Need to learn statistics as part of your job, or want some help passing a statistics course? Statistics in a Nutshell is a clear and concise introduction and reference that's perfect for anyone with no previous background in the subject. This book gives you a solid understanding of statistics without being too simple, yet without the numbing complexity of most college texts.
You get a firm grasp of the fundamentals and a hands-on understanding of how to apply them before moving on to the more advanced material that follows. Each chapter presents you with easy-to-follow descriptions illustrated by graphics, formulas, and plenty of solved examples. Before you know it, you'll learn to apply statistical reasoning and statistical techniques, from basic concepts of probability and hypothesis testing to multivariate analysis.
Organized into four distinct sections, Statistics in a Nutshell offers you:
Different ways to think about statistics
Basic concepts of measurement and probability theory
Data management for statistical analysis
Research design and experimental design
How to critique statistics presented by others
Basic inferential statistics:
Basic concepts of inferential statistics
The concept of correlation, when it is and is not an appropriate measure of association
Dichotomous and categorical data
The distinction between parametric and nonparametric statistics
Advanced inferential techniques:
The General Linear Model
Analysis of Variance (ANOVA) and MANOVA
Multiple linear regression
Business and quality improvement statistics
Medical and public health statistics
Educational and psychological statistics
Unlike many introductory books on the subject, Statistics in a Nutshell doesn't omit important material in an effort to dumb it down. And this book is far more practical than most college texts, which tend to over-emphasize calculation without teaching you when and how to apply different statistical tests.
With Statistics in a Nutshell, you learn how to perform most common statistical analyses, and understand statistical techniques presented in research articles. If you need to know how to use a wide range of statistical techniques without getting in over your head, this is the book you want.
Chapter 1 Basic Concepts of Measurement
Levels of Measurement
True and Error Scores
Reliability and Validity
Chapter 2 Probability
Enough Exposition, Let’s Do Some Statistics!
Chapter 3 Data Management
An Approach, Not a Set of Recipes
The Chain of Command
The Rectangular Data File
Spreadsheets and Relational Databases
Inspecting a New Data File
String and Numeric Data
Chapter 4 Descriptive Statistics and Graphics
Populations and Samples
Measures of Central Tendency
Measures of Dispersion
Chapter 5 Research Design
Gathering Experimental Data
Inference and Threats to Validity
Example Experimental Design
Chapter 6 Critiquing Statistics Presented by Others
The Misuse of Statistics
Chapter 7 Inferential Statistics
Independent and Dependent Variables
Populations and Samples
The Central Limit Theorem
Chapter 8 Thet-Test
The t Distribution
Repeated Measures t-Test
Unequal Variance t-Test
Effect Size and Power
Chapter 9 The Correlation Coefficient
Graphing Associations Through Scatterplots
Pearson’s Product-Moment Correlation Coefficient
Coefficient of Determination
Spearman Rank-Order Coefficient
Chapter 10 Categorical Data
The R×C Table
The Chi-Square Distribution
The Chi-Square Test
Fisher’s Exact Test
McNemar’s Test for Matched Pairs
Correlation Statistics for Categorical Data
The Likert and Semantic Differential Scales
Chapter 11 Nonparametric Statistics
Between Subjects Designs
Chapter 12 Introduction to the General Linear Model
The General Linear Model
Analysis of Variance (ANOVA)
Chapter 13 Extensions of Analysis of Variance
Repeated Measures ANOVA
Chapter 14 Multiple Linear Regression
Multiple Regression Models
Common Problems with Multiple Regression
Chapter 15 Other Types of Regression
Chapter 16 Other Statistical Techniques
Discriminant Function Analysis
Chapter 17 Business and Quality Improvement Statistics
Chapter 18 Medical and Epidemiological Statistics
Measures of Disease Frequency
Ratio, Proportion, and Rate
Prevalence and Incidence
Crude, Category-Specific, and Standardized Rates
The Risk Ratio
The Odds Ratio
Confounding, Stratified Analysis, and the Mantel-Haenszel Common Odds Ratio
Sample Size Calculations
Chapter 19 Educational and Psychological Statistics
Sarah Boslaugh holds a PhD in Research and Evaluation from the City University of New York and have been working as a statistical analyst for 15 years, in a variety of professional settings, including the New York City Board of Education, the Institutional Research Office of the City University of New York, Montefiore Medical Center, the Virginia Department of Social Services, Magellan Health Services, Washington University School of Medicine, and BJC HealthCare. She has taught statistics in several different contexts and currently teaches Intermediate Statistics at Washington University Medical School. She has published two previous books: An Intermediate Guide to SPSS Programming: Using Syntax for Data Management (SAGE Publications, 2004) and Secondary Data Sources for Public Health (forthcoming from Cambridge U. Press, 2007) and am currently editing the Encyclopedia of Epidemiology for SAGE Publications (forthcoming, 2007).
Paul A. Watters PhD CITP, is Associate Professor in the School of Information and Mathematical Sciences and Centre for Informatics and Applied Optimization (CIAO) at the University of Ballarat. Until recently, he was Head of Data Services at the Medical Research Council's National Survey of Health and Development, which is the oldest of the British birth cohort studies, and an honorary senior research fellow at University College London. He uses multivariate statistics to develop orthogonal and non-orthogonal methods for feature extraction in pattern recognition, especially in biometric applications.
The animal on the cover of Statistics in a Nutshell is a thornback crab, also knownas a spiny spider crab (Maja squinado, Maja brachydactyla). Found in the northeast Atlantic Ocean and the Mediterranean Sea, the thornback crab is the largestof the European crabs, with a carapace diameter of two to seven inches. It is easily identifiable by the two hornlike spikes between its eyes, and the six or so smaller spikes that extend from each side of its shell. The thornback's body is reddish, with pink, brown, or yellow markings, and its surface is also covered with small spikes, as the crab's name implies.
Thornback crabs are occasionally found on the shore, but they prefer depths of 90 to 600 feet. They are solitary animals except during mating season, when they form large breeding mounds. In years when their numbers are particularly abundant, they can be a source of frustration for lobster fisherman, as they infest the lobster pots. Thornbacks are themselves fished for their delicious claw meat.
Male thornbacks are effective predators; their delicate-looking claws are actually quite powerful and can open small mussels to feed on them. Their claws are also double-jointed, so although it is generally safe for a person to hold crustaceans by each side of their shells, thornbacks are able to reach over their backs to pinch the offender. Females have smaller, less flexible claws and are thus more vulnerable to attack. To defend against their predators-which include lobsters, wrasses, and cuttlefish-many species of spider crabs decorate their spiny shells with seaweed, sponges, or aquatic debris to better blend in against the seabed.
The cover image is from Lydekker's Library of Natural History. The cover font is Adobe ITC Garamond. The text font is Linotype Birka; the heading font is Adobe Myriad Condensed; and the code font is LucasFont's TheSansMonoCondensed.
Wanting to firm up my knowledge of statistics, I picked up a book titled "Statistics in a Nutshell: A Desktop Quick Reference" by Sarah Boslaugh and Paul Andrew Watters (ISBN: 978-0-596-51049-7).
I jumped right to Chapter 9: The Correlation Coefficient to check on the exact definition of this concept. On page 171 is the explanation ==QUOTE== In the general form of the model y = +/-ax +/- b: * y is the dependent variable * +/-a is the _slope_ (i.e., where the value of y when x = 0); -a indicates a negative association; +a indicates a positive association * x is the independent variable * +/-b is the _intercept_ of the straight line ==UNQUOTE== That parenthesized clause lacks a verb "where the value of y [is]", and it describes not the slope but the intercept; it should appear after the words "straight line".
And on the very same and the next page is the wording ==QUOTE== -- the first values plotted are (1, -2), (2, -4), and (3, -6). The model for this relationship is y = -2x, which is multiplicative in nature. Using the linear model, a = -2, which is positive, so the relationship is positive. Again, b = 0, since the intercept is the origin (0,0). ==UNQUOTE== Huh? -2 is positive? And in these two quotes, the word "intercept" is being used in two mutually incompatible ways: in the first quote, the intercept is a number; in the second quote, the intercept is a point on the xy plane. Moreover, the word "intercept" is not listed in the index (p. 447).
So here's a nice statistics problem for ya: Having found four obvious errors in a random sample of two pages of a 453-page book, what degree of confidence should we assign to whatever this book says about a given topic?
What is going on here? Is this some kind of joke? Did no editor read the text of this college textbook?
Page xx of this book has a section titled "We'd Like to Hear from You", which says, "To comment or ask technical questions about this book, send email to: firstname.lastname@example.org ". It also says: "We have a web page for this book, where we list errata, examples, and any additional information. You can access this page at: http://www.oreilly.com/catalog/9780596510497 " (Get it? That number is the book's ISBN number). But if you go there, you instead are taken to http://shop.oreilly.com/product/9780596510497.do , which, as far as I can see, lists no errata at all for a book that was released over four years ago, in July 2008.
I wanted to send a copy of this note to the two authors too, but I can't find their e-mail address in the book or on the O'Reilly page. Let's try an online search... http://www.ofcs.org/sarah-boslaugh/ says that Sarah Boslaugh has given up biostatistics in favor of being a movie critic(!). And a search on Paul Andrew Watters leads to http://dblp.uni-trier.de/pers/hd/w/Watters:Paul_Andrew.html , which is a page of the Universitaet Trier, the town of a cousin of my mother's (kleine Welt! Hallo, Porta Nigra!). But still I can't find the authors' e-mail addresses. If anybody can find them, feel free to send them this note.
The surnames Boslaugh and Watters are interesting too. I can't figure out whether to say "BOSS laff" or "BOSS law", or whether "Watters" rhymes with "waters" or with "matters".
-- Mark Spahn (West Seneca, NY)
Bottom Line No, I would not recommend this to a friend
Go to Page and 157, first paragraph. The population mean is 1.708<mu<2.292. Is this right?
Bottom Line Yes, I would recommend this to a friend
Merchant response: Thanks for writing about this. Yes, the author has confirmed that this is an error. You can find confirmed errata and submit errors you find at http://oreilly.com/catalog/errata.csp?isbn=9780596510497
If you have any questions, feel free to email our book technical support at email@example.com.
The book makes it difficult to grasp new concepts because of the errors. Reading requires two-hands: one for the highlighter and the other for the Errata sheet. The Errata sheet has 12 pages, so you can't fold it and slip it into the book. Additionally, if you print the Errata sheet, the formatting is messed up. Text are on top of other text, making it unreadable in certain spots.
I've been reading through the book as a quick review of basic statistics while trying to learn some statistics computing packages. So far, the book has been riddled with computational errors, basic errors for variable definitions, and figures that say they will show something they don't. For example, trying to define the binomial distribution, the authors confuse the variables for trials and successes in a Bernoulli process. The authors do not explain how to really calculate a t-test. For example, the authors provide a value for the mean in an example of the t-test, and then use a different mean later in the example for calculating confidence intervals without actually defining what numbers should be going into the confidence interval calculation. Not at all proof read, as another reviewer has stated.
Without a reasonable strong statistics background for a masters degree in epidemiology, I would be left confused by a great deal of the explanations in this book. Not a tried and true reference...
Bottom Line No, I would not recommend this to a friend
I bought this hoping for the one 'magic' stats book that explained statistics in plain english. Unfortunately this is not the one. I concur that this book is riddled with errors, and it's written not so much as a reference book, but as an explanatory text. Very few worked examples. Bummer.
Bottom Line No, I would not recommend this to a friend
I had high hopes for this book. The world really needs a reference guide for those of us who have had exposure to statistics in college, but often don't remember which tool to turn to (t-test? F-test? Chi-square? a non-parametric test? I don't remember which to use when!). It would have been perfect to provide handy reference tables for which tools to look into in various circumstances, along with the assumptions that must be met to use that tool and a reference to a page number where the tool is briefly explained with some examples. I would have been happy to use this to get me started and then google for more detailed information.
This isn't that book. There are no handy reference tables (of any sort) in this book.
Despite the "Nutshell" title, this book is written as if it is a linear, beginner primer on statistics.
I could be OK with that if it were a good beginner primer on statistics. Instead, the authors seem to assume that "Nutshell" in the title is a license to cut corners in their explanations, leaving the beginning reader confused.
The worst part is that the book is riddled with errors. If you're trying to follow the examples and learn from them, you're never sure if you just don't understand or if this is yet another typo/error. Chapter 8 on t-test is so full of computation errors and typos that I doubt it was ever proof-read.
While many of the "Nutshell" books assume you already know the subject and simply need a reference book, Statistics In A Nutshell takes a decidedly different approach. While this book assumes you understand basic math concepts, it does not assume you have any prior background in statistics. It then proceeds to cover nearly every fundamental concept taught during an introductory statistics course.
While many introductory statistics courses go through basic descriptive statistics and move through more advanced concepts like ANOVA or liner regression, I found that this book also covered such concepts as non-parametric tests, design of experiments, and the general linear model. Certainly, these concepts are not covered in as much depth as say a college-level courses dedicated to non-parametric statistics. However, the concepts are there, and the authors provide enough information to make the discussion valuable.
The last few sections of the book discuss the use of statistics in a variety of professions, including manufacturing, business, medical, and education fields. Individuals who may not feel comfortable with their math skills can take comfort in the fact that the book provides a section on basic math skills, however, those who may be mathematically challenged may argue with the term "basic".
All in all, I think this is an excellent book for individuals who are looking to implement the scientific method and statistics in their business. The author provides sound explanations of the concepts and plenty of figures and tables to explain difficult concepts. I'd highly recommend this book for the reader who is not afraid of a little math and wants to understand statistics and statistical concepts.