The picture below shows that, in reality, no ONE study can answer a question. In fact, when trying to reproduce 56 big published studies, Dr. Glen Begley found that only SIX studies reached the same result. Scary right?
Therefore, to claim that something is true we need MANY studies pointing to the same result. Ideally both long term observational (prospective cohort studies) and experimental studies (RCTs). Systematic reviews and meta-analyses are a good way of getting an overview of all studies trying to answer the same research question.
Sometimes there is no evidence, and in that case the best thing to do is to speculate based on experience and existing studies (which means that you need to know what evidence there IS).
This does NOT mean that you should make up an explanation based on cherry picked studies (i.e. LCHF “carbs make you fat because insulin”).
So the next time you assume something to be true: Find out if the “fact” is level A true, with many observational and experimental studies to back it up, of level D to G BS! 😋