Similarity Factor (F2) – false and illusionary “statistics”!
In case, anyone is looking for sophisticated “statistical” jugglery, F2 provides an excellent example of thoroughly confusing people and science. Everyone has to use it (compliance requirement), as it is suggested in the regulatory (FDA) guidance documents. This parameter has not been described in statistics and absolutely has no relevance to the assessment of quality of manufactured products (tablet/capsule). It has been developed using drug dissolution data, and then applied to such data, which are invalid and irrelevant to start with. Calculation wise, it is simpler than calculating typical standard deviation more like a skill-test arithmetic exercise often provided at the back of lottery tickets or some sort of promotion or advertisement (link). Now consider this, a 2-day workshop/conference (yes two full days) is organized at the university level to explain and teach about it to make it look like science or statistics!!! (link or here). There certainly is a serious disconnect here between science and its practice at the regulatory level which requires attention.
