Asking a testable question

A broad question can open a subject. A testable question helps you decide what evidence to seek next. It identifies something observable and leaves room for an answer that differs from your expectation. Refining a question is part of the investigation itself.

Start with something you noticed

Begin with a description rather than an explanation. Perhaps a paper bridge sags when a small weight is placed at its centre. Write down what you saw, including the shape, support positions, and loading conditions. You do not need to decide immediately whether the important feature is thickness, folding, or the distance between supports.

The first question might be, “What makes a paper bridge stronger?” That is a useful starting point, but “stronger” needs a meaning before you can compare answers. Does it mean less bending under the same load, a larger load before failure, or recovery after the load is removed?

Name the comparison and the outcome

A more focused question might be: “For strips cut from the same sheet, how does one lengthwise fold affect the centre's downward movement under a fixed small load, with the same distance between supports?” This names a change, a measurable outcome, and several conditions to keep consistent.

It also suggests the record you would need: the geometry of each strip, the position and size of the load, and the displacement measured in a defined way. A sketch can clarify the setup more quickly than several vague sentences. The example is a planning exercise, not a claim about which design will perform better.

Describe what would change your mind

A prediction should expose an idea to comparison with observations. If you expect the folded strip to bend less, decide how you would interpret similar movement, greater movement, or inconsistent results. Do this before collecting the measurements.

One observation rarely settles every explanation. The test may reveal a problem in the measurement, a condition you overlooked, or a difference worth investigating further. Keep the question narrow enough that each possible result suggests a useful next step.

Match the method to the question

A simple controlled comparison is one approach. Other questions call for observations over time, comparisons of existing records, or model predictions checked against data. You cannot deliberately vary every phenomenon you want to understand. The method should make clear which conclusions the available evidence can support.

Consider feasibility early. Can you measure the outcome with enough detail to distinguish the possibilities? Can you repeat observations without changing the setup unintentionally? Are the conditions safe and appropriate? Questions involving people, animals, or sensitive information need the relevant permissions and oversight before any collection begins.

A question worth keeping

Read your draft aloud and check four things:

Someone else can identify what you intend to observe.
The key terms have practical definitions.
The comparison and its conditions are explicit.
Different possible results would lead you to reconsider the explanation.

Save the broad question beside the focused one. It records why the work matters, while the narrower question gives you a manageable place to begin. After the first investigation, revisit both. A clearer next question can be a valuable result.