Rewriting Research Questions
YAML--- name: rewriting-research-questions description: Transforms broad, vague, or yes/no research questions into focused, open-ended, researchable questions by narrowing scope, adding boundaries, and reframing with analytical language. Use when a research question is too broad, too narrow, closed-ended, unfocused, or needs refinement for a paper, thesis, or study proposal. --- # Rewriting Research Questions
Given a weak research question, apply this transformation:
Input: "Do teachers use tablets in classrooms?" Output: "How do middle school teachers in under-resourced districts integrate tablets into daily lesson planning?"
Changes made: closed yes/no → open "how"; added population (middle school teachers), context (under-resourced districts), and specific focus (daily lesson planning).
If given a bare topic instead of a question, first convert it to a question, then apply the same narrowing process.
Progress:
- Step 1: Diagnose the problem with the current question
- Step 2: Identify the core gap, tension, or relationship
- Step 3: Reframe closed questions as open-ended (how/why/to what extent)
- Step 4: Add boundaries (population, place, time, variable)
- Step 5: Check feasibility and answerability
- Step 6: Present rewritten question(s) with rationale
Step 1 — Diagnose. Identify which failure mode(s) apply:
- Too broad (no boundaries, entire field in scope)
- Too narrow (no literature/data will exist)
- Closed/yes-no (answerable in one word)
- Purely descriptive with no analytical angle
- Vague or ambiguous terms ("effective," "impact," "better")
- Not researchable (unfalsifiable, purely normative/opinion-based)
Step 2 — Find the core gap. Ask: What specific tension, unknown relationship, or missing piece of knowledge is actually being pursued? Strip away the general topic to find the real angle of interest.
Step 3 — Reframe as open-ended. Replace "Is/Do/Does/Are" openers with:
- "How" — process, mechanism, experience
- "Why" — causes, motivations, explanations
- "What" — characteristics, factors, relationships (when paired with a clear object)
- "To what extent" — degree, strength of relationship
Step 4 — Add boundaries. Insert at least two of:
- Population/sample (who)
- Geographic/institutional context (where)
- Timeframe (when)
- Specific variable or outcome (what exactly)
Step 5 — Check feasibility. Verify:
- Data/sources are realistically accessible
- Scope fits the time and resources available (course paper vs. dissertation vs. systematic review)
- Question isn't so narrow that no literature exists to support it
- Question isn't so broad it can't be answered in the given format
Step 6 — Present output. Give 1–3 rewritten versions (e.g., broad/medium/narrow, or quant/qual framing) with a one-line rationale for each, plus note what still needs refining.
Example 1: Input: "Is social media bad for teenagers?" Output: "How does daily Instagram use affect self-reported body image among female high school students aged 14–18?" Rationale: Converted yes/no to "how"; replaced vague "social media" with a specific platform; replaced vague "bad" with a measurable construct (self-reported body image); added population and age boundary.
Example 2: Input: "Climate change and agriculture" (bare topic) Output: "To what extent has increased drought frequency affected smallholder maize yields in sub-Saharan Africa between 2010 and 2023?" Rationale: Converted topic to question; added variable (drought frequency), outcome (maize yields), population (smallholder farmers), region, and timeframe.
Example 3 (too narrow → broadened): Input: "What is the exact correlation coefficient between left-handedness and chess rating among 12-year-old players in one specific school?" Output: "Is there a relationship between handedness and competitive chess performance among youth players (ages 10–14)?" Rationale: Original was unanswerable due to insufficient data pool; widened population and dropped overly specific single-school constraint while keeping focus.
Example 4 (descriptive → analytical): Input: "What are the symptoms of burnout in nurses?" Output: "Why do nurses in high-acuity ICU settings report higher burnout rates than those in outpatient settings?" Rationale: Shifted from purely descriptive (listing symptoms) to explanatory/comparative, which supports deeper analysis and a defensible research design.
- Default to "how" or "why" openers when the goal is analysis, not just description.
- Add boundaries incrementally — two or three specific constraints are usually enough; too many makes the question unanswerable with available data.
- Keep one key variable or relationship at the center; avoid stacking multiple unrelated questions into one.
- Match question scope to the project type: a course paper needs tighter scope than a dissertation; a dissertation needs tighter scope than a field-wide review.
- When uncertain whether the question is too broad or too narrow, do a 5-minute literature gut-check: too many hits = broad, too few = narrow.
- Preserve the researcher's original interest/angle — narrow the scope, don't replace their topic with a different one.
- Offer multiple rewritten variants when the original intent is ambiguous, rather than guessing a single "correct" version.
- Don't just tack on jargon ("exploring the nuanced interplay of...") without actually adding boundaries — that's padding, not narrowing.
- Don't over-narrow to the point no data or prior research exists to support the study.
- Don't leave "yes/no" phrasing hidden inside a longer sentence (e.g., "Does X have an impact on Y" is still closed even if wordy).
- Don't conflate a research question with a thesis statement, hypothesis, or title — a research question stays a question and remains open until the research answers it.
- Don't ignore feasibility; an elegant question the researcher cannot actually investigate (no access to population, no time, no data) is still a bad question.
- Don't rewrite in isolation from the researcher's actual purpose/discipline — a question reframed for a quantitative study looks different than one for a qualitative or systematic review context.