Imagine being asked these two questions: 

1. If legitamacy is defined as "the will of the people," do you think George W. Bush is a legitimate president? 
2. When George W. Bush is sworn in as president, will he be a legitimate president? 

Only 50% of the people said yes to the first question.  70% said yes to the second. Why the discrepancy?  Because research depends on asking the right kinds of questions. When you research, you have to define what you are looking for, but you must ask questions that assure your readers that your results are reliable.