September 1, 2019
It is known to experts, that in the nonlinear regression analysis, because numerous curve fitting methods exist, which allow the statistician to cook up the data according to what he/she wants to see. Such a deep problem in nonlinear regression methods will be discussed in particular in the context of analyzing Hubble diagram from various existing data. It is our aim to distinguish the raw data and foregone conclusions, in order to arrive at a model-independent conclusion. As a preliminary remark, we deem that it remains possible that the Hubble law exhibits nonlinearity, just like proposed by Segal & Nicoll long time ago. More researches are needed to verify our proposition.