From Daren and Josh:
In "Classic" 6e (and 4e/5e), inference chapters are organized by the type of inference: confidence intervals for single samples (Chapter 8), significance tests for single samples (Chapter 9), and significance tests and confidence intervals for two samples (Chapter 10). In "Updated" 6e, inference chapters are organized by type of data, to match the CED: inference for proportions (Chapters 8–9) and inference for means (Chapters 10–11). In both books, chi-square tests and inference for slope wrap up the chapters on inference.
—

A number of  important structural changes were made  in the UPDATED sixth edition to ensure tight alignment with the AP® Statistics Course and Exam Description, including:  

• Transforming to achieve linearity moved from Section 12.2 to Section 3.3 in UPDATED TPS 6e, so Chapter 3 now aligns with College Board’s Unit 2: Exploring Two-Variable Data in the Course Framework.
• Sampling distribution of a difference in proportions moved from Section 10.1 to Section 7.2 in UPDATED TPS 6e; sampling distribution of a difference in means moved from Section 10.2 to Section 7.3 in UPDATED TPS 6e.
• Inference completely reorganized to match Course Framework units:

The 4-step “State-Plan-Do-Conclude” process that is primarily intended to help students answer inference questions on the AP® exam, is now only developed only in the inference chapters (8-12) of the Updated TPS6e.  Further, a streamlined two-step process for Normal distribution and Binomial distribution calculations has been introduced in the UPDATED sixth edition to match recent AP® rubrics. 

Here are specific content changes made in the UPDATED sixth edition based on the AP® Statistics Course Framework.

Chapter 1

• Added mosaic plots.
• Added margin note discussing mean ± 2 or 3 SD criterion for outliers.
• Emphasized terminology:
  - discrete vs. continuous quantitative variables
  - descriptive statistics vs. inferential statistics
  - statistic vs. parameter
  - unimodal (single-peaked), bimodal (double-peaked), approximately uniform
  - variance vs. standard deviation

Chapter 2

• Changed definition of percentile to less than or equal to, with equivalent changes to cumulative relative frequency graphs.
• Emphasized terminology:
  - empirical rule (68-95-99.7 rule)

Chapter 3

• Removed the term “outlier” in Section 3.1, as it has a more restrictive meaning in the course framework. Replaced with “unusual value.”
• Changed the equation of the least-squares regression line to .
• Introduced term “high-leverage point” as one whose x value is far from . Also, restricted the use of “outlier” to points with big residuals. Both high-leverage points and outliers can be influential points.
• Emphasized terminology:
  - univariate vs. bivariate data

Chapter 4

• Added systematic random sampling.
• Emphasized terminology:
  - homogeneous vs. heterogeneous
  - prospective vs. retrospective observational studies

Chapter 5

• Changed “chance process” to “random process”.
• Emphasized terminology:
  - trial
  - empirical vs. theoretical probability
  - disjoint (mutually exclusive)
  - joint probability

Chapter 6

• Expanded discussion of 10% condition and independent observations.
• Increased coverage of geometric random variables to include shape, center (mean), and variability (SD) of the probability distribution.
• Emphasized terminology:
  - probability distribution vs. population distribution; parameter
  - cumulative probability distribution
  - linear transformation vs. linear combination

Chapter 7

• Increased emphasis on the 10% condition, including idea that the traditional formulas for  and overestimate the actual SD when sampling without replacement (but the difference is negligible when the 10% condition is met).
• Emphasized terminology:
  - point estimator
  - randomization distribution
  - independent observations vs. independent samples

Chapters 8–11

• For each inference procedure, the conditions are introduced by stating that we need to check for independence in data collection methods (Random condition, 10% condition when sampling without replacement) and that the sampling distribution is approximately Normal (Large Counts, Normal/Large Sample).

Chapter 8

• Emphasized that a confidence interval is an interval estimate of plausible values for a parameter based on sample data.
• Emphasized that a confidence level represents the long-run capture rate when conditions are met.
• Modified generic formula for a confidence interval to use “standard error” rather than “standard deviation” to align with the new formula sheet.
• Emphasized that a confidence interval for a proportion can be used to create other intervals (e.g., for the total number of successes in a population).

Chapter 9

• Modified generic formula for a standardized test statistic to include standard deviation (error) of statistic in the denominator to align with the new formula sheet.
• Emphasized that the P-value is the probability of getting evidence for Ha as strong or stronger than the observed evidence when H0 is true, assuming the probability model is valid (conditions are met).
• Adjusted the decision rule for rejecting H0 to P-value ≤ α instead of P-value < α.
• For a two-sample z test of H0: p1 – p2 = 0, the Large Counts condition now uses expected counts calculated with the pooled (combined) proportion instead of observed counts.

Chapters 10 and 11

• Encouraged use of df from technology for two-sample t procedures.
• Placed paired data procedures immediately after two-sample procedures to help students differentiate these methods.

Chapter 12

• Emphasized that the chi-square statistic measures how far observed counts are from expected counts, relative to expected counts.
• Changed the equation of the population regression model to .
  Modified the Normal condition for inference about slope: At each x value, the distribution of y values is approximately Normal OR n ≥ 30.
• Added mention of a one-parameter model (with y intercept = 0) that uses df = n – 1 (otherwise df = n – 2 when estimating both the slope and y intercept).

In addition to the content changes, the Updated 6th edition has been refined to provide  more and better support for the diverse population of students that takes AP® Statistics today.

NEW – Many of the Examples now include Teacher Talk boxes that feature the voice of the teacher to guide students through the appropriate steps in the solution.

NEW -  More detailed Chapter Review Exercise Videos that augment the Section Example and Exercise Videos and help students prepare for the Chapter AP® Practice Test.

Activities and NEW Chapter Projects – with more than 30 built in Activities and 5 Chapter Projects, students are encouraged to learn by simulation and by “doing” statistics.   The activities and projects may be used to support collaborative learning as well as deeper investigation and study