Related Experiment Videos
On the use of historical controls in the analysis of laboratory data
1School of Engineering, Nagoya University, Japan.
Statistics in Medicine
|June 1, 1991
Abstract:
Our purpose is to propose a method of data analysis incorporating historical controls. It consists of three steps of hypothesis testing procedure. The first step is the check of concurrent controls. The second is the significance testing of treatment groups in comparison with historical controls which are composed of negative and positive controls. The third is the significance testing of a linear trend of the dose-response relationship. The superiority of the proposed procedure is also verified.