Or else it's biased. It's easy to fool yourself with a large number of anecdotes (listening only to like-minded people, for example), but "data" implies some di

Or else it's biased. It's easy to fool yourself with a large number of anecdotes (listening only to like-minded people, for example), but "data" implies some directed collection was done and some thought was given as to what you're actually sampling.