Statistics can be a tricky self-learn adventure, as a lot of the field is non-intuitive and sometime counter-intuitive. I would usually advise, for anything beyond rudimentary statistics, to have some sort of mentor (traditional teacher style, online, whoever you can get). Especially since it is easy for a small error to make a large difference in your conclusion.
However, as with any learning there is no ubiquitous "best place". Nor is there a universal "best method". Generally, you need to find what works for you - and I can almost guarantee it is going to be a combination of different mediums. You'll want reference books for the heavy lifting, videos for engaging, interactive content for practice, etc.
However, as with any learning there is no ubiquitous "best place". Nor is there a universal "best method". Generally, you need to find what works for you - and I can almost guarantee it is going to be a combination of different mediums. You'll want reference books for the heavy lifting, videos for engaging, interactive content for practice, etc.