Pls let me know (yudi.pawitan@ki.se) if you find other errors.

- A full list of errors in the
**first print**is given in errors.ps or errors.pdf.

- Corrected exercise 5.9: ps or pdf.

List of known errors found

**Page 21 ** line 13 from top: theta=1/3 -> theta=0.3.

**Page 45 ** line 7 from top: L(g(theta)) -> L*(g(theta))

**Page 45 ** line 15 from top: reverse 3.6 and 4.5.

**Page 119 ** line 7 from bottom: b_1/n^2 -> b_2/n^2

**Page 184 ** line 6 from bottom: 160.26 -> 122.74 (missing sqrt for sigma in the Rcode EX6-19A.R)

**Page 188 ** line 1 from bottom: add 'Investigate also the
interaction between age and sex.'

**Page 290 ** line 6 from top: t_9 -> t_10

**Page 332 ** line 14 from top: beta_0w -> beta_0

List of known errors in the

**Page 82 ** line 12 from bottom: maximum likelihood function ->
maximum likelihood estimate.

**Page 93 ** line 8 from bottom: ybar=840.5 an s^2 = 4604.8 ->
ybar = 838.2 and s^2 = 4161.5.

**Page 119 ** line 15 from top: theta = e^{-2theta} -> theta=e^{-2mu}

**Page 140 ** line 7-8: Title 'Pivotal statistics' -> 'Pivotal quantity', and
the start 'A pivot is a statistic' -> 'A pivot is a random variable'. Since a pivot
is a function of both data and unknown parameters,
it should not be called a statistic. (Thanks to A. Azzalini)

**Page 155 ** line 3: theta_1 -> theta_i

**Page 199 ** line 7: Section 7.2 -> Section 3.2.

**Page 279 ** line 14: mu_i -> sqrt(2) mu_i

**Page 287 ** line 1: in the denominator L(thetahat,etahat_theta) -> L(thetahat,etahat)

**Page 287 ** line 9,10,12 from bottom: log missing in front of likelihood L(.)

**Page 289 ** line 6 from bottom: n/(2sigma-hat_mu^2) -> n/(2sigma-hat_mu^4)

**Page 289 ** last line: (n-1) -> (n-2)

**Page 380 ** line 3: lambda(theta_k)=0 -> lambda'(theta_k)=0