Characteristics of Therapy in Statistical Analysis
One of the most central features of the controlled trial is the allocation of treatments to subjects by chance, a procedure known as randomization. This serves two functions: in the first place it guarantees an unbiased allocation of the treatments (and there is no better method of doing so) and therefore permits a fair comparison between treatments. Secondly, it provides the basis for the application of the theory of probability which underlies statistical techniques. It is true that statistical tests are applied to data in which there has been no randomization, but the inferences in such cases are insecure.
KeywordsError Variance Smoking Habit Fair Comparison Covariant Variable Emeritus Professor
Unable to display preview. Download preview PDF.