A central position: indecision, trade-off or refusal of the frame?

The first situation is that of the person who has no formed position on the question, for want of having thought about it, for want of information, or because the subject does not concern them. The middle is here a sincere answer, and treating it as a moderate opinion would be an error of interpretation.

This situation is frequent and perfectly legitimate. Nobody has a considered view on every subject, and an honest questionnaire must leave room for it rather than forcing a choice that would manufacture a non-existent opinion.

The second is that of the person who has examined the question and gives weight to both sides, because they see a real interest on each side or because they consider that the answer depends on the case. The middle here reflects a constructed position, often more worked through than a clear-cut one.

These first two situations give the same point on the scale and cannot be distinguished in the raw result. What distinguishes them is stability: a trade-off position is found identical a few weeks later, an indecision shifts with the news.

A second indication separates them: response time. An indecision is settled quickly, for want of material to weigh; a trade-off takes longer, because there is something to balance.

The third is that of the person who rejects the way the question is put: they consider that the alternative offered is not the right one, that the vocabulary is badly chosen, or that the real subject lies elsewhere. Unable to say so, they place themselves in the middle, and their answer is then the most badly interpreted of the three.

That is why a central position deserves a cautious commentary. Announcing to someone that they are moderate when they were rejecting the question is the best way of losing their trust, and it is an error many tests commit with assurance.

The statistical centre of a scale has no ideological content: it is a landmark, which shifts with the population measured. Being in the middle of a distribution therefore does not mean defending an intermediate position, but standing where most of the answers are.

Conflating the two produces a frequent error in commentaries on results: treating the average position as wisdom, or on the contrary as lukewarmness. Neither can be deduced from a measurement.

Go back to a question where you answered in the centre and ask yourself three questions. Could you name the two opposing arguments in play? Would you have answered the same six months ago? And if you could rewrite the question, would you change anything?

Three yeses designate a trade-off. A no to the first designates an indecision. A yes to the third designates a refusal of the frame, and it is the only one of the three situations that deserves to be reported to whoever wrote the question.