Journal of biopharmaceutical statistics
-
The probabilistic rationale for statistical design and analysis of clinical trials is random assignment. While arithmetic and mathematical formulations may be identical to those used with random samples, we should not indiscriminately borrow tools from survey sample methods. Specifically, the confidence interval should be used sparingly, if at all. ⋯ The clinical trial is the model research tool for clinical medical research, founded on randomization. The confidence interval is a statistical tool for parameter estimation based on population sampling concepts. These tools are incompatible.
-
Statistical estimates and significance tests address distinct (but related) questions using the same data. Point estimates and confidence intervals of differences are statistical estimates that address: "How LARGE is the difference in the population of interest?" A significance test addresses the question: "How LIKELY was the difference to have occurred by chance?" Because p-values deal with the existence of a real nonzero difference between treatments but not the size of that treatment difference, they cannot be used to assess clinical (practical) significance. ⋯ The point estimate is the outcome difference actually observed in the study sample; it is also the best single-number estimate of the unknown difference in the sampled population. Point estimates, confidence intervals, and p-values extract complementary information from study data and should all be reported for major results.