Errors Found in the Text

Note: These errors were ones I found in the first printing of the 5th edition. I reported them to the authors, and so hopefully some of them have been corrected in later printings of the 5th edition.


p. 10, near the middle of the page
in the expression for the mgf of a + bX, the a and b are reversed (the expression given is for the mgf of aX + b)
p. 11
pmf for the Bernoulli distribution isn't correct; the binomial coefficient should be omitted and n should be set to 1
p. 11
for the Poisson distribution the allowable parameter values should be μ > 0 instead of μ < 0
p. 12
pmf for the hypergeometric distribution isn't correct; in the denominator x should be n
p. 12
expectation for the hypergeometric distribution isn't correct; N should be n
p. 23, midway down page
in T - θ there should be a subscript of 0 on the θ
p. 31, 5th line above Fig. 2.2.1
instead of n, the parameter needs to be the specific value 10 in order for everything which follows to hold
p. 35, 4th line above Theorem 2.3.2
it should be Corollary 2.3.1.2 instead of 2.3.2
p. 35, statement of Theorem 2.3.2
there is a stray/extra [ right after the sup
p. 57, 2 lines above the top line of expression (2.10.1)
the word variance should be changed to standard deviation (or else the σ should be squared)
p. 67, in part (c) of Problem 2.4
the denominator of the density should be multiplied by b
p. 95, in 1st line of Example 3.5.1
n should equal 13 instead of 3
p. 101, in 2nd line of Sec. 4.1
the word "is" at the end of the line shouldn't be there
p. 169, last line
the 2nd probability should be replaced by 0.5 (i.e, it should indicate θ < 0.5)
p. 216, 8 lines from bottom
the lower confidence bound should be -1.00 instead of 1-1.00
p. 243, expression (6.4.3)
when giving the possible values of u, the 2nd value won't always be 1 ... e.g., if max(0, t - n) = 5, the 2nd smallest value would be 6
p. 275, last sentence of first paragraph
unless certain assumptions are made, one cannot rely on the ranks to make accurate inferences about the relationship of two distribution means --- it's possible that a negatively skewed distribution can have a smaller mean than a positively skewed distribution even though the probability mass is such that typically the ranks of the negatively skewed distribution will generally be larger than the ranks of the positively skewed distribution
p. 279, expression (7.3.3)
dun should be duN
p. 291, 8 lines above Table 8.2.1
it should be >= 17 instead of <= 17
p. 297, 5 lines from bottom
I'm not sure, but should it be 0.5 instead of 0.05?
p. 316
in the expression for the sum of the first N integers raised to the 4th power, the denominator should be 30 instead of 180
p. 343
I think the chapter title is a bad choice since the tests are about the equality of distributions and not the equality of samples --- could make it Tests for the Equality of k Distributions Based on Independent Samples (or perhaps The General k-Sample Problem (although other situations are also addressed in the chapter))
p. 346
in the last line of expression (10.2.2) it should be i = 1 instead of i = t under the summation symbol, and it should be nit/N instead of uit/N
p. 347 and p. 349
in the examples on both of these pages, the conclusions are statements about the distribution medians --- with no qualifying statements (e.g., that a shift model is assumed), I'd prefer that the conclusions just be that the distributions differ
p. 492, 12 lines below Table 13.3.1
the uniform distribution is characterized as having heavier tails than the normal, and while from one point of view that makes sense (since, where it's positive, the pdf of a uniform distribution is the same everywhere, while the pdf of a normal distribution gets smaller in value as we consider values farther from the mean), from another point of view uniform distributions are (extremely) light-tailed since the pdf goes down to 0 when one considers values more than 1.7321 standard deviation away from the mean (and uniform distributions do not have a tendency to produce extreme outliers like most heavy-tailed distributions do)