Move from Research Question to Defensible Conclusion

Make the question operational

Research begins with a consequential gap, not with a preferred statistical test. State what is already known, what remains uncertain, and what decision the answer could change. For an intervention question, define population, intervention, comparator, outcomes, and time. For an exposure question, replace intervention with exposure. For qualitative inquiry, specify the population, phenomenon, and context. Translate broad concepts into operational variables: how food security, adherence, dietary intake, or quality of life will be measured; when; by whom; and with which instrument. Name one primary outcome that drives the main inference, then prespecify secondary outcomes and harms. A narrow, answerable question protects the study from collecting attractive but irrelevant data and from changing the outcome after results are seen.