Learning styles are not an exact science. Many tests with very large samples have been conducted and there is evidence of correlation between how people start their learning, how they base their decisions, and even how it links to their personality traits. Even though you may be sceptic about the scientific value of the model, there is value in its usage. Overreliance on models and treating it as the sole truth can be hazardous. It’s a model. It’s fallible. But it can be useful.