modern epi cover
The following is a list of corrections for
Modern Epidemiology, Second Edition,
by Kenneth J. Rothman and Sander Greenland.



Last Updated March 8, 2009

See bottom of page for cross reference list of errata by page number.

Report Corrections EMAIL

Buy the new edition AMAZON

Download Episheet Episheetlogo

Epidemiology -- An Introduction introepi



1. Page xiii, line 2: Add "Constantine Daskalakis"

2. On page xiii, line 9, "colleages" should be "colleagues"

3. Pg. 8, last paragraph, line 1: "we are" should be "one may be";

4. Pg. 8, last paragraph, line 2: "causal status" should be "status"

5. Pg. 11, 2nd paragraph, line 1: "conditions" should be "condition"

6. Pg. 11, 2nd paragraph, line 2: "are" should be "is"

7. Pg. 11, 2nd paragraph, line 7: "E to act" should be "E to cause the disease"

8. Pg. 11, 3rd paragraph, line 1: "set of conditions" should be "condition"

9. Pg. 11, 3rd paragraph, line 2: "(A and U) or (B and U)" should be "A or B but not both"

10. Pg. 11, 4th paragraph, lines 6-7: "identity of the constituent components of the cause amounts to"should be "identities of the components of a sufficient cause are part of"

11. Pg. 11, 4th paragraph, lines 8 and 10: "complements" should be "complement"

12. Pg. 11, 4th paragraph, line 11: "complements" should be "cofactors"

13. Pg. 22, 1st paragraph, line 7: "(Kahneman et al., 1982)" should be "(Kahneman et al., 1982; Piattelli-Palmarini, 1994)."

14. Pg. 41, 2nd paragraph, line 9: "incidence rates as observed" should be "incidence rates observed"

15. Pg. 41, 3rd paragraph, line 4: Italicize "S" in "proportion S no longer holds,"

16. Pg. 50, 4th paragraph, third inequality: "<" should be ">" signs, so the inequality should be

R1

R1S 0

1

>

----

>

--------

R0

R0S 1

17. Pg. 50, 4th paragraph, last sentence: "the risk ratio is always closer" should be "the risk ratio will be closer"

18. Pg. 56, line 1: "Equation 4-1" should be "Formula 4-1"

19. Pg. 64, last paragraph, line 7: should be "expected years of life lost"

20. Pg. 64, last paragraph, line 8: add "Boshuizen and Greenland (1997)" to reference list

21. Pg. 77, last paragraph, line 6: "Incidence or mortality is the preferable measure" should be "Incidence and mortality are commonly used."

22. Pg. 104, last paragraph, line 5: "the effect" should be "adverse effect"

23. Pg. 104, last paragraph, last sentence: should begin "The irrelevance of considering opportunity in isolation from other relations ..."

24. Pg. 106, 1st paragraph, line 6-7: "proper analysis of the resulting data is more complex than that required for single control groups" should be "the case and control groups represent multiple outcomes; hence proper analysis may become complex"

25. Pg. 106, 2nd equation: remove "j107" at end of equation, i.e., equation should be

RRA/RR B = .... = ORAB .

26. Pg. 125, last line of footnote: "(1998)" should be "(1999)"

27. Pg. 187, 3rd paragraph, line 9: "resulting in a probable" should be "which probably resulted in"

28. Pg. 193, 2nd paragraph, line 4: "12-3" should be "12-1"

29. Pg. 194, line 4: "12-3" should be "12-4"

30. Pg. 198, 5th paragraph, line 8: add comma after "et al."

31. Pg. 203, line 11 from bottom: "coded as either" should be "coded as" (delete "either")

32. Pg. 214, 5th paragraph, line 1: the upper case Greek character Pi should be lower case

33. Pg. 222, 3rd paragraph, line 5: "to the right" should be "to the left"

34. Pg. 234, 4th paragraph, line 11: should begin "reference population are known to be"

35. Pg. 235, last paragraph, line 1: should begin "Assuming E has negligible sampling variability compared to A, it follows from..."

36. Pg. 236, line 7 from bottom: should begin "These results suggest that, if the Poisson model is correct and the sampling variability of E is negligible, there is..."

37. Pg. 242, 4th equation from bottom: "116" should be "16" , so the equation should be

^

RD = 12/14 - 7/16 = 0.42

38. Pg. 255, 1st paragraph, line 13: "(1998)" should be "(1999)."

39. Pg. 259, Table 15-2: reference is "Doll and Hill, 1966"

40. Pg. 276, 2nd paragraph, line 10: "context of regression modeling" should be "context of testing product terms in regression models"

41. Pg. 277, 3rd paragraph, last line: "to alter the conclusions" should be "to alter the inferences"

42. Pg. 279, last paragraph, lines 1-2: "and as homogeneity tests" should be "and homogeneity tests"

43. Pg. 293, 2nd paragraph, line 5: "survival regression" should be "incidence-time (survival) regression"

44. Pg. 297, line 2: "scrutinized any" should be "scrutinized in any"

45. Pg. 297, 1st paragraph, line 6: "would lead" should be "would probably lead"

46. Pg. 305, 2nd paragraph, line 4 after equation 17-1: lower subscript "P" in P2

47. Pg. 305, line 2 after equation 17-3: "17-1" should be "17-3"

48. Pg. 305, Equation after 17-3 which is not labeled: last term in middle sum should be

[209 - 488(387)/976]2

---------------------------

488(387)/976

49. Pg. 308, 5th paragraph, line 1: "presented so far in" should be "presented in"

50. Pg. 309, Fig. 17-2 label: should read "80% and 99% confidence limits from data in Table 17-4" (not Table 17-2).

51. Pg. 310, 2nd paragraph, line 4: "prevented fractions" should be "preventable fractions"

52. Pg. 310, Fig. 17-3 label: should be "80% and 99% confidence limits from data in Table 17-4" (not Table 17-2).

53. Pg. 312, Fig. 17-4 label: should be "80% and 99% confidence limits from data in Table 17-4" (not Table 17-2).

54. Pg. 318, 2nd paragraph, line 2: "can" should be "may"

55. Pg. 328, 2nd paragraph, line 3: "to capture" should be "to try to capture"

56. Pg. 334, last paragraph, line 3: "however equation" should be "however: Equation"

57. Pg. 334, last paragraph, line 4: "cohorts. In" should be "cohorts; in"

58. Pg. 336, line 11: "R11 - R00 + R01 - R00" should be "R10 - R00 + R01 - R00"

59. Pg. 337, last paragraph, line 5: "D, E, F, G, H" should be "D, E, G, H" (delete F)

60. Pg. 340, 2nd formula, " - R00" should be "+ R00"

61. Pg. 341, 2nd paragraph, line 1: "RDij" should be "RDjk"

62. Pg. 342: in equation, "0.19" should be "0.019" and "0.10" should be "0.010"

63. Pg. 342, last paragraph, line 1: "case load" should be "case-load"

64. Pg. 347, 3rd paragraph, line 6: "Greenland, 1987" should be "Greenland, 1987d"

65. Pg. 347, 3rd paragraph, line 7: "Chapter 19" should be "Chapter 32"

66. Pg. 353, in equation number add left bracket: "[19-10]"

67. Pg. 356, 4th paragraph, line 4: "endometrial cancer)" should be "endometrial cancer; see Chapter 9)"

68. Pg. 364, line 2: "set to x" should be "set to x (see Chapter 4)"

69. Pg. 371, line 3 under Rescaling: "interested" should be "interest"

70. Pg. 385, 3rd paragraph, line 2: "intercept beta" should be "intercept alpha"

71. Pg. 389, 2nd full paragraph, line 1-2: "dominate if they get in the way" should be "prevent one from making"

72. Pg. 391, 5th line from bottom of page: "the power model" should be " the fractional polynomial model"

73. Pg. 392, 1st paragraph, line 8: "delta-beta" should be "influence"; "Chapter 21" should be "Chapter 13"

74. Pg. 394, 2nd paragraph, line four, end of line: printing blotch on word "be"

75. Pg. 412, 2nd paragraph, line 1: "One important and simple diagnostic tool" should be "One important form of influence analysis"

76. Pg. 416, 3rd paragraph, line 6: italicize x in "a total of T(x) person-time"

77. Pg. 421, line 1: delete quotation marks

78. Pg. 422, bottom title: should be "Time-Dependent Exposures and Covariates"

79. Pg. 532, 2nd paragraph, line 3: "the same parasite" should be "for a given parasite"

80. Pg. 533, 2nd paragraph formula: "SAR" should be

^

SAR

81. Pg. 533, 4th paragraph, line 1: "SAR" should be "the SAR"

82. Pg. 534, Fig. 27-3 label: "attach" should be "attack"

83. Pg. 535, 1st line after formula: "SAR" should be

^

SAR

84. Pg. 536, 1st paragraph, line 16: "infectives" should be "infections"

85. Pg. 536, 3rd paragraph, line 5: "a population" should be "a large population"

86. Pg. 545, 3rd paragraph, line 2: "fixed cohort" should be "closed cohort"

87. Pg. 545, 3rd paragraph, line 8 or last line: "fixed cohort" should be "closed population"

88. Pg. 548, 3rd paragraph, line 6: "fixed cohort" should be "study cohort"

89. Pg. 548, 3rd paragraph, 3rd line from bottom: "fixed cohort" should be "study cohort"

90. Pg. 548, last paragraph, line 1: "incidence as a function of prevalence" should be "incidence, dynamics, and prevalence"

91. Pg. 548, last paragraph, line 5: "number of new clinical cases" should be "incidence rate"

92. Pg. 549, 3rd paragraph, line 6: "higher number" should be "higher rate"

93. Pg. 553, 3rd paragraph, line 1: "in a person" should be "among persons"

94. Pg. 618, Table 30-6: "U0" should be "V0", "V0" should be "U0", "U1" should be "V1", "V1" should be "U1"

95. Pg. 655, 1st paragraph, last line: "1/723" should be "1/7231/2"

96. Pg. 662, 1st paragraph, line 3, add references after "a single number": (Greenland, 1987d; Rubin, 1992).

97. Pg. 663, 4th paragraph, line 1: bk should be

-

bk

98. Pg. 665, 6th paragraph, line 3: remove ", for example,"

REFERENCE ADDITIONS AND CORRECTIONS

99. Greenland S, Pearl J, Robins JM. Causal diagrams for epidemiologic research. Epidemiology 1999; in press.

100. Hertz-Picciotto I, Gravitz N, Neutra RR. How do cancer risks estimated from animal bioassays compare with the epidemiologic evidence? The case of ethylene dibromide. Risk Analysis 1988;8:205-214.

101. Krieger N. Epidemiology and the web of causation: has anyone seen the spider? Soc Sci Med 1994;39:887-903.

102. Michels KB, Greenland S, Rosner BA. Does body mass index adequately capture the relation of body composition and body size to health outcomes? Am J Epidemiol 1998;147:167-172.

103. Piattelli-Palmarini M. Inevitable Illusions. New York, Wiley, 1994.

104. Robins, JM. Causal inference from complex longitudinal data. In: Berkane M, ed. Latent variable modelling with application to causality. New York: Springer-Verlag, 1997, 69-117.

105. Robins JM, Greenland S. The probability of causation under a stochastic model for individual risks. Biometrics 1989b;45:1125-1138.

106. Robins JM, Tsiatis AA. Correcting for non-compliance in randomized trials using rank-preserving structural failure-time models. Commun Stat 1991;20:2609-2631

107. Rothman KJ, Boice JD. Epidemiologic analysis with a programmable calculator. 2nd ed. Newton, MA: Epidemiology Resources, 1982.

SUBJECT INDEX ADDITIONS AND CORRECTIONS

108. Alpha error, 186

109. Alpha level, 186

110. Bayesian analysis

subjective, 197-198, 366

111. Beta error, 186

112. Bonferroni adjustments, 227-228

113. Case-control studies

control series for (control group), 74, 94

multiple, 105-107

multiple controls, 105-107

114. Categorization

open-ended, 206-207

percentile, 206

115. Competing risks, 36, 41, 288, 289

116. Complementary log-log risk model, 399

117. Conditional logistic regression, 420

118. Cumulative-odds model, 397

119. Direct adjustment. See Standardization

120. Error. See also Misclassification

alpha, 186

beta, 186

121. Expected years of life lost, 64

122. External adjustment

for confounding, 344-347, 649-652

for misclassification, 347-355, 653

for selection bias, 355-356, 652-653

123. Fractional polynomial model, 391

124. Hypothesis testing, 23

statistical, 183-188, 202

125. Indirect adjustment. See Standardized morbidity ratio

126. Influence analysis, 209-210

for regression analysis, 412. See also Delta-beta analysis

127. Missing values, handling of, 207-208

128. Multilevel models, 428. See also Hierarchical regression

129. Multiple-comparisons analysis, 225-229, 322-328

130. Neyman-Pearson hypothesis tests, 186-189

alpha error, 186

beta error, 186

131. Open-ended categories, 206-207

132. Percentile categories, 206

133. Polynomial model, 391

134. Power, 186

135. Preventable fractions, 55, 57

in graphical analysis, 310

136. Proportional-odds model, 397

137. Quantiles, 206

138. Rate-advancement period, 64

139. Recurrent outcomes, 424

140. Regression models

proportional hazards, 376-377, 422, 647

survival, 375-377

141. Risk advancement period, 64

142. Sample size, 232. See also Study size

143. Standardization, 45-46, 63-64, 234, 260-265

model-based, 415

144. Standardized morbidity ratio, 234-235, 262, 295, 655-656

145. Statistical power, 186

146. Study size, 117

147. Time-dependent covariates, 422-423

148. Type I error, 186

149. Type II error, 186

150. Page 654, first full paragraph, line 9: the first R within parentheses should have an overbar

151. Page 447, last paragraph, line 1: "such a payment" should be "such as payment"

152. Page 636, last line: omit the line (it is duplicated on the next text page)

153. Page 637 (Table 31-1): second line under column "Population" should be "(n=194)" instead of "n=194)"

154. Page 272 (Formula 15-24): the denominator of the middle term should be 2(Sumi)Gi(Sumi)Hi rather than 2(Sum)Gi Hi

155. Page 276, lines 10-11: the risk difference heterogeneity formulas should have the expressions Var[RD1] and Var[RD2] instead of Var[ln(RD)1] and Var[ln(RD)2]

156. Page 270, line 14 (formula for IDMH): the number "11,47" should be "1,147"

157. Page 202, par 4, line 4: the phrase "for machine entry, usually by keypunching." should be replaced by the phrase "for machine entry."

158. Page 429, lines 5-7: the right sides of the three equations should each be placed within the expression "expit( ... )"

159. Page 697, Reference: "Nurminen N. Asymptotic efficiency ..." should be "Nurminen M. Asymptotic efficiency ..."

160. Page 697, Reference: "Nurminen M, Mutanen P. Exact Bayesian analysis of risk measures." should be "Nurminen M, Mutanen P. Exact Bayesian analysis of two proportions."

161. Page 12, 9th line from bottom: "two mechanisms, I and II" should be "two mechanisms, II and III"

162. Page 198, par. 3, line 6: "like that in Fig. 12-3" should be changed to "similar in shape to Fig. 12- 5"

163. Page 269, par. 2, line 8: "is widely available" should be changed to "is available"

164. Page 431, par. 3, 2nd to last line: "with mean zero" should be "with mean zero and standard deviation of 1"

165. Page 667, par. 5, line 1: "usually does" should be "often does"

166. Page 709, line 5: "in press" should be "9:16-20"

167. Page 716, Add index entry: "Collapsibility, 53-54, 60"

168. Page 728, Add index entry: "Noncollapsibility, 53-54, 60"

169. Page 699, line 1: "D Thomas" should be "DC Thomas"

170. Page 22, par. 1, line 8: "such reasoning" should be "sound reasoning"

171. Page 22, par. 1, last line: Add "(Greenland 1998ab)" at end

172. Page 429, par. 2, last line: Add "and Witte et al. (1998)" at end

173. Page 687, add references:

Greenland S. Induction versus Popper: substance versus semantics. Int J Epidemiol 1998a;27:543-548.

Greenland S. Probability logic and probabilistic induction. Epidemiol 1998b;9:322- 332

174. Page 708, Add reference:

Witte JS, Greenland S, Kim LL. Software for hierarchical modeling of epidemiologic data. Epidemiology 1998;9:563-566.

175. Page 263, line 3 after formula 15-11: "risk ratios" should be "risk differences and risk ratios"

176. Page 735, under "Standardized morbidity ratio": "262" should be "262-264"

177. Page 735, under "Standardized rate": "45-46" should be "45-46, 260-262"

178. Page 735, under "Standardized rate difference": "63-64" should be "63-64, 261-263"

179. Page 735, under "Standardized rate ratio": "262" should be "262-265"

180. Page 735, under "Standardized risk difference": "64" should be "64, 261, 263"

181. Page 350, formula 19-7: the second plus sign should be a minus sign

182. Page 269, formula 15-15: in the middle expression, the final term in the numerator should be I0i, not I01

183. Page 713, under "Case-control data," add "two-stage, 112-114, 278-279, 418"

184. Page 737, add "Two-stage data, analysis of, 278-279, 418" and "Two-stage sampling, 112-114"

185. Page 401, last line, and page 402, first line: delete "for example"

186. Page 700, under Robins JM, Greenland S, Hu F: replace "(to appear)" by "1999; 94: in press"

187. Page 522, first formula for PV+: "LR" should be "LR+"

188. Page 319, par 3, line 2: "Table 17-2" should be "Table 17-4"

189. Page 259, footnote under Table 15-2 should read: "Deaths per 10,000 person-years."

190. Page 55, start of the line before formula 4-1: delete "the"

191. Page 56, start of equation 4-3: should begin "Pr(C) = F = "

192. Page 56, first line of third paragraph: "fraction of cases" should be "fraction of A0 cases"

193. Page 56, third line of third paragraph: "fraction of cases under exposure" should be "fraction of A1 cases"

194. Page 284, equation 16-4, first line: replace " + 1/(U+W) + 1/(V+W)" with " - 2/P"; next two lines replace " +(U+V+2W)/(U+W)(V+W)" with " - 2/P".

195. Page 284, lines 1-3 after equation 16-4 should read: "This quantity is larger than that in formula 16-3 if the pair outcomes are positively associated, i.e., if TW > UV, as is usually the case. In this case, the crude variance estimate ... "

196. Page 284, line 2 after 16-5 should begin: "timate if TW > UV. The ..."

197. Page 286, line 3 after 16-8: "N1i" should be "B1i"

198. Page 290, right hand side of equation 16-14: "Nk-1" should be "Nk - 1"

199. Page 296, lines 2-4 after equation 16-25: delete the carets (hats) from AFp

200. Page 296, unnumbered variance equation: Var should have caret (hat)

201. Page 296, 12 lines from bottom, 95% limits for ln(1 - AFp): lower limit should be - 0.4740 rather than 0.4740"

202. Page 672, last line: end of sentence should be " ... to affect bias (see chapter 19)."

203. Page 698, under Poole C. Beyond the confidence interval": "195-9" should be "195-199"

204. Page 18, end of first full paragraph: "the latter on the former" should be "sewer service on heart disease"

205. Page 20, fourth line under heading "Bayesianism": "British philosophers, but the" should be "philosophers. The"

206. Page 86, Fig 6-1: in the upper part of the figure, the shaded area should begin at 1965, not 1960

207. Page 688, under Greenland, Pearl, Robins: "In press" should be "10:37-48."

208. Page 717, under Correlation Coefficient(s): add ". See also R2"

209. Page 278, first two formulas under "TWO STAGE DATA": The term " A0i* " should be " A0i " with no asterisk, as should the second occurrence of " B0i* " in the next line. The first occurrence of " B0i* " in that line should have a caret above it.

210. Page 673, start of line 1: "main" should be "problems remain"

211. Index addition, p. 725: "Inverse-variance weighting, 266"

212. Index addition under "Precision," p. 730: "weighting, 266"

213. Page 269, the line above formula 15-16: "RDMH" should be "IDMH"

214. Page 655, line 11 of second paragraph: "standard error" should be "squared standard error"

215. Page 485, Table 24-1: First column of rates should be headed "1880", not "1860"

216. Page 688, Greenland, Pearl and Robins reference: year should be "1999" rather than "1998"

217. Page 216:

218. Page 199, line 11: 1992 should be 1992a

219. Page 264, line 8 under "Standardized Case-control Estimates": B01 should be B0i

220. Page 293, second and third line from the bottom: "negative" should be "positive" and "positive" should be "negative"

221. Page 662, line 10 from bottom: "0.027/0.035" should be "0.048/0.049".

222. Page 11, line 4 of first full paragraph: "incidence proportion for E" should be "effect of E".

223. Page 11, line 7 of first full paragraph: "for the two populations" should be "in the two populations".

224. Page 11, first full paragraph, last sentence: change to "In contrast, the difference in the association of B with disease in the two populations reflects differences in the prevalence of complementary cause A in the two populations (a phenomenon referred to as confounding -- see chapter 4)."

225. Page 700, Robins, Greenland and Hu (1999): After "outcome" add "(with discussion)", and at end of line replace "in press" by "687-712."

226. Page 354, 2nd line from bottom (eq. 19-12): "P-" should be removed from the beginning of the line.

227. Page 111, line 4 of 2nd paragraph of "Case-Only Studies": "equally likely location" should be "location that would be equally likely were there no exposure effect."

228. Page 714: Add "Case-specular studies, 111."

229. Page 682: "Deeley JE" should be "Deely JE."

230. Page 476: In table 23-5, Group 1 total cases should be 90 instead of 100 for both the correctly classified and the misclassified data, and the corresponding rates should also be 90 instead of 100.

231. Page 123, last sentence of 3rd full paragraph: delete the word "chance."

232. Page 127, next to last line of "differential misclassification" section: "induce" should be "change nondifferential error to".

233. Page 85, third line after header "Categorizing Exposure": replace "at all, but rather to" with "at all. Instead, one may use".

234. Page 714, second column, entry for "Category indicator variables": should have p. 387- 391.

235. Page 724, first column, add entry: "Indicator variables, 387-391".

236. Page 701, under Rubin DB. Bayesian inference for causal effects: vol. no. should be 6, not 7.

237. Page 340, third full paragraph, line 3: substitute "type 2" for "type 8".

238. Page 666, Table 32-4: delete the second "(" in the row labeled "Z statistic".

239. Page 667, second to last line of first full paragraph: "at" should be "of".

240. Page 667, fourth full paragraph: all three occurrences of "RSS/f" should be "RSS/df".

241. Page 109, last sentence of last full paragraph: replace "effect estimation" with "rate-ratio estimation" and replace "not related to risk of getting the study disease" with "either not associated with exposure or not associated with disease risk".

242. Page 271, 3rd line from bottom, and page 272, first line after equation 15-23: "Ni" should be "N+i".

243. Page 311, top paragraph, second to last line: "R2/R0" should be "R2/R1".

244. Index addition: p. 721: add "Exponential rate model, 375, 416-420".

245. Index additions: p. 725: Under "Linear odds model" add 420; Under "Linear rate model" add 418-420; Add "Linear relative-risk model, 418-420".

246. Page 318, third line after the last formula, and page 321, fourth line from bottom: "wx(u) should be wu(x)".

247. Page 319, second equation: "w0.95(0.94) should be w0.94(0.94)".

248.

249. Page 307, second line above figure: the gamma after "100" should not be subscripted.

250. Page 312, sixth line of text: "third column" should be "second column".

251. Page 397, models 20-52 and 20-53: All four Pr expressions should have "|X = x" inserted before the closing parenthesis ")", as in model 20-51.

252. Page 397, line 3 from end of second paragraph: "(e.g., blood count)" should be "(e.g., blood pressure)".

253. Page xiii, line 5: add "Trung Nam Tran" after "James Robins".

254. Page 198, line 15 from bottom: "Freedman, 1986" should be "Freedman, 1985, 1987"

255. Page 194, paragraph 4, line 5: Oakes (1986) should be Oakes (1990)

256. Page 256, last paragraph, line 1: "Miettinen, 1976" should be "Miettinen, 1976b"

257. Page 257, first paragraph, last line: "Maldonado and Greenland, 1993" should be "Maldonado and Greenland, 1993a"

258. Page 257, paragraph 3, line 2: "Miettinen, 1976" should be "Miettinen, 1976b"

259. Page 295, paragraph 3, line 2: "Greenland, 1984" should be "Greenland, 1984b"

260. Page 296, paragraph 2, 2 lines after eq. 16-25: "Greenland, 1987" should be "Greenland, 1987c"

261. Page 299, paragraph 4, 2nd-to-last line: "Robins, 1995" should be "Robins, 1997"

262. Page 357, paragraph 2, lines 1-2: "Greenland and Robins, 1985" should be "Greenland and Robins, 1985a"

263. Page 391, paragraph 3, line 4: "Royston and Altman, 1993" should be "Royston and Altman, 1994"

264. Page 418, lines 1-2: "Thomas 1981" should be "Thomas 1981b"

265. Page 424, lines 1-2: "Greenland et al., 1998" should be "Greenland et al., 1999"

266. Page 54, paragraph 4, line 2: "etiologic fraction" should be italicized

267. Page 56, line 1: "etiologic fraction" should not be italicized

268. Page 191, paragraph 3, line 5: "Foulkes,1978" should be "Folks, 1978"

269. Page 684, insert reference: Sullivan KM, Foster DA. Use of the confidence interval function. Epidemiology 1990;1:39-42.

270. Page 191, paragraph 3, line 6: "Foster and Sullivan" should be "Sullivan and Foster"

271. Page 730, second column, add primary entry: "Probability of Causation, 56"

272. Page 698, add reference: Poole C. Exceptions to the rule about nondifferential misclassification (abstract). Am J Epidemiol 1985;122:508.

273. Page 225, line 17: 0.10 should be 0.11.

274. Page 283, end of third line after equation 16-1: "U type-3" should be "U type-2".

275. Page 283, 4th line after equation 16-1: A1i should be A0i.

276. Page 402, line 3 before new section: delete "the search".

277. Page 664, second, fifth and last line of third full paragraph and first line of fourth full paragraph, change "f" to "df".

278. Page 708, under Witteman et al: replace "to appear" by "148:390-401".

279. Page 257, end of first paragraph and end of third paragraph: "1993" should be "1993a".

280. Page 694, reverse the "a" and "b" for the two Maldonado and Greenland 1993 references.

281. Page 243, 4th line from the bottom and 10th line from the bottom, delete "95%".

282. Page 472, line 2: insert "a" before "confounder".

283. Page 472, line 4: delete "especially".

284. Page 424, lines 5 and 10 of second paragraph under "Recurrent Outcomes": "Robins et al., 1998" should be "Robins et al., 1999".

285. Page 333, title of Table 18-2: Replace title with the following: "Response types given a binary outcome and two binary exposures".

286. Page 243, paragraph 2, line 7: "M0 - N1 - E" should be "M0 - N1 + E".

287. Page 50, line 16: both equals signs, not just the second, should have a dot above them.

288. Page 18, line 16: "1945" should be "1939"

289. Page 701, Add reference:

Russell B. Dewey's new 'Logic.' In The philosophy of John Dewey (ed. P. A. Schlipp) Tudor, New York, 1939.

290. Page 727, add to index: "Model selection, 401-404"

291. Page 730, add to index: "Potential outcomes. See Counterfactual definitions"

292. Page 275, 4th line above "Testing Homogeneity", "0.0008" should be "0.008"

293. Page 247, Formula for variance of ln(RR_hat) in middle of the page: both occurrences of M01 should be M10

294. Page 315, Formula for V, first line: sj2 should be sj2

295. Page 315, Formula for V, first line: E2 should be (E/M1)2

296. Page 315, Formula for V, second line: 2870.82 should be 5.88282

297. Page 271, Title of table 15-4: "lung cancer" should be "coronary death"

298. Page 684, Fifth reference listed: "Ephron E" should be "Efron E"

299. Page 278, Formula 15-28: The bracketed term in the numerator sum should be divided by "[N+i]2", the square of the stratum-i total.

300. Page 278, Line following formula 15-28: "(Sato, 1992b)" should be "(Sato, 1992a)"

301. Page 263, fourth line of text from bottom: "3.60 X 104 years" should be "3.60 per 104 years"

302. Page 352, lines 13, "1%" should be "10%" and line 14, change both instances of "0.01" to "0.1"

303. Page 194, second paragraph under "Guidelines for Practice": add to start of list of references "Rozeboom (1960)".

304. Page p. 701, add: Rozeboom WM. The fallacy of null-hypothesis significance test. Psych Bull 1960; 57:416-428.

305. Page 293, line 3 from bottom: "positive" should be "negative", and "negative" should be "positive".

306. Page 378, eq. 20-25: remove comma at end of second line; remove R(x1) at start of third line; on both superscript betas the 1 should be a subscript.

307. Page 515-517: the 3 section headings should be capitalized.

308. Page 690, in the reference Hosmer DW, Hosmer T,..., "Lee Cessie" should be "Le Cessie".

309. Page 690, Howe GM should be Howe GR.

310. Page 698, the Poole 1985 reference should be moved to the top of the Poole references.

311. Page 725, col. 1, add: Internal validation, 353-355.

312. Page 731, col. 2: "Rate models, 375" should be "Rate models, 375, 420"

313. Page 737, col. 2, add: Validation study, 353-355, 636

314. Page 722, add: External Validity, See Validity, external

315. Page 695, The title of the citation for McNemar Q should read: "Note on the sampling error of the difference between correlated proportions or percentages."

316. Page 681, fourth line from bottom, "Campbell OT" should be "Campbell DT" and "Quasi-experimentation" should be italicized.

317. Page 681, 8th citation, "1990d" should be "1991a", title should be "Acquired Immunodeficiency Syndrome --- United States, 1981 - 1990", pages should be "358-363", and citation should be moved to below the one that follows it.

318. Page 681, 9th citation, change "1990d" to "1990c", change pages from "421-424" to "421-423", and move citation above the one that precedes it.

319. Page 681, 10th citation, "1991e" should be "1991b".

320. Page 454, 7th line from bottom, "1990d" should be "1991a".

321. Page 126, line 4, insert at end of sentence "or on errors in other variables".

322. Page 388, line 3 from bottom, change "formerly" to "ever".

323. Page 389, line 1, change "not currently" to "formerly".

324. Page 389, line 2, insert at end of sentence, "among those ever married (Z2 = 0)".

325. Page 271, formula 15-21, "B0i" should be "B1i" and "B1i" should be "B0i".

326. Page 379, line 6 from bottom: "x" at start of line should be "x1".

327. Page 394, line 5 (second equation): "-5.00" should be "-6.00".

328. Page 404, line 5 before the section "MODEL FITTING": "following" should be "from".

329. Page 10, line 3, "throughout the book" should be "in most of the book".

330. Page 731, for "Recurrent outcomes", the page numbers should be "424, 548-549".

331. Page 350, lines 8-9: change "cases are always more likely to be classified as exposed than noncases" to "that sensitivity is higher among cases than among noncases".

332. Page 686, lines 25-26 (Greenland 1982 paper): change "1982" "1982a", and then insert new reference: "Greenland S. Interpretation and estimation of summary ratios under heterogeneity. Statistics in Medicine 1982b;1:217-227."

333. Page 154, last line before title "Benefits of Matching": change "1982" to "1982a".

334. Page 347, line 5 under title "Analyses of Misclassification": change "1982" to "1982a".

335. Page 276, line 7 from bottom of page: change "1982" to "1982b".

336. Page 243, footnote under table 14-4: change "antibiotics" to "antibodies".

337. Page 528, 4th line from bottom: change "1986" to "1996".

338. Page 271, formula 15-22: N01 should be N0i.

339. Page 356, second full paragraph, line 7: "selection" should be changed to "omission", and everything from ", where tA and" on line 7 to "of control selection)" on line 10 of the paragraph, including all of lines 8 and 9, should be dropped.

340. Page 723, Add index entry: "Healthy Worker Effect, 119, 138"

341. Page 333, Table 18-2: in rows 1 and 16, "effects" should be "effect"

342. Page 367, lines 5-6: "need only predict the distances to 2% accuracy" should be "can tolerate up to 2% error in our predictions"

343. Page 367, lines 9-10: "need to predict the distances to within .2% accuracy" should be "can tolerate up to .2% error in our predictions"

344. Page 415, formula in lines 10 and 11 from bottom: second term, x1 should be x12; third term, x2 should be x22; and all covariance terms should have a 2 in front.

345. Page 274, line 14: "one-tailed" should be "one-sided" and "two-tailed" should be "two-sided".

346. Page 275, line 11: "one-tailed" should be "one-sided" and "two-tailed" should be "two-sided".

347. Page 275, lines 17 and 18: "one-sided lower" should be "one-sided upper".

348. Page 389, line 1: "not currently married" should be "formerly married".

349. Page 241, first line after first equation: "P = 0.0002" should be "P = 0.002".

350. Page 667, lines 12 from bottom and 9 from bottom: both occurrences of "(0.00738)" should be "(0.00738)5"

351. Page 734: Add "Selection confounding, 355"

352. Page 56, line 6, equation 4-2 (twice), and equation 4-3: F should be Pr(C).

353. Page 702, Sheps 1958 reference: page numbers should be 1210-1214.

354. Page 734: Entry for "Sparse-data problem, 155" should be "Sparse-data problem, 155, 258"

355. Page 734, Add entry: "Sparse-data methods, 268, 420"

356. Page 250, line 2 below subheading "Single Study Group": (14.6) should be (14-5)

This is the end of erratum list. The list below is a cross-reference between page number and erratum number as listed above:

Page no.

Erratum no.

xiii

1.

xiii

2.

xiii

253.

8

3.

8

4.

10

329.

11

5.

11

6.

11

7.

11

8.

11

9.

11

10.

11

11.

11

12.

11

222.

11

223.

11

224.

12

161.

18

204.

18

288.

20

205.

22

13.

22

170.

22

171.

41

14.

41

15.

50

16.

50

17.

50

287.

54

266.

55

190.

56

18.

56

191.

56

192.

56

193.

56

267.

56

352.

64

19.

64

20.

77

21.

85

233.

86

206.

104

22.

104

23.

106

24.

106

25.

109

241.

111

227.

123

231.

125

26.

126

321.

127

232.

154

333.

187

27.

191

268.

191

270.

193

28.

194

29.

194

255.

194

303.

198

30.

198

162.

198

254.

199

218.

202

157.

203

31.

214

32.

216

217.

222

33.

225

273.

234

34.

235

35.

236

36.

241

349.

242

37.

243

281.

243

286.

243

336.

247

293.

250

356.

255

38.

256

256.

257

257.

257

258.

257

279.

259

39.

259

189.

263

175.

263

301.

264

219.

269

163.

269

182.

269

213.

270

156.

271

242.

271

297.

271

325.

271

338.

272

154.

274

345.

275

292.

275

346.

275

347.

276

40.

276

155.

276

335.

277

41.

278

209.

278

299.

278

300.

279

42.

283

274.

283

275.

284

194.

284

195.

284

196.

286

197.

290

198.

293

43.

293

220.

293

305.

295

259.

296

199.

296

200.

296

201.

296

260.

297

44.

297

45.

299

261.

305

46.

305

47.

305

48.

307

249.

308

49.

309

50.

310

51.

310

52.

310

311

243.

312

53.

312

250.

315

294.

315

295.

315

296.

318

54.

318

246.

319

188.

319

247.

328

55.

333

285.

333

341.

334

56.

334

57.

336

58.

337

59.

340

60.

340

237.

341

61.

342

62.

342

63.

347

64.

347

65.

347

334.

350

181.

350

331.

352

302.

353

66.

354

226.

356

67.

356

339.

357

262.

364

68.

367

342.

367

343.

371

69.

378

306.

379

326.

385

70.

388

322.

389

71.

389

323.

389

324.

389

348.

391

72.

391

263.

392

73.

394

74.

394

327.

397

251.

397

252.

401

185.

402

276.

404

328.

412

75.

415

344.

416

76.

418

264.

421

77.

422

78.

424

265.

424

284.

429

158.

429

172.

431

164.

447

151.

454

320.

472

282.

472

283.

476

230.

485

215.

515

307.

522

187.

528

337.

532

79.

533

80.

533

81.

534

82.

535

83.

536

84.

536

85.

545

86.

545

87.

548

88.

548

89.

548

90.

548

91.

549

92.

553

93.

618

94.

636

152.

637

153.

654

150.

655

95.

655

214.

662

96.

662

221.

663

97.

664

277.

665

98.

666

238.

667

165.

667

239.

667

240.

672

202.

673

210.

677

350.

681

316.

681

317.

681

318.

681

319.

682

229.

684

269.

684

298.

686

332.

687

173.

688

207.

688

216.

690

308.

690

309.

694

280.

695

315.

697

159.

697

160.

698

203.

698

272.

698

310.

699

169.

700

186.

700

225.

701

236.

701

289.

701

304.

702

353.

703

183.

708

174.

708

278.

709

166.

714

228.

714

234.

716

167.

717

208.

721

244.

722

314.

723

340.

724

235.

725

211.

725

245.

725

311.

727

290.

728

168.

730

212.

730

271.

730

291.

731

312.

731

330.

734

351.

734

354.

734

355.

735

176.

735

177.

735

178.

735

179.

735

180.

737

184.

737

313.