Errata

Second Edition

Page Location Issue
25 printed: tally(death ~ victim, data = DeathPenalty)
correct: tally(death ~ defendant, data = DeathPenalty)
Exercise 1.20 geyser is in the MASS package (see page 14). Use MASS::geyser to access this data set.
285 last line printed: $$-2n\pi$$
correct: $$-n\pi$$
286 lines 9-10 printed: correct when $$\pi_0 = 0$$ but overestimates … when $$\pi_0 > 0$$ correct: correct when $$\pi_0 = 1/2$$ but overestimates … when $$\pi_0 \neq 1/2$$
331 The problem could be clarified to indicate that the first moment is not useful. The second moment could be used.
373 printed: $$W = -2 \log\left(\frac{L(\theta_0)}{L(\hat\theta)} \right) = 2 \left(\log(\hat \theta) - \log(\theta_0)\right) \approx I(\hat \theta) (\theta_0 - \hat\theta)^2$$
correct: $$W = -2 \log\left( \frac{L(\theta_0)}{L(\hat\theta)} \right) = 2 \left(l(\hat \theta) - l(\theta_0)\right) \approx I(\hat \theta) (\theta_0 - \hat\theta)^2$$
380 lines 6, 11, 13 printed: pnorm(), pchisq(), pchisq()
correct: qnorm(), qchisq(), qchisq().
388 printed: The assymptotic distribution of $$\Lambda$$
correct: The assymptotic distribution of $$W = -2 \log(\Lambda)$$
389 printed: makLik
correct: maxLik
446 Exercise 5.18 printed: log of the odds ratio
correct: log of the odds
447 line 1 printed: poison
correct: Poisson
498 line 5 printed: internally studentized residuls
correct: internally studentized residuals
747 Exercise A.11 Unfortunately, the weatherData package has been removed from CRAN. Similar data now live in mosaicData::Weather, but the data for several cities are already contained in a single data frame, so there is no need to merge from multiple sources.