A simulation-based comparison of tests for equivalence in clinical trials with application to tobacco data

The equivalence problem is concerned with statistical methodology to assess the equivalence between two medications or two formulations in clinical trials. Frequently, researchers incorrectly use difference tests to evaluate equivalence. This study introduces three common equivalence tests (two one-sided tests, power analysis, and the Hauck-Anderson method) and investigates their performance. We conclude by describing appropriate experimental design and testing procedures for assessing equivalence between tobacco composition developed by Health Canada (HC) and the World Health Organization (WHO) and explore their performance through a simulation-based comparison.