Page 471/ Solution to Exercise 10.25, part b: t = 3.30; P-value=0.0011; the probability of observing a test statistic this extreme or even more extreme if there is no difference in the mean HONC scores (i.e., if the null hypothesis is true) is very small, 0.11%. There is strong evidence that gender has an effect: In the population that the survey refers to (seventh graders who have tried tobacco), the mean female HONC score differs significantly from the mean male HONC score.
Page 487: In Exercise 10.47 "Dominance of Politicians" the data show the dominance scores for male and female politicians. Part a asks to run a permutation test to see whether there is a difference in the ratings on competence between female and male speakers. Please replace "competence" by "dominance" in this sentence. (The data on the book's website, in addition to the dominance scores also contain a column for perceived competence. Using these as the responses, one does not find a significant difference between the distribution of competence scores for males and females, while the distributions of dominance scores are significantly different.)
Page 459: In Exercise 10.12 Obama A/B testing, the numbers in the X and N columns of the displayed computer output need to be switched, as follows:
Sample X N Sample p SignUp 5851 77858 0.075150 LearnMore 6927 77729 0.089117
Page 85: Exercise 112. There is an "e" missing at the end in "Hurrican" in the title of the Histogram.
Page 42: The "100" x-axis label for histogram on the top of page 42 should really be "160" and histogram should look like this:
Page 44: "everyday" in the label of the bar graph on the very bottom should read "every day".
Page 76: Exercise 2.82. The two outliers are not discernible for the CO2 values for Central & South America. There really should be two symbols (almost on top of each other) to label the outliers, as in the boxplot below. In addition, the typo that read "Emmissions" for the y-axis label was corrected:
Page 118: Beginning of line 3, second paragraph: The +2.61(0.267) should be a +26.1(0.267) since the estimated slope is 26.1 and not 2.61. The final result of that equation is correct, though.
Page 201: In the box for Activity 1, very last paragraph: The "it's" should be "its" as in "After each click, the cumulative proportion of 6s will tend to fall closer and closer to its expected value of 1/6."
Page 222, Table 5.3: The total shown in the margin for row "More than $1,000,000" should be 0.00249 and not 0.00222.
Page 95: Line two, last paragraph before Example 3 should say "..., which we found in part d of Example 2, ..." and not "...in part c...".
Page 321:Example 6, Think It Through, part a: The standard deviation of X should be $9.99 and not $10.05. The correct sentence should read: "Additionally, the standard deviation, sigma, of the winnings on a single $10 bet equals $9.99. Subsequently, the standard deviation of the sample mean x-bar should be $9.99/sqrt(40) = $1.58. Finally, in part b, the standard deviation of the sample mean for 400 spins should be $9.99/sqrt(400) = $0.499, or approximately 50 cents. (Thanks to Yoo Young B for pointing out this mistake.)
Page 180: The caption of Figure 4.2 is incorrect and confusing. Please replace Figure 4.2 and its caption with the following:
Page 353, Exercise 8.28: "The exercise should read like this: Using the Explore Coverage web app, let’s check that the large-sample confidence interval for a proportion may work poorly with small samples. In the app, set p = 0.30, n = 10 and leave the confidence level at 95%." Early versions of the 4th edition refer to the Explore Coverage Tab of the Inference for a Proportion App, which is now a stand-alone app.
Page 363: The answers to parts b and c of Exercise 8.33 in the back of the book are not in the correct order.
Page 378: Under the fourth bullet point, the part of the sentence "we use z = 1.96 for 95% confidence" appears twice.