What makes a question good?

It was an experience I had dozens of times during my pregnancies. The nurse asked “Do you have any allergies?”

I answered without hesitation: “Hay fever, and pets.” Occasionally they followed up “Any medication allergies?” I always answered “Not that I know of” – I didn’t really want to get locked in if it turned out I was allergic to some weird medication I’d never tried before.

Then one day I was reading the nursing page on reddit. A poster commented how funny it is that patients mention environmental allergies. One person joked “I promise we won’t treat your broken arm with cats.”

It hadn’t occurred to me before that I was potentially over-answering the allergy question. To me, my answers made perfect sense. To the nurses on that thread, it was clear that they were interested only in the allergies they might trigger through my treatment, not more general allergy issues.

Depending on the medical setting, my answer still makes sense – allergy management is one thing medical professionals can help with. In fact, increased allergies were among the few treatable issues I had during pregnancy.

Their question also probably makes sense even when patients over-answer. From the nurse point of view, they needed to know all allergies that might impact my care. They wouldn’t want to ask “Any medication allergies?” and have a patient leave out aspirin because they think of it as a “drug” not a “medication”.

The goal of a question is a transfer of knowledge. Since people don’t typically do things for no reason, there should be some kind of purpose attached to the questioner’s inquiry. We can use the future achievement of that purpose to define a “good” question.

Turning this definition into a set of rules for defining good questions is not easy. Here is my attempt:

• A good question states what information the questioner is seeking (goals, possible tradeoffs, details of the options, expected outcomes, etc). These are the components I use in my Meta-Problem Method, which is how I’ve taken to framing decisions.

• A good question includes the relevant context.

• A good question includes a minimum of irrelevant information.

• A good question which for whatever reason can’t cover the above, should explain why. For example: “I don’t know if this information is relevant, but it might be?”

These four rules in some sense boil down to “a good question asks precisely what you want to know.” But, especially when we are asking for advice, we include extra information so the answerer can act as our agent.

When your question directly relates to a future decision, that decision becomes a guide for identifying good questions.