Time-varying exposure subject to misclassification : bias characterization and adjustment
Measurement error occurs frequently in observational studies investigating the relationship between exposure variables and a clinical outcome. Error-prone observations on the explanatory variable may lead to biased estimation and loss of power in detecting the impact of an exposure variable. When th...
| Main Author: | |
|---|---|
| Language: | English |
| Published: |
University of British Columbia
2010
|
| Online Access: | http://hdl.handle.net/2429/27839 |