You're out to dinner with friends, when someone mentions they have a headache. A friend sees you offering another friend an aspirin and says,"Why are you doing that? Those don't work. You should take a walk instead. Those always work for my cousin."
You can tell your friend that for ninety percent of people, aspirin is an effective painkiller. You can go online and print out research papers that back your claim. You can hand them to him the next time you go out to dinner. But that friend still says, "Aspirin doesn't work for my cousin, so I know it doesn't work at all."
Social psychologists call that phenomena "the man who" effect.
For some people, face-to-face testimony seems much more convincing than any statistics. Even though the statistics represent the many different experiences of the people studied, including the people that reported aspirin wasn't an effective painkiller, they still believe personal anecdotes rather than compelling statistical evidence.