Design a Household Survey
Create a structured survey questionnaire for household or beneficiary data collection.
Draft: Design a Household Survey
Your Prompt
- What You Need
- Share the survey purpose, target households and respondents, geography, indicators or topics, collection mode, and privacy constraints.
- Choose This Prompt When
- Use this for a general household questionnaire. Choose Design a Health Household Survey or Design a WASH Household Survey for those sector-specific instruments, or Design a KoboToolbox/ODK Survey Form for platform-ready form logic.
Describe your goal, audience, available information and constraints. Your words are added to the prompt as supplied. You can also use the guided fields below.
Kept in this tab while you work. Copy and Start includes this description.
Copy this prompt and paste it into your AI chat to start. If it needs source material, the AI will ask for it.
Evidence and applicability: use only supplied or verifiable evidence for historical facts, values, quotations, completed activities and approvals. Preserve conflicting sources with their provenance; do not choose a preferred value without adjudication. Mark missing evidence as unknown or not evidenced, not as proof an activity never occurred. Keep proposed plans and illustrative examples separate from findings. Never invent a baseline, quote, pilot, validation or sign-off to complete a template. Match methods to the purpose and population; a suitable single-method design is valid. Distinguish descriptive change from causal attribution. Check obligations in the applicable funding agreement rather than assuming a universal donor rule.
You are a senior MEAL specialist. Design a household survey for the user's program using its stated purpose, population, geography, constructs, collection mode, and data protection needs. Do not presume a sector, age range, topic set, question count, response format mix, sampling frame, or anonymity.
Design requirements:
- State the survey purpose, intended population, unit of observation, geography, eligibility, and decisions the data should inform. Distinguish a household's characteristics from an individual respondent's answers.
- Draft only the questions needed for the priority constructs. For each question, give the exact wording, response options or recording instruction, reference period where needed, and linked measure or analysis purpose. Choose closed, scaled, or open response formats only when each suits the information sought. Avoid leading, double-barreled, and unnecessary sensitive questions.
- Organize a respondent-friendly flow with eligibility, consent, core sections, and closing. Specify routing and skip conditions where needed. Avoid assuming one household member can reliably report private facts about another.
- Include disaggregation items only if justified by the purpose, safe to ask, and analytically usable. Minimize personal data and explain how missing, refused, and not-applicable answers will be recorded separately.
- Give enumerator instructions for recruitment, voluntary informed consent, refusal or withdrawal, neutral probing, privacy during interview, and escalation of distress or safeguarding issues where relevant. Describe confidentiality protections that the actual data plan can support. Promise anonymity only if no identifiers or linkable records will be collected.
- Propose appropriate sampling and recruitment steps based on an actual frame, coverage, feasibility, and analysis goal. If these details are unknown, give options and the decisions needed, not a falsely settled methodology or sample size.
- State a feasible pretest and quality plan, such as cognitive review, pilot, training, supervision, and validation checks. Label all as planned until completed. Provide a short analysis plan for descriptive estimates or comparisons the design can support; do not claim impact attribution from a single cross-sectional survey.
- Outline a high-level implementation sequence and timing dependencies without inventing dates.
If essential material is missing: Ask for the survey purpose, target population and geography, core topics or indicators, collection mode, and data protection constraints before drafting a field-ready instrument. If the user asks for a starter, provide a clearly labeled draft plus decisions and pretest requirements.
Output: A purpose and population note; a numbered questionnaire with response codes and skips; enumerator and consent guidance; proposed sampling, quality, privacy, and analysis plans; a high-level implementation sequence; and open design decisions. Keep length proportionate to the measurement needs and respondent burden.
Check: Score It with the Survey Instrument Review Rubric
The Review Prompt
Paste the draft at the bottom of this prompt and run it. It scores each criterion from 1 to 5 and says why.
Establish the document purpose, design, stage, applicable requirements, and available source evidence before rating. Distinguish N/A with a reason, not evidenced, unresolved conflicting evidence, and observed inadequate performance. Missing records do not prove an activity did not occur. Missing records alone justify evidence requests and a deferred readiness decision, not an inferred failure or mandatory repeat activity. Do not invent documentary, signature or approval requirements. Cite document locations for every finding. Keep supported facts separate from proposals, assumptions, and illustrative examples. Never invent completed fieldwork, validation, consent, approval, values, dates, quotations, or assigned ownership. Label unconfirmed owners and proposed revisions for confirmation. Report blocking ethics, safety, confidentiality, measurement-validity, or inference failures first regardless of scores. Give a total and band only when all five dimensions are applicable and sufficiently evidenced with no unresolved evidence conflict; otherwise show the dimension profile, coverage, and evidence requests without a total, band, or rescaled equivalent. A documentation score never authorizes use, fielding, compliance, or approval.
You are an expert M&E data specialist. Score the survey instrument I will provide using the rubric below.
SCORING RUBRIC - Survey Instrument Review Rate applicable dimensions 1-5. For item-rate anchors, report defective/applicable reviewed items. Zero defects among applicable reviewed items earns 5; no applicable items is N/A. Blocking defects override the rate:
DIMENSION 1: Question Design
- Score 5: All questions use one-topic-per-question format. No leading or loaded words. Language matches target population literacy level. Single-select options are non-overlapping and sufficiently exhaustive; multi-select items allow legitimate co-selection and prevent duplicate coding. Sensitive-item placement is justified by privacy, rapport, routing, and safety; no universal last-section rule applies.
- Score 4: More than 0% and at most 10% of reviewed applicable items have defects in this dimension; no blocking defect.
- Score 3: More than 10% and at most 20% of reviewed applicable items have defects in this dimension; no blocking defect.
- Score 2: More than 20% and at most 50% of reviewed applicable items have defects in this dimension; no blocking defect.
- Score 1: More than 50% of reviewed applicable items have defects, or a critical item has a blocking defect. Report the blocking issue separately; no aggregate approval follows.
DIMENSION 2: Instrument Structure
- Score 5: Introduction covers all five elements (who is collecting, purpose, data use, time estimate, voluntary participation). Questions ordered general to specific within sections. Skip logic present, clear, and consistently applied. Length appropriate for context.
- Score 4: Introduction covers four of five elements. Question order logical. Skip logic present with no more than two minor inconsistencies.
- Score 3: Introduction covers three of five elements. Question order mostly logical with grouping issues. Skip logic exists for major branches but is missing or unclear for minor ones.
- Score 2: Introduction covers two or fewer elements, OR skip logic is absent or systematically inconsistent, OR survey includes questions not needed for indicator measurement.
- Score 1: No introduction. No logical structure. No skip logic. Questions not traceable to indicator needs.
DIMENSION 3: Ethical Standards
- Score 5: The consent procedure covers six elements: purpose, voluntariness, refusal/withdrawal, confidentiality and its limits, contact route, and storage information. Applicable safeguards are tailored. Assess the script as a design artifact; completed consent requires separate records.
- Score 4: Five of six elements are usable and core protections are intact; one noncritical detail needs clarification. A material consent defect is a separate blocker.
- Score 3: Three or four elements are usable, including purpose, voluntariness, and appropriate confidentiality information; material gaps require correction.
- Score 2: Only one or two elements are usable, or a core protection is materially inadequate. Identify blocking consent/safeguarding failures separately.
- Score 1: No usable consent procedure is present in the complete instrument, or it is coercive/materially misleading. Missing completed-consent records do not prove consent was never sought.
DIMENSION 4: Indicator Alignment
- Score 5: Every question maps to a specific indicator or stated data need. No orphan questions. Every required indicator has at least one question.
- Score 4: At least 80 percent of questions trace to indicators. The remainder are useful for context but not explicitly linked. All core indicators have coverage.
- Score 3: Half or more questions trace to indicators, with up to four orphan questions. Most required indicators are covered, but one or two have weak or indirect coverage only.
- Score 2: Less than half of questions trace to indicator requirements, OR several required indicators have no corresponding question.
- Score 1: No traceable link between questions and program indicators.
DIMENSION 5: Pretesting and Adaptation
- Score 5: Before testing, a usable proposed plan specifies target-population scope and selection rationale; after testing, actual records show pretesting with documented scope and selection rationale appropriate to relevant participant characteristics and measurement risks. Findings, revision or no-change decisions, and any further testing needed are documented. Enumerator preparation includes practice appropriate to the administration method. Participant count alone does not establish validity or readiness.
- Score 4: Before testing, the proposed plan specifies pretesting scope, participant-selection rationale, and how findings and revision or no-change decisions will be recorded; after testing, these are documented from actual records. Relevant participant coverage or follow-up testing detail is partly documented. Enumerator guidelines are present where applicable.
- Score 3: Stage-appropriate pretesting plans or actual records are mentioned but only partly documented: planning details or evidenced findings lack specific revision-decision procedures or notes, OR enumerator guidelines are generic rather than instrument-specific. Completed testing is not required to assess a prospective plan.
- Score 2: Required testing/preparation is inadequately specified for the declared stage. An unavailable external test record is not evidenced, not proof no test occurred.
- Score 1: Complete supplied records show fielding without required testing/preparation, or testing is falsely claimed. Missing external records are not evidenced. Before testing, assess the proposed plan and defer readiness.
ADDITIONAL TASK: List every specific question that fails Dimension 1 (Question Design). For each, state the problem type (leading / double-barreled / ambiguous / wrong literacy level / response options) and provide a corrected version.
OUTPUT FORMAT:
| Dimension | Score (1-5), N/A, or not evidenced | Key Finding | Priority Action |
|---|---|---|---|
| Question Design | |||
| Instrument Structure | |||
| Ethical Standards | |||
| Indicator Alignment | |||
| Pretesting |
Blockers and evidence gaps: Report these first, with source locations. Coverage: State applicable and sufficiently evidenced dimensions out of five. Total: X/25, only when all five dimensions are applicable and evidenced with no unresolved evidence conflict. Otherwise omit the total and band. Optional descriptive band (eligible totals only; never approval): Strong (22-25) / Adequate (17-21) / Needs Revision (11-16) / Substantial Revision (5-10) Single Most Important Revision: [One sentence]
Then list flagged questions with corrections.
SURVEY INSTRUMENT TO SCORE: [Paste your questionnaire here]
Fix: Revise What Scored Low
The Revise Prompt
Paste the draft and its scores at the bottom of this prompt. It revises the parts that scored below 4.
Establish the document purpose, design, stage, applicable requirements, and available source evidence before rating. Distinguish N/A with a reason, not evidenced, unresolved conflicting evidence, and observed inadequate performance. Missing records do not prove an activity did not occur. Missing records alone justify evidence requests and a deferred readiness decision, not an inferred failure or mandatory repeat activity. Do not invent documentary, signature or approval requirements. Cite document locations for every finding. Keep supported facts separate from proposals, assumptions, and illustrative examples. Never invent completed fieldwork, validation, consent, approval, values, dates, quotations, or assigned ownership. Label unconfirmed owners and proposed revisions for confirmation. Report blocking ethics, safety, confidentiality, measurement-validity, or inference failures first regardless of scores. Give a total and band only when all five dimensions are applicable and sufficiently evidenced with no unresolved evidence conflict; otherwise show the dimension profile, coverage, and evidence requests without a total, band, or rescaled equivalent. A documentation score never authorizes use, fielding, compliance, or approval.
You are an expert M&E data specialist. Review the draft questionnaire I will provide and fix every problem that falls below the Good (4) threshold on the Survey Instrument Review rubric.
WHAT TO LOOK FOR AND FIX:
Question Design problems:
- Double-barreled questions - split into two separate questions
- Leading questions - rewrite as neutral
- Loaded or assumptive language - remove the assumption and rewrite
- Technical jargon the target population likely does not use - replace with plain language
- Response options that overlap unintentionally or have relevant gaps - for single-select items, make options non-overlapping and sufficiently exhaustive; for multi-select items, preserve legitimate co-selection and clarify instructions
Structure problems:
- Missing or incomplete introduction - add or complete with all required elements (who, purpose, use, time, voluntary)
- Sensitive-item order creates a privacy, rapport, or safety risk - revise it with rationale
- Missing or unclear skip logic - add explicit skip instructions
Ethical problems:
- Absent or incomplete consent - add a full informed consent statement
Indicator alignment problems:
- Questions not traceable to any indicator - either cut them or note what indicator they serve
- Required indicators not covered - add questions to fill the gap
Pretesting and adaptation problems: Request actual testing, revision-decision, and enumerator-preparation records when missing. Propose further testing with a justified scope where needed. Keep proposed work separate from completed work; do not invent pilot participants, findings, changes, or training.
OUTPUT FORMAT:
- List every problem found by dimension, with original text and corrected version.
- Retain unsupported facts as evidence requests and mark proposed owners, dates, and procedures for confirmation. Then provide the complete revised questionnaire incorporating all corrections.
DRAFT QUESTIONNAIRE TO REVISE: [Paste your draft questionnaire here]