TY - JOUR TI - Hypothesis testing of bio-equivalence DO - https://doi.org/doi:10.7282/T3T43RP4 PY - 2013 AB - This thesis considers two statistics problems in bio-equivalence. The first problem concerns 2-stage testing for bio-equivalence of parameters emanating from two populations. The first stage of the procedure can be thought of as a pilot sample which is used to determine the feasibility of taking an additional sample that would lead to inferring bio-equivalence. If a second sample is taken, the combined sample can then be used for inference purposes regarding bio-equivalence. Many models, including normal, Poisson, binomial, matched pairs, testing means and variances or both simultaneously are considered. The second part of the thesis is concerned with multiple testing of bio-equivalence. Here an actual data set concerned with analyzing different types of iron content with different instruments is studied for various bio-equivalent outcomes among pairs. To perform the statistical analysis, two standard statistical methods are used along with a new method. Both equivalence testing and simultaneous interval estimates are offered. KW - Statistics and Biostatistics KW - Equivalence relations (Set theory) KW - Biometry--Research LA - eng ER -