This one is becoming a classic in my office.
One of my colleagues, who thinks a lot of herself, gave a presentation. On one slide, she had a chart showing the percentage of responses given by a set of people on how they felt about a number of ideas.
Let's say it looked like this:
47% said they supported X
36% said they supported Y
24% said they supported Z
The presenter then said "the majority of respondents supported X" blah blah blah. Er, no, the majority actually didn't support X, it just received the highest level of support. That's probably an easy semantic mistake to make, but she went on to defend X on the basis that most people wanted it.
Incidentally, her approach to leading her team is very much "I, the people".