1985 年 105 巻 8 号 p. 796-800
The difference was examined in the nature of judgment drawn by the test of hypothesis, interval estimation, and Bayesian analysis with the same set of data obtained in a balanced two-period crossover bioequivalence trial. The conventional method and the method of Westlake, in which and interval symmetric with respect to the mean of the standard preparation was calculated, were used to estimate 95% confidence intervals. As to the Bayesian analysis, two different methods of calculation were employed. The bioequivalent region of the coefficient of intrasubject error variation, the ratio of the square root of the intrasubject error variance to the mean of bioavailability parameter for the standard preparation, was calculated against the ratio of the means for the test and standard preparations by using these methods. The region accepted as being bioequivalent based on the Westlake method was slightly larger than that based on the conventional method and agreed with that obtained by one method of calculation based on the Bayesian analysis.