|
|||||
|
|
Precision Estimation in Zelen's Single-Consent Design |
|
| Authors: | J. P. Matts R. B. McHugh |
| Abstract: | The precision of a treatment contrast, as conventionally estimated in Zelen's Single-Consent Design, is shown under randomization theory to be positively biased. The magnitude of this bias is demonstrated and a new statistic is derived which is a consistent estimator of the precision. Confidence interval estimation for the treatment contrast is also presented. |
| Keywords: | Randomized clinical trial Consent design Randomization analysis Precision Confidence interval estimation |