3.3 Criteria of a Research Problem

A well-formulated research problem's tests (the checklist): Specific and bounded — one problem, not a province ("the enforcement gap in caste- discrimination cases in district courts of one province", not "discrimination in Nepal"); Researchable/answerable — the evidence exists or is collectable; the question can be answered by data, not only by opinion (the moral sermon and the unframed value question are not research problems); Original or usefully replicable — the literature's check shows the gap (or the replication's purpose: the earlier study's findings tested in a new setting or era); Neutral in statement — the question without the smuggled conclusion ("Untouchability's criminalisation has failed" is a conclusion; "What explains the enforcement gap...?" is a question); Feasible — within the researcher's time, resources, competence and access; Significant — the answer matters (to doctrine, practice, policy or theory); Ethical — the inquiry can be conducted within the research-ethics disciplines (consent, confidentiality, no harm).