The headlines read "Police union members vote no confidence in Chief Lanier." The poll results as reported by local media: "A survey of D.C. Police Union members finds they do not have confidence in Metropolitan Police Chief Cathy Lanier."
But, lets look a little closer at the numbers in the poll:
1150 members responded, with 97.5 percent, or 1,122 answering "No."
But, over 3600 union members were polled. Which means that more than 2/3rds of union members didn't respond.
This is a form of selection bias. Do we know how the non-responders really feel about Chief Lanier? Do we know if we can extrapolate meaningfully from the 1150 who did respond? These are the types of questions you should ask yourself as a consumer of everydata.