A study sponsored by the marketer of password management software, by definition, isn't the most independent research in the world. But I've seen hundreds of these sponsored studies and the data is usually pretty good. It is hard to get researchers to fake it, as their reputations are too much on the line.
