For creating a survey with some meaning, you really have to have a good idea of what you are looking for. Otherwise you end up with the grammar of the last sentence.

Sometime this year two young students surveyed people on foreign affairs. They wanted to know if people thought our international reach was too much (isolationist) or too little. In reality, these questions have no meaning with questions that direct answers to nonsense conclusions.

Naturally they found some percentage favoring isolationism. Did they force a certain amount of people into that category? I'm sure they must have given some thought on the matter, yet the arguments may have been about the basic premise of more (or less) is better. Perhaps the isolationists just didn't like the confrontational international relationships and concluded less is better when done badly.

When the argument remains contained within one "sphere" of perception strange conclusions result. What lowest level of influence produces a predictive activity? Without getting to the influence level, many survey creators never ask the right questions for marketing or investment decisions.

People sense when you ask the "right"  questions. Not only do you mine into the data you are principally examining, additionally people will tell you what they really think and make sure you know it!

I use surveys to prove my theories. I pontificate occasionally to make sure I can explain philosophies to people who don't care at all about theory. I try to translate into what they understand.

Most people create surveys to create a thesis. I don't believe these "before theory" surveys produce useful activities to implement anything useful. Mostly they are make-work projects since a spreadsheet will get a paycheck.

If we look at the three styles of groups, intellectual, warrior, and tribal, can we say a single question will evoke the same response for everyone? How should a question be framed to get the same result in all three groups? If you examine the results based on one group using the perceptions of another,  can you cause misinterpretations resulting in confusing conclusions.

If you don't know what to ask you probably don't know what's really being examined. Odd conclusions result, many of them emotional arguments. The the point is to sound very serious and, if possible, frighten people with TV appearances.

The next time you get a bad survey, probably on some e-newsletter link, think about  the waste in resources.