A research question that a team can test
Turn an early research idea into a bounded question with a defined population, measurement, comparison, and decision record.
Move from interest to a workable question
A promising research idea often begins broadly: an observation, a surprising pattern, or a gap in explanation. A team can begin work more effectively when that interest becomes a question with clear boundaries. Write down what you want to learn, the evidence that could answer it, and the limits of the available materials.
For a learning project, keep the first question modest enough to examine carefully. An initial literature exercise might ask how a concept is defined across a specified set of papers. A data exercise might examine one prespecified variable in an authorized dataset. Each activity needs a purpose that another team member can explain in their own words.
Define the unit and the comparison
Specify what one observation represents. A person, a visit, an image, an institution, and a published article are different units. Confusing them can make a count appear more informative than it is. Write the unit beside every planned measurement, and check whether repeated observations from the same source require special handling.
Then identify the comparison. You may be comparing methods, groups, periods, or definitions. State why that comparison fits the question. If the project is descriptive, explain what will be described and within which boundaries. A clear descriptive question can be a valuable contribution to a larger research program.
Separate the plan from what you discover
Before examining outcomes, record the main question and intended analysis. The Center for Open Science describes preregistration as specifying a research plan in advance and distinguishes planned analyses from exploration. Use the approach appropriate to the project, its stage, and its governing requirements.
As the work develops, preserve dated amendments. Explain what changed and why, including whether the change occurred before or after you saw relevant results. Exploration can reveal useful new questions; label it accurately so readers understand the path by which an observation arose.
Make a small evidence record
Create a project note with five fields: question, permitted source, unit of observation, intended comparison, and completion criterion. Add the responsible reviewer and the next decision. Keep the note in the approved project workspace, using the permissions and data rules that already govern the work.
Public learning pages are appropriate for methods and general educational examples. Identifiable research records and clinical information belong in their approved systems. For practice, use an explicitly synthetic example or an authorized public source, and label its status at the point of use. A synthetic example should never become a reported study result.
Ask for a reproducible handoff
Before moving to the next stage, invite a teammate to read the question and explain the planned work. Can they identify the source, the measurement, and the decision the output supports? Can they locate the current version of the plan? Any ambiguity is a useful signal to refine the handoff.
The next useful deliverable may be a short protocol note, a source table, or a clearly annotated exploratory notebook. Choose the smallest artifact that carries the question forward with enough context for review. Careful beginnings make later interpretation more transparent and teamwork more dependable.