Jaywalker
So long, and thanks for all the fish.
But when dealing with 2 extremes, a black and white... a yes and no, a binairy choice, not answering should yield "no result".
Not necessarily, particularly if only one extreme is tested for.
You are suggesting that, given options disagree, neither agree nor disagree, and agree, the response values should be -1, 0, and 1, respectively.
It entirely possible, and quite probable, that the actual values on some questions are actually 0, 1, 2.
The easiest way to test your hypothesis, however, is to look at the strength of the preferences in the results for all neutral responses. If they are all 1%, marginal or no preference, then your hypothesis about the scoring is justified. A different result would indicate the scoring to be more complex than you hypothesize.