Support for Gays in the Military Depends on the Question

http://www.cbsnews.com/blogs/2010/02/11/politics/politicalhotsheet/entry6198284.shtml

In the poll, 59 percent say they now support allowing "homosexuals" to serve in the U.S. military, including 34 percent who say they strongly favor that. Ten percent say they somewhat oppose it and 19 percent say they strongly oppose it.

But the numbers differ when the question is changed to whether Americans support "gay men and lesbians" serving in the military. When the question is asked that way, 70 percent of Americans say they support gay men and lesbians serving in the military, including 19 percent who say they somewhat favor it. Seven percent somewhat oppose it, and 12 percent strongly oppose it.

This is a great example of why statistics are so easy to manipulate.  There are obvious things you can do, like only call old people, but this wording change leads to a non obvious change.  If you told me to guess which wording favored gays more I would have a hard time guessing.  So someone could do a small sample with a bunch of different wordings find which one favored their opinion and then ask that to a larger group.  They could even publish the question and all and it wouldn't seem dishonest.