Pretty much echoing everything from above, they’re useful to push you in a certain direction, but then you have to use actual research to go any farther.
And as stated above, Case Studies can also be very useful, though they’re a little different then both anecdotes or scientific research. They are very well documented and printed recordings of events, so it’s not “a story about my cousin”, but “a very detailed and immutible* record of my cousin’s state at this time”, and thus far more useful for, say, determining treatments (since they seem from my work to mostly be used in medicine), but they wouldn’t take the place of clinical trials to determine the efficacy of a drug, since they’re only a single data point, and there could be, and probably are, many confounding factors that are unaccounted for. So they’re not quite anecdotes, but there’re also still not useful in the same way solid research is.
*technically they are mutable, but not in the same way that someone’s memory is mutable