161:
74:
226:
202:
124:
145:
221:
231:
195:
236:
188:
51:
of a study, risk estimates themselves are not themselves attenuated (as would be the case where
176:
135:
114:
8:
47:
in cases where exposure data are highly aggregated. While this kind of error reduces the
40:
91:
141:
120:
48:
160:
87:
83:
44:
32:, Berkson error causes little or no bias in the measurement. It was proposed by
29:
172:
33:
215:
52:
21:
25:
36:
in an article entitled “Are there two regressions?,” published in 1950.
168:
95:
140:(Second ed.). London: Chapman & Hall. pp. 26–32.
43:
in epidemiological studies. Berkson error may predominate over
133:
134:
Carroll, R. J.; Ruppert, D.; Stefanski, L. A. (2006).
116:
72:Berkson, J. (1950). "Are There Two Regressions?".
213:
75:Journal of the American Statistical Association
196:
112:
203:
189:
71:
214:
65:
39:An example of Berkson error arises in
137:Measurement Error in Nonlinear Models
227:Statistical deviation and dispersion
155:
13:
106:
14:
248:
159:
88:10.1080/01621459.1950.10483349
1:
119:. CRC Press. pp. 76–78.
113:Buonaccorsi, John P. (2010).
58:
175:. You can help Knowledge by
7:
10:
253:
154:
24:(or misclassification) in
222:Accuracy and precision
171:-related article is a
232:Errors and residuals
20:is a description of
41:exposure assessment
18:Berkson error model
184:
183:
126:978-1-4200-6658-6
244:
237:Statistics stubs
205:
198:
191:
163:
156:
151:
130:
100:
99:
82:(250): 164–180.
69:
252:
251:
247:
246:
245:
243:
242:
241:
212:
211:
210:
209:
148:
127:
109:
107:Further reading
104:
103:
70:
66:
61:
55:predominates).
45:classical error
30:classical error
12:
11:
5:
250:
240:
239:
234:
229:
224:
208:
207:
200:
193:
185:
182:
181:
164:
153:
152:
146:
131:
125:
108:
105:
102:
101:
63:
62:
60:
57:
34:Joseph Berkson
9:
6:
4:
3:
2:
249:
238:
235:
233:
230:
228:
225:
223:
220:
219:
217:
206:
201:
199:
194:
192:
187:
186:
180:
178:
174:
170:
165:
162:
158:
157:
149:
147:1-4200-1013-1
143:
139:
138:
132:
128:
122:
118:
117:
111:
110:
97:
93:
89:
85:
81:
77:
76:
68:
64:
56:
54:
50:
46:
42:
37:
35:
31:
27:
23:
19:
177:expanding it
166:
136:
115:
79:
73:
67:
53:random error
38:
22:random error
17:
15:
26:measurement
216:Categories
169:statistics
59:References
28:. Unlike
96:2280676
144:
123:
94:
167:This
92:JSTOR
49:power
173:stub
142:ISBN
121:ISBN
16:The
84:doi
218::
90:.
80:45
78:.
204:e
197:t
190:v
179:.
150:.
129:.
98:.
86::
Text is available under the Creative Commons Attribution-ShareAlike License. Additional terms may apply.
