No, as any seasoned sociologist will tell you having a middle option (or N/A) is a prerequisite to having accurate results.
How often is there an issue where I don't have an opinion? Not often. But it's not that I have or don't have opinions on issues, it's that the opinion I have may be irrelevant to the context of the survey or that the question is incorrectly worded (or 'loaded', as we would say). For example:
The question presupposes an invalid premise. The correct answer is N/A. As for political questions, the compass asks:
Now I do have an opinion on this issue. But the opinion is too complex to express with a simple agree/disagree response format. I think that it's worrying from a social perspective but that doesn't at all imply that I think the government should do anything about it. Answering 'agree' there most certainly shifts me down the authoritarian axis, which is incorrect and exactly opposite of my philosophy. Ergo the results are now inaccurate.
Then there's the loaded questions:
Who honestly would disagree with such a thing? But from the perspective of government and law, my philosophy is one of non-interference. So agreeing or disagreeing is again inaccurate.
And that's just the first page.
|