{{ summaryTitle }}
{{ summaryValue }}

{{ summaryLine }}

Survey use{{ surveyUseLabel }} Open prompts{{ computation.ok ? values.open_question_count : '—' }} Review cues{{ computation.ok ? values.review_cue_count : '—' }}
The context changes respondent-facing language and analysis segmentation.
Example: understand export workflow friction before roadmap planning.
A narrow audience produces more useful stems and answer options.
Up to eight retained topics are combined with context defaults to prevent accidental coverage gaps.
The selection changes question types, response options, timing, and the response-mix chart.
{{ params.question_count }} questions
Eight to twelve questions usually fit a short survey; longer instruments need a careful pilot.
This affects rating and agreement response anchors only.
{{ params.open_question_cap }}
One through eight open prompts; discovery mode may raise the effective cap to 35% of the survey.
Use a screener only when direct experience affects eligibility or analysis.
{{ params.include_screener ? 'Screener included' : 'No screener' }}
Leave this off for contexts where referral intent would be awkward or misleading.
{{ params.include_nps ? 'Score included' : 'No score' }}
Collect only a segment that supports a planned comparison or decision.
{{ params.include_demographic ? 'Segment included' : 'No segment' }}
{{ values.draft_text }}
#SectionTypeTopicQuestionResponse optionsAnalysis useGuardrailCopy
{{ row.number }}{{ row.section }}{{ row.type }}{{ row.topic }}{{ row.question }}{{ row.response_options }}{{ row.analysis_use }}{{ row.guardrail }}
CheckStatusEvidenceRecommended actionCopy
{{ row.check }}{{ row.status }}{{ row.evidence }}{{ row.action }}
Interview Questions Generator Build a structured interview guide from a role or research brief with tailored questions, timed probes, scoring anchors, and bias checks.
User Story Generator Draft a user story with a clear role, goal and benefit, then turn acceptance checks into structured criteria with readiness or splitting warnings.
Acceptance Criteria Generator Draft testable acceptance criteria from behaviors and edge cases, then check Gherkin or QA handoff wording for missing coverage and vague terms.
Checklist and Log Generator Turn one-item-per-line notes into Markdown checklists or logs while keeping source order and duplicates and adding optional context when needed.

A useful survey begins with a decision, not a list of interesting questions. The research goal should name what the responses will help someone understand or change, and the audience should be narrow enough that respondents share the relevant experience.

Question wording shapes the data. A leading phrase can push answers in one direction, a double question can leave respondents unsure which part to answer, and incomplete response options can force a choice that is not true. Even a neutral question may fail when respondents interpret an important term differently.

Response typeUseful forMain caution
Single or multiple choiceCurrent behavior, drivers, and segmentationOptions must cover realistic answers without overlapping.
Rating or agreement scaleComparable satisfaction, ease, confidence, or agreement measuresEach item should measure one construct with clear end anchors.
RankingForced priorities when tradeoffs matterLong lists increase burden and can produce weak ordering.
Open textExamples, causes, and language not anticipated in fixed optionsIt takes more effort to answer and more work to code consistently.

Questionnaire quality and survey quality are not the same. Clear questions cannot repair a sample that misses part of the population, a collection method that excludes important respondents, or analysis that treats a convenience sample as representative. Privacy also depends on what is collected, how the survey is fielded, and who can see individual responses.

Draft questions should therefore be treated as hypotheses about what respondents will understand. Cognitive interviews and a pilot reveal problems that automated wording checks cannot see, including confusing terms, missing answer choices, recall difficulty, and sensitive questions that people prefer not to answer.

How to Use This Tool:

Define the research decision and respondent experience before choosing a question mix.

  1. Select the Survey use, then enter a specific Research goal and Respondent audience. Replace broad phrases such as “all customers” with the group whose experience matters.
  2. List the Topic areas that the survey must cover. Keep each line focused on one subject; up to eight unique topics are retained and supplemented by context defaults.
  3. Choose the Response mix, question count, scale size, and open-question cap. Add a screener, recommendation score, or segment item only when it changes eligibility or analysis.
  4. Review the question ledger and Design Quality Checklist. Rewrite loaded goals, combined questions, weak response coverage, or excessive burden, then test the revised draft with representative respondents.

Interpreting Results:

Every generated draft remains in a review-required state. The question count, response mix, and completion estimate describe the deterministic draft; they do not validate the instrument or predict response quality.

  • A Pass check means one explicit rule was satisfied, not that the question is unbiased or understood consistently.
  • A Watch check identifies wording, answer coverage, open-text load, or fatigue that deserves revision.
  • An Action check calls for a human decision, including whether a segment item is necessary and how the questionnaire will be piloted.

Technical Details:

The drafting model combines a use-specific topic set with the supplied topics, reserves requested special items, then cycles through a fixed sequence for the selected response mix. The same normalized inputs always produce the same wording and order; no language model or random draw is involved.

Rule Core:

DecisionExact ruleEffect
TopicsSplit on lines, semicolons, or commas; collapse whitespace; remove case-insensitive duplicates; retain at most eight combined supplied and default topics.Questions rotate through the retained topics.
Question countAccept 4 through 25 questions. Requested screener, recommendation, and segment items reserve positions first.The remaining positions follow the response-mix sequence.
Open questionsUse the selected cap from 1 through 8. More Open Discovery may raise the effective cap to 35% of the questionnaire, rounded up and limited to eight.Later open positions become rating or single-choice items after the cap is reached.
Scale anchorsUse a five-point or seven-point satisfaction scale for ratings and the corresponding agreement scale for Likert items.Scale points change response anchors, not topic selection.
Special itemsPlace a screener first; place recommendation and segment items after the main sequence.The total still stays within the selected question count.

Completion time is a workload estimate built from fixed seconds per question type. It rounds the total upward to a whole minute and never displays less than one minute.

Formula Core:

Ts = ∑i=1n ttype(i) , Tmin = max(1,⌈Ts60⌉)
Question typeSeconds
Recommendation score12
Segment16
Screener18
Rating scale or agreement scale18
Single choice20
Multiple choice28
Ranking36
Open text45

The quality checks are conservative prompts for review rather than statistical tests. Fatigue passes only when the estimate is at most six minutes and the draft has at most 18 questions. More than two generated questions containing an and or slash cue triggers a one-concept warning. A supplied goal containing persuasive terms such as “best,” “terrible,” “should,” or “perfect” triggers a neutral-wording warning. Pilot review is always an action, with a recommendation to test with 3 to 5 representative respondents.

Limitations and Privacy Notes:

The draft generator cannot establish construct validity, sampling quality, accessibility, legal compliance, translation equivalence, or respondent safety. Its generic answer options may miss domain-specific choices, and its completion estimate does not model reading ability, skip logic, device use, or sensitive-topic hesitation.

  • Generation happens in the browser and the entered goal, audience, and topics are not uploaded by this tool.
  • The generated introduction's aggregate-review wording is draft copy, not a guarantee about the survey system where the questions are later fielded.
  • Collect only necessary personal data and follow the privacy, consent, retention, and review requirements that apply to the study.

Worked Examples:

A twelve-question product-discovery draft

With the balanced mix, no optional special items, and an open-question cap of two, the 12-position sequence contains two open-text prompts. Its fixed time weights total 312 seconds, so the estimate rounds up to 6 minutes. That passes the built-in fatigue boundary, but the questions still need respondent testing and a review of whether every answer choice fits the audience.

References: