Make Your Question Evidence-Answerable and Measurable
Your question is now narrow and bounded, and that is real progress. But a scoped question can still fail in two quiet ways: it can be one that no evidence could ever actually reach, or one so loosely worded that no two people would measure it the same way. This chapter puts your question through two final checks before you go looking for anything, so you never spend hours chasing evidence that was never there to find.
Tell “Answerable in Principle” from “Answerable in Practice”
In the last chapter you learned to keep only questions that evidence could settle. But “could settle” hides an important gap. A question can be answerable in principle, meaning it is about a real fact of the world, and still be answerable in practice only if the evidence actually exists and you can actually reach it.
Take “How many people in my city skipped breakfast this morning?” That is a count, so in principle it has an answer. In practice, no such record exists and you could never survey everyone, so the question is a dead end as written. Now compare “How many meals did the local food bank serve last month?” That is also a count, but a real record almost certainly exists, and a food bank may even publish it. Same kind of question; completely different odds of ever getting an answer.
The test is simple and you run it before searching, not after. For your scoped question, name at least one realistic place the evidence could actually come from: a public dataset, a published report, a set of observations you could make yourself, or a small survey you could run. If you cannot name a single realistic source, your question is not yet answerable in practice, and you reshape it until it is.
Run a Quick Evidence-Availability Check
That source test is the single most important check, but it is not the only one. Before you commit to a question, widen it into four fast checks, each catching a different way a question quietly becomes unanswerable.
Could this be settled by a count, a measurement, a record, or a set of observations, rather than by opinion alone? If only opinion could answer it, go back and separate out the values part.
Can you name at least one realistic, public source for that evidence? If you cannot name even one, the question fails in practice.
Is the evidence likely to be recent enough and specific enough to fit the boundaries you set? A national figure will not answer a question scoped to your street.
Would two careful people, using the same sources, land near the same answer? If the wording lets them reach wildly different results, tighten it.
If any check comes back “no”, you have found the fix before it cost you a wasted afternoon: name a source, adjust a boundary, or sharpen the wording, then run the checks again.
Add a Measurable Element So Numbers Can Speak
Many honest questions start out fuzzy: “Is the library well used?” or “Is this route to work faster?” To answer them, you have to attach something countable. Something is measurable when you can count it, size it, or put a number on it. Evidence in the form of those numbers, counts, amounts, rates, or measured outcomes, is called quantitative evidence, as opposed to evidence made only of words.
Adding a measurable element turns a vague yardstick into something numbers can settle. “Well used” becomes visits per week, items borrowed per month, or seats occupied at midday. “Faster” becomes minutes door to door, averaged over several trips. The move has two steps:
Pick the countable thing that best stands for what you actually care about: cost, usage, frequency, survey results, or a measured outcome.
Name the unit, so the number is unambiguous: per week, per household, per hour, per trip.
Not every question needs a number. But wherever a decision turns on “how much”, “how many”, or “how often”, a measurable element makes the answer checkable instead of arguable, which is exactly what you want when the stakes are real.
Walk Through a Measurability Check
Take a fresh everyday question. A town is debating whether to keep the public library open on Sundays, and someone asks, “Is Sunday opening really worth it?”
As stated, “worth it” is a values judgment, so first pull out the evidence-answerable part: “How much is the library used on Sundays compared with other days?”
Now run the two checks from this chapter.
Evidence reach: is there a real source? Yes. Libraries routinely record entries and loans, and many publish them; if not, you could count entries yourself across a few Sundays. The question is answerable in practice.
Measurable element: what do you count? Visits per Sunday against visits per typical weekday, items borrowed on Sundays, and seats occupied at midday. Each has a clear unit.
Put it together and the question becomes: “Over the last three months, how many visits and loans did the library record on Sundays compared with a typical weekday?” It is evidence-answerable, it names real sources, and it is measurable. Note the honest-measure caution too: visits count footfall, not benefit, so pairing visits with loans gives a fuller picture than either alone. You now have a question numbers can actually answer.
Let’s Recap!
A question can be about the world and still be a dead end if the evidence does not exist or you cannot reach it.
Test every scoped question by naming at least one realistic, public source that could answer it before you start searching.
Measurable means something you can count or put a number on; quantitative evidence is evidence in the form of numbers.
Where a decision turns on how much, how many, or how often, add a measurable element and name its unit.
Choose the measure honestly, because a number is only as useful as what it truly stands for.
Your question can now be reached by real, measurable evidence. The final framing step is to say why the question matters and what a trustworthy answer must contain, so you know when you are actually done. That is what the next chapter builds.
