I got the following (unsolicited) email from a publisher today:
We are developing a new, introductory statistics textbook with a data analysis approach, and would value your answers to our brief survey regarding the proposed table of contents (attached). . . .
Having read the table of contents (see below), all I can say is . . . yuck! It’s gotta be tough being a book publisher if you’re expected to be coming up with new intro texts all the time.
OK, here it is:
Chapter 1 Introduction
1.1. Variables, parameters, and statistics
1.2. Experiments and Observational Studies
1.3. Surveys
1.4. Basic Probability and odds
Chapter 2 Summarizing Data – Quantitative Variables
2.1 Numerical Summaries of Center
2.2 Numerical Summaries of Variation
2.3 Graphical Summaries
2.4 The Normal Distribution
2.5 Scatterplots/Regression
Chapter 3 Summarizing Data – One Categorical Variable
3.1 Numerical and Graphical Summary
3.2 Random Variables and Expectation
3.3 Binomial Distribution
3.4 Chi-Square test for a single variable
Cumulative Test and Data Analysis Projects
Chapter 4 Summarizing Data – Two Categorical Variables
4.1 Two Way Contingency Tables
4.2 Chi-Square Test of Independence
4.3 Relative Risk and Odds ratios
Chapter 5 Estimating and Testing Proportions
5.1 Discovering the Sampling Distribution of the Sample Proportion
5.2 Estimating with Confidence
5.3 Hypothesis Testing a Single Proportion
5.4 Estimating the Difference and Testing Hypotheses for Two Population Proportions
5.5 Determining the Sample Size and Power for Proportions
Cumulative Test and Data Analysis Projects
Chapter 6 Estimating and Testing one and two sample means
6.1 Sampling variability with a mean
6.2 Estimation of a mean with confidence
6.3 Testing a hypothesis concerning a mean
6.4 Paired estimation and testing for two dependent samples
6.5 Estimating and testing two means for two independent samples
Chapter 7 Correlation and Regression
7.1 Correlation
7.2 Estimation and Inference for slope
7.3 Residuals
7.4 Coefficient of determination
7.5 Spearman Rank Correlation
Cumulative Test and Data Analysis Projects
Chapter 8 Analysis of Variance
8.1 One Way ANOVA
8.2 Multiple Comparison Procedures
8.3 Blocked/Repeated Measures ANOVA
8.4 Two Way ANOVA
8.5 Special Cases
Chapter 9 Nonparametric Statistics
9.1 Wilcoxon Rank Sum Test (independent samples)
9.2 Wilcoxon Signed Rank Test (dependent samples)
9.3 Kruskal Wallis Test
9.4 Friedman Rank Test
9.5 Spearman Rank Correlation
Cumulative Test and Group Project
Chapter 10 Multiple and Logistic Regression
10.1 Quadratic Regression
10.2 Categorical Predictor with continuous predictor
10.3 Interaction
10.4 Dummy variables
10.5 Variable Selection
10.6 Logistic Regression
Chapter 11 Time Series Analysis and Control Charts
11.1 Time Series Components
11.2 Smoothing Techniques
11.3 Trend and Seasonal Effects
11.4 Forecasting
1. 11.5 and S Charts
11.5 p chart
11.6 R charts
Cumulative Test and Data Analysis Projects
In all seriousness, there’s nothing wrong with most of these topics, but is there really a need for one more book presenting this sort of grab-bag of methods? I’d think there are a zillion existing books that would do the job quite well already.
Yes, there is, because the author isn't getting any royalties from such a book and wants a share of the mandatory undergrad stats class pie! :)
Freshman calculus texts don't need to be redone either. it's a scam.
It's because they sell. An awful lot of people buy an introductory statistics textbook. So the market is much larger than for other texts. A small percentage of that market is better than a much larger percentage of a smaller market.
Also, when they make it big, they make it really big. I've heard (from someone who should know) that the publisher Sage's best selling text in the UK is Andy Field's "Discovering statistics using SPSS". Type SPSS into the search at amazon.co.uk, and you get 808 books back, many of which have similar titles:
The top seller is Andy Field's, with a sales rank of 1,123. Second is 11,000th, 3rd is 13,000th, and so on.
Try Bayes, and the top seller is in 34,000th place, Bayesian, and the top seller is in 92,000th place. Regression (and ignore all the past life stuff) and it's 30,000th. Multivariate and it's 26,000th. Generalized linear is 86,000th, and so on.