Can you believe polls

[droidamus2]

Was just reading a Gallup Poll http://www.pollingreport.com/business.htm that is real interesting. It is actually to different polls one of 1009 people and the second of 508 people. What they did is basically ask the same question of each group just worded slightly differently and the received dramatically different results. to paraphrase

1. Would you favor oppose the government giving aid to the big 3 auto companies
Results : 47 Favor 49 oppose 4 unsure

2. Suppose some of the big 3 were certain to fail without government help would you favor or oppose?
Results : 18 Favor 79 oppose 2 unsure

These polls were done over the same period. This goes to show that how you word a polling question and what group you poll no matter how many statistical adjustments you make you may get results the are drastically different. If anybody wants to take a look at the link. I did read this right that they basically asked the same question just reworded it and got two different results.