What a conversation knows that a form never will

Picture weighing an office move, and running a survey to test the water. The result lands: sixty-eight per cent against. Now you are cornered. Push on and you have overruled a clear majority. Back down and you have lost the opportunity, perhaps a better building or a lower rent, on the strength of a number. The figure has boxed you in before you have learned anything useful.

And it has hidden the only thing worth knowing. Ask the same people to talk, and the sixty-eight per cent comes apart into something far more usable. Most of them are not against the move at all. They are worried about the commute. The school run. Whether the new place has anywhere quiet to take a call. That is not a vote against the office. It is a list of solvable problems, and it is a completely different brief for the person making the decision. With it in hand you can keep two days remote, add a travel allowance, design in some quiet rooms, and make the move you were always going to make, with people feeling met rather than overruled.

This is the structural difference between a form and a conversation, and the research backs it two ways. First, open questions surface what closed ones cannot. A multiple-choice list can only ever collect the answers the question-writer already thought of. An open question can turn up the reason that was never on the list. Survey methodologists have understood this since the 1940s: open-ended questions reveal motivations and explanations the researcher did not anticipate. A rating scale measures how strongly people feel about your idea. An open question can tell you that you were measuring the wrong thing.

Second, a conversation can follow up, and that quietly dissolves problems that wreck surveys. Take the double-barrelled question, the one that asks about two things at once, like how you feel about the price and the quality. On a form it is one of the cardinal sins of question design, because a single answer cannot tell the two apart and you never know which half the person meant. In a conversation it barely matters, because the interviewer can simply ask which part they were talking about, and chase each one. The flaw that ruins the survey is harmless the moment a follow-up is possible.

The form flattens everything into a single figure. The conversation branches toward the reason. And the reason is the part you can actually act on. A percentage tells you where people landed. Only a conversation tells you why they landed there, which is the only part you can do anything about.

References Foundations of Social Work Research, Designing effective questions and questionnaires (open-ended questions reveal unanticipated motivations).

CultureMonkey, Double-barrelled survey questions (the flaw is far less of a problem in conversational, follow-up settings).

For the wider trade-off, see Lazega et al., Open vs closed-ended questions in attitudinal surveys.