

Subject and Experimenter Artifacts 216

Demand Characteristics and Their Control 220

Interactional Experimenter Effects 223

Experimenter Expectancy Effects and

Their Control 226

Concluding Commentary 230

Chapter 8 Nonrandomized Research and FunctionalRelationships
233

Nonrandomized and Quasi-Experimental Studies 233

Nonequivalent Groups and Historical Controls 235

Interrupted Time Series and the Autoregressive IntegratedMoving
Average 238

Single-Case Experimental Designs 239

Cross-Lagged Correlational Designs 242

Invisible Variables and the Mediation Problem 245

Path Analysis and Causal Inference 246

The Cohort in Longitudinal Research 250

Different Forms of Cohort Studies 252

Subclassifi cation on Propensity Scores 256

Multiple Confounding Covariates 257

Chapter 9 Randomly and Nonrandomly SelectedSampling Units
260

Sampling a Small Part of the Whole 260

Bias and Instability in Surveys 262

Simple Random-Sampling Plans 264

Improving Accuracy in Random Sampling 266

Confi dence Intervals for Population Estimates 269

Speaking of Confi dence Intervals 270

Other Selection Procedures 271

Nonresponse Bias and Its Control 273

Studying the Volunteer Subject 276

Characteristics of the Volunteer Subject 278

Implications for the Interpretation of

Research Findings 284

Situational Correlates and the Reduction of

Volunteer Bias 285

The Problem of Missing Data 288

Procedures for Dealing With Missing Data 289

PART IV FUNDAMENTALS OF DATA ANALYSIS 291

Chapter 10 Describing, Displaying, and Exploring Data 293

Descriptions of Sampling Units 293

Frequency Diagrams and Stem-and-Leaf Displays 294

Box Plots 297

Comparing Distributions Back to Back 298

Measures of Central Tendency 299

Measures of Spread 300

The Normal Distribution 304

Standard Scores 305

Data Not Distributed Normally 306

Precision of Estimating Population Means 307

Defi ning Outliers 309

Coping With Outliers 310

Exploring the Data 311

Chapter 11 Correlation 314

Pearson r 314

Proportion of Variance Interpretation of Correlation 316

Binomial Effect-Size Display 318

Confi dence Intervals for Effect-Size Correlations 322

Small Correlations, But Important Effects 324

Counternull Values of Effect Sizes 328

Spearman Rank Correlation 330

Ranks as a Transformation 332

Observations of Disproportionate Infl uence 333

Point-Biserial Correlation 336

Exact Tests for Rho 339

Phi Coeffi cient 340

Curvilinear (Quadratic) Correlation 344

Five Product-Moment Correlations 346

Comparing Correlations 347

Considering Third Variables 347

Effects of Variability on Correlations 349

Chapter 12 Statistical Power and Effect Size Revisited 354

Why Assess Statistical Power? 354

The Neglect of Statistical Power 357

The requivalent Statistic 359

Cohen’s Multipurpose Power Tables 360

The t Test for Comparing Two Means 363

The Signifi cance of a Product-Moment r 366

Differences Between Correlation Coeffi cients 369

The Test That a Proportion is .50 373

The Difference Between Proportions 373

The Focused Chi-Square Test 374

F Tests for Focused Comparisons 375

Additional Strategies for Improving Power 376

PART V ONE-WAY DESIGNS 379

Chapter 13 Comparing Means by Standard t Tests 381

Gosset and the t Test 381

Two Components of t Tests 382

Maximizing t 383

Effect Sizes and Adjustments for Unequal Sample Sizes 385

Interpreting the Independent Sample t 388

Computing the Independent Sample t 391

Reporting the Results 392

t Tests for Nonindependent Samples 395

Effect Size and Study Size Components of Nonindependent Sample t
397

Assumptions Underlying t Tests 401

Nonparametric Procedures 403

The Bootstrap, the Jackknife, and Permutation Tests 405

Chapter 14 Analysis of Variance and the F Test 409

The F Test and the t Test 409

The Analysis of “Variances” 410

Illustration of an Omnibus F 412

Dividing Up the Total Variance 413

ANOVA Summary Tables 414

Distributions of F 417

After the Omnibus F 418

Protecting Against “Too Many t Tests” 421

Bonferroni Procedures 422

Bonferroni Tolerance Value 425

Comparing Two Independent Variabilities 425

Illustration Using Transformations 426

Comparing Two Correlated Variabilities 429

Comparing Three or More Independent Variabilities 431

Comparing Three or More Correlated Variabilities 433

Summary of Procedures for Comparing Variabilities 433

Chapter 15 One-Way Contrast Analyses 434

Focusing Our Questions and Statistical Tests 434

Contrast F Tests on Original Data 436

Contrast t Tests on Original Data 440 Ca