The scientific method doesn't guarantee correctness. It's inherent in it's process to be wrong at times.
The way the scientific method works is by proposing a theory that explains a natural process within some scope. Then scientists don't try to prove it, they try to disprove it. If a theory can't be disproved, it becomes accepted until someone finds a way to disprove it, at which point a new theory gets developed.
Most modern theories, e.g. gravity or evolution, are accepted within their scopes because no one has managed to disprove them despite thousands of scientists trying their best to do so. Disproving popular theories and creating new theories to replace them that don't get disproved themselves results in quite a bit of fame. That doesn't mean that they are necessarily correct, just that no one has managed to prove that they are wrong. At some point in the future a new theory might replace them, but until then the scientific consensus accepts them as true, because they are the best explanation we have for what we observe in nature.
With the scientific method, there aren't really facts or truths. There are theories that no one managed to disprove yet and so they are accepted as true until someone manages to do so. The scientific method consists of constantly doubting theories.
This is opposed to accepting something, e.g. the word of an authority figure, as a fact or truth. If you accept it as a truth, you won't have the doubt required to look for flaws.
Taking evolution as an example. You can propose the theory that god created all animals the way they are now. You can accept it, because it comes from authority. Or you follow the scientific method, interpret it as a theory, use your doubt and look for ways to disprove it, for example by showing that animals changed even just over the last few hundred years. Once you've disproven the theory, you develop a new theory, evolution, which explains the change and look for ways to disprove that.
The scientific method relies on doubt, the appeal to authority on belief.