Synopses & Reviews
Praise for the First Edition of Common Errors in Statistics" . . . let me recommend Common Errors to all those who interact with statistics, whatever their level of statistical understanding . . . "
—Stats 40
" . . . written . . . for the people who define good practice rather than seek to emulate it."
—Journal of Biopharmaceutical Statistics
" . . . highly informative, enjoyable to read, and of potential use to a broad audience. It is a book that should be on the reference shelf of many statisticians and researchers."
—The American Statistician
" . . . I found this book the most easily readable statistics book ever. The credit for this certainly goes to Phillip Good."
—E-STREAMS
A tried-and-true guide to the proper application of statistics
Now in a second edition, the highly readable Common Errors in Statistics (and How to Avoid Them) lays a mathematically rigorous and readily accessible foundation for understanding statistical procedures, problems, and solutions. This handy field guide analyzes common mistakes, debunks popular myths, and helps readers to choose the best and most effective statistical technique for each of their tasks.
Written for both the newly minted academic and the professional who uses statistics in their work, the book covers creating a research plan, formulating a hypothesis, specifying sample size, checking assumptions, interpreting p-values and confidence intervals, building a model, data mining, Bayes' Theorem, the bootstrap, and many other topics. The Second Edition has been extensively revised to include:
- Additional charts and graphs
- Two new chapters, Interpreting Reports and Which Regression Method?
- New sections on practical versus statistical significance and nonuniqueness in multivariate regression
- Added material from the authors' online courses at statistics.com
- New material on unbalanced designs, report interpretation, and alternative modeling methods
With a final emphasis on both finding solutions and the great value of statistics when applied in the proper context, this book is eminently useful to students and professionals in the fields of research, industry, medicine, and government.
Review
“In summary, I think the book does achieve its mission to highlight common errors in statistical analysis.” (
Significance, 1 December 2006)
"I regard this interesting book to be of potential use to a broad audience." (Statistical Papers, August 2007)
"A very engaging and valuable book for all who use statistics in any settings." (CHOICE, October 2006)
"...a concise guide to the basics of statistics, replete with examples, explaining in common language...a valuable reference for more advanced statisticians as well..." (MAA Reviews, June 22, 2006)
“All statistics students and teachers will find in this book a friendly and intelligent guide to…applied statistics in practice.” (Journal Of Applied Statistics, April 2007)
Review
“In summary, I think the book does achieve its mission to highlight common errors in statistical analysis.” (
Significance, 1 December 2006)
"I regard this interesting book to be of potential use to a broad audience." (Statistical Papers, August 2007)
"A very engaging and valuable book for all who use statistics in any settings." (CHOICE, October 2006)
"...a concise guide to the basics of statistics, replete with examples, explaining in common language...a valuable reference for more advanced statisticians as well..." (MAA Reviews, June 22, 2006)
“All statistics students and teachers will find in this book a friendly and intelligent guide to…applied statistics in practice.” (Journal Of Applied Statistics, April 2007)
Synopsis
With a final emphasis on both finding solutions and the great value of statistics when applied in the proper context, this book is eminently useful to students and professionals in the fields of research, industry, medicine, and government.
Synopsis
The previous editions have proven to be successful guides for choosing and using the right techniques. Common Errorsis consistently coherent and provides a consistent level throughout. The Third Editionelaborates on many key topics such as response variables, errors in testing hypothesis, higher order experimental design, curve fitting and magic beans, Poisson and negative binomial regression, correcting for confounding variables, dynamic models, factor analysis, general linear models, decision trees, etc. One new chapter has been added on "Statistical Analysis" and includes sections on data quality assessment, data review, and design review. Modifications have been included throughout the book, and many new figures have also been added. Topics covered include creating a research plan, collecting data, formulating and testing a hypothesis, estimating coefficients, specifying sample size, checking assumptions, interpreting p-values and confidence intervals, building a model, reporting results, data mining, Bayes' Theorem, the bootstrap, and many others.
Synopsis
Praise for the First Edition of Common Errors in Statistics
" . . . let me recommend Common Errors to all those who interact with statistics, whatever their level of statistical understanding . . . "
--Stats 40
" . . . written . . . for the people who define good practice rather than seek to emulate it."
--Journal of Biopharmaceutical Statistics
" . . . highly informative, enjoyable to read, and of potential use to a broad audience. It is a book that should be on the reference shelf of many statisticians and researchers."
--The American Statistician
" . . . I found this book the most easily readable statistics book ever. The credit for this certainly goes to Phillip Good."
--E-STREAMS
A tried-and-true guide to the proper application of statistics
Now in a second edition, the highly readable Common Errors in Statistics (and How to Avoid Them) lays a mathematically rigorous and readily accessible foundation for understanding statistical procedures, problems, and solutions. This handy field guide analyzes common mistakes, debunks popular myths, and helps readers to choose the best and most effective statistical technique for each of their tasks.
Written for both the newly minted academic and the professional who uses statistics in their work, the book covers creating a research plan, formulating a hypothesis, specifying sample size, checking assumptions, interpreting p-values and confidence intervals, building a model, data mining, Bayes' Theorem, the bootstrap, and many other topics. The Second Edition has been extensively revised to include:
* Additional charts and graphs
* Two new chapters, Interpreting Reports and Which Regression Method?
* New sections on practical versus statistical significance and nonuniqueness in multivariate regression
* Added material from the authors' online courses at statistics.com
* New material on unbalanced designs, report interpretation, and alternative modeling methods
With a final emphasis on both finding solutions and the great value of statistics when applied in the proper context, this book is eminently useful to students and professionals in the fields of research, industry, medicine, and government.
About the Author
PHILLIP I. GOOD, PhD, is Operations Manager of Information Research, a consulting firm specializing in statistical solutions for private and public organizations and has published eighteen books.
JAMES W. HARDIN, PhD, is Associate Research Professor in the Department of Epidemiology and Biostatistics at the University of South Carolina.
Table of Contents
Preface.
PART I: FOUNDATIONS.
1. Sources of Error.
1. Prescription.
2. Fundamental Concepts.
3. Ad-hoc, post-hoc hypotheses.
2. Hypotheses: The Why of Your Research.
1. Prescription.
2. What is a hypothesis?
3. How precise must a hypothesis be?
4. Found data.
5. Null hypothesis.
6. Neyman-Pearson theory.
7. Deduction and Induction.
8. Losses.
9. Decisions.
10. To Learn More.
3. Collecting Data.
1. Preparation.
2. Response Variables.
3. Determining Sample Size.
4. Fundamental Assumptions.
5. Experimental Design.
6. Four Guidelines.
7. Are Experiments Really Necessary?
8. To Learn More.
PART II: STATISTICAL ANALYSIS.
4. Data Quality Assessment.
1. GIGO.
2. Objectives.
3. Design Review.
4. Data Review.
5. Estimation.
1. Prevention.
2. Desirable and Not-so-desirable estimators.
3. Interval Estimates.
4. Improved Results.
5. Summary.
6. To Learn More.
6. Testing Hypotheses: Choosing a Test Statistic.
1. First Steps.
2. Test Assumptions.
3. Binomial Trials.
4. Categorical Data.
5. Time to Event Data (survival analysis).
6. Comparing Means of Two Sets of Measurements.
a. Multivariate comparisons.
b. Options.
c. Testing equivalence.
d. Unequal variances.
e. Dependent observations.
7. Comparing Variances.
8. Comparing the Means of K Samples.
9. Subjective Data.
10. Independence vs. Correlation.
11. Higher Order Experimental Designs.
f. Errors in interpretation.
g. Multi-factor designs.
h. Cross-over designs.
i. Factorial designs.
j. Unbalanced designs.
12. Inferior Tests.
13. Multiple Tests.
14. Summary.
15. To Learn More.
7. Miscellaneous Statistical Procedures.
1. Bootstrap.
2. Bayesian Methodology.
3. Meta-Analysis.
4. Permutation Tests.
5. To Learn More.
PART III: REPORTS.
8. Reporting Results.
1. Fundamentals.
a. Treatment Allocation.
b. Adequacy of Blinding.
c. Missing Data.
2. Descriptive Statistics.
d. Binomial trials.
e. Categorical data.
f. Rare events.
g. Measurements.
h. Which mean?
i. Ordinal data.
j. Tables.
k. Dispersion, precision, and accuracy.
3. Standard Error.
4. p-values.
5. Confidence Intervals.
6. Recognizing and Reporting Biases.
7. Reporting Power.
8. Drawing Conclusions.
9. Summary.
10. To Learn More.
9. Interpreting Reports.
1. With A Grain of Salt.
2. Rates and Percentages.
3. Interpreting Computer Printouts.
10. Graphics.
1. Five Rules for Avoiding Bad Graphics.
2. Displaying the 3rd Dimension.
3. The Misunderstood Pie Chart.
4. Effective Display of Subgroup Information.
5. Two Rules for Text Elements.
6. Choosing The Right Display.
7. To Learn More.
PART IV: BUILDING A MODEL.
11. Univariate Regression.
1. Model Selection.
a. Scope.
b. Ambiguous relationships.
c. Confounding variables.
2. Estimating Coefficients.
3. Further Considerations.
a. Bad data.
b. Practical v. statistical significance.
c. Goodness-of-fit v. prediction.
d. Indicator variables.
e. Transformations.
f. When a straight line won’t do.
g. Curve fitting and Magic Beans.
4. Summary.
5. Checklist.
6. To Learn More.
12. Alternate Modeling Methods.
1. LAD Regression.
2. Demming or EIV Regression.
3. Quantile Regression.
4. The Ecological Fallacy.
5. Poisson and Negative Binomial Regression.
6. Nonsense Regression.
7. Summary.
8. To Learn More.
13. Multivariate Regression.
1. Caveats.
2. Correcting for Confounding Variables.
3. Keep it Simple.
4. Dynamic Models.
5. Factor Analysis.
6. Reporting Your Results.
7. A Conjecture.
8. Decision Trees.
9. Building a Successful Model.
10. To Learn More.
14. Modeling Correlated Data.
2. Common Sources of Error.
3. Panel Data.
4. Fixed and Random Effects Models.
5. GEE’s.
a. Subject-specific or population averaged?
b. Variance estimation.
6. Quick Reference for Popular Panel Estimators.
7. To Learn More.
15. Validation.
1. Objectives.
2. Methods of Validation.
a. Independent verification.
b. Split sample.
c. Resampling.
3. Measures of Predictive Success.
4. Long Term Stability.
5. To Learn More.
Appendix A.
Appendix B.
Appendix C.
Glossary .
Bibliography.
Author Index.
Subject Index.