Quantitative Data Analysis: Doing Social Research to Test Ideas
Author: 
Donald J. Treiman 
Publisher: 
JosseyBass (a Wiley imprint) 
Copyright: 
2009 
ISBN13: 
9780470380031 
Pages: 
480; paperback 
Price: 
$54.50 



Comment from the Stata technical group
Quantitative Data Analysis, by Donald J. Treiman, is a wellwritten
demonstration of how to answer questions using statistics. While the preface
states that the book is “designed for a course to be taken after a
firstyear graduate statistics course in the social sciences”, the
thought processes and techniques illustrated are useful and interesting to a
much wider audience. The range of techniques is broad, ranging from simple
advice for making tables readily readable through linear and logistic
regression to loglinear and randomeffects models.
Treiman writes using clear, precise language. This style makes the book
accessible to readers from many fields and especially worthwhile to
statistical consultants or others who work with clients of different
backgrounds. The book is highly applied—the examples stem from
published papers with real datasets. Many of the examples stem from
Treiman’s long history of research in applied sociology, yet these
examples are still interesting and approachable to those outside the field.
The main material is nicely supplemented with “callouts”
containing biographical and historical background information, as well as
tips on Stata usage. Treiman also takes the time and effort to explain how
to avoid common pitfalls of data analysis.
Because this is an applied book, there is little derivation of the
mathematics behind the statistical techniques. This is not a drawback,
though, because Treiman includes references and a large bibliography, which
can be followed by those curious about statistical theory. Stata was used
for the computation of all statistical results, and all the Stata dofiles
(Stata code) and datasets are available from the web. (In the book itself
there is advice about how to use Stata for some analyses, but there is no
explicit Stata code.)
Quantitative Data Analysis is worth a look for those wanting to see
the applications of a wide variety of statistical techniques to a variety of
problems or for those who are interested in the thought process behind
assessing the results of techniques.
Table of contents
Tables, Figures, Exhibits, and Boxes
Preface
The Author
Introduction
1 CrossTabulations
What This Chapter Is About
Introduction to the Book via a Concrete Example
CrossTabulations
What This Chapter Has Shown
2 More on Tables
What This Chapter Is About
The Logic of Elaboration
Suppressor Variables
Additive and Interaction Effects
Direct Standardization
A Final Note on Statistical Controls Versus Experiments
What This Chapter Has Shown
3 Still More on Tables
What This Chapter is About
Reorganizing Tables to Extract New Information
When to Percentage a Table “Backwards”
CrossTabulations in Which the Dependent Variable is Represented by a Mean
Index of Dissimilarity
Writing About CrossTabulations
What This Chapter Has Shown
4 On the Manipulation of Data by Computer
What This Chapter Is About
Introduction
How Data Files Are Organized
Transforming Data
What This Chapter Has Shown
Appendix 4.A Doing Analysis Using Stata
Tips on Doing Analysis Using Stata
Some Particularly Useful Stata 10.0 Commands
5 Introduction to Correlation and Regression (Ordinary Least Squares)
What This Chapter Is About
Introduction
Quantifying the Size of a Relationship: Regression Analysis
Assessing the Strength of a Relationship: Correlation Analysis
The Relationship Between Correlation and Regression Coefficients
Factors Affecting the Size of Correlation (and Regression) Coefficients
Correlation Ratios
What This Chapter Has Shown
6 Introduction to Multiple Correlation and Regression (Ordinary Least Squares)
What This Chapter is About
Introduction
A Worked Example: The Determinants of Literacy in China
Dummy Variables
A Strategy for Comparisons Across Groups
A Bayesian Alternative for Comparing Models
Independent Validation
What This Chapter Has Shown
7 Multiple Regression Tricks: Techniques for Handling Special Analytic Problems
What This Chapter Is About
Nonlinear Transformations
Testing the Equality of Coefficients
Trend Analysis: Testing the Assumption of Linearity
Linear Splines
Expressing Coefficients as Deviations from the Grand Mean (Multiple Classification Analysis)
Other Ways of Representing Dummy Variables
Decomposing the Difference Between Two Means
What This Chapter Has Shown
8 Multiple Imputation of Missing Data
What This Chapter Is About
Introduction
A Worked Example: The Effect of Cultural Capital on Educational Attainment in Russia
What This Chapter Has Shown
9 Sample Design and Survey Estimation
What This Chapter Is About
Survey Samples
Conclusion
What This Chapter Has Shown
10 Regression Diagnostics
What This Chapter Is About
Introduction
A Worked Example: Societal Differences in Status Attainment
Robust Regression
Bootstrapping and Standard Errors
What This Chapter Has Shown
11 Scale Construction
What This Chapter Is About
Introduction
Validity
Reliability
Scale Construction
ErrorsinVariables Regression
What This Chapter Has Shown
12 LogLinear Analysis
What This Chapter Is About
Introduction
Choosing a Preferred Model
Parsimonious Models
A Bibliographic Note
What This Chapter Has Shown
Appendix 12.A Derivation of the Effect Parameters
Appendix 12.B Introduction to Maximum Likelihood Estimation
Mean of a Normal Distribution
LogLinear Parameters
13 Binomial Logistic Regression
What This Chapter Is About
Introduction
Relation to a LogLinear Analysis
A Worked Logistic Regression Example: Predicting Prevalence of Armed Threats
A Second Worked Example: Schooling Progression Ratios in Japan
A Third Worked Example (DiscreteTime HazardRate Models): Age at First Marriage
A Fourth Worked Example (Case–Control Models): Who Was Appointed to a Nomenklatura Position in Russia?
What This Chapter Has Shown
Appendix 13.A Some Algebra for Logs and Exponents
Appendix 13.B Introduction to Probit Analysis
14 Multinomial and Ordinal Logistic Regression and Tobit Regression
What This Chapter Is About
Multinomial Logit Analysis
Ordinal Logistic Regression
Tobit Regression (and Allied Procedures) for Censored Dependent Variables
Other Models for the Analysis of Limited Dependent Variables
What This Chapter Has Shown
15 Improving Causal Inference: Fixed Effects and Random Effects Modeling
What This Chapter Is About
Introduction
Fixed Effects Models for Continuous Variables
Random Effects Models for Continuous Variables
A Worked Example: The Determinants of Income in China
Fixed Effects Models for Binary Outcomes
A Bibliographic Note
What This Chapter Has Shown
16 Final Thoughts and Future Directions: Research Design and Interpretation Issues
What This Chapter Is About
Research Design Issues
The Importance of Probability Sampling
A Final Note: Good Professional Practice
What This Chapter Has Shown
Appendix A: Data Descriptions and Download Locations for the Data Used in This Book
Appendix B: Survey Estimation with the General Social Survey
References
Index