Survey Design Framework (Insights Without Bias)

Design a survey for {RESEARCH_OBJECTIVE}.

Input:
- Research objective: {OBJECTIVE}
- Target respondents: {AUDIENCE}
- Decisions this will inform: {DECISIONS}
- Target sample size: {SAMPLE_SIZE}

Rules:
- Start with clear research questions
- Use unbiased, neutral wording
- Mix question types appropriately
- Keep it as short as possible
- Plan analysis approach upfront

Output format:

SURVEY DESIGN BRIEF

Research objective: [What you're trying to learn]

Key research questions:
1. [Specific question you need answered]
2. [Specific question you need answered]

Hypotheses (if applicable):
- [What you expect to find]

Target respondents: [Who will take this]
Sample size goal: [Number needed for validity]
Estimated completion time: [X minutes]

SURVEY INTRODUCTION (what respondents see)

"[Survey title]

Thank you for taking the time to complete this survey. Your feedback will help us [purpose].

This survey will take approximately [X] minutes to complete. All responses are [anonymous/confidential] and will only be used for [purpose].

Please answer honestly—there are no right or wrong answers."

SURVEY QUESTIONS

Section 1: [Section name, e.g., "Background"]
Purpose: [What this section establishes]

Q1: [Question text]
Type: [Multiple choice / Rating scale / Open-ended / etc.]
Options: [If applicable]
- Option 1
- Option 2
- Prefer not to answer [if sensitive]

Required: [Yes/No]
Rationale: [Why asking this]

Q2: [Question]
Type: Likert scale
Scale: Strongly disagree (1) to Strongly agree (5)
Statement: "[Neutral statement to rate]"

[Continue all questions...]

Section 2: [Section name]
Purpose: [What this section measures]

Q5: [Question]
Skip logic: [If Q4 = X, show this; otherwise skip]

[Continue...]

QUESTION TYPE GUIDANCE USED

Rating scales:
- Using consistent 1-5 scales throughout
- Labels on endpoints for clarity
- Neutral midpoint included

Multiple choice:
- Mutually exclusive options
- "Other" with text box where appropriate
- "None of the above" or "Prefer not to answer" when needed

Open-ended:
- Limited to [X] questions to avoid fatigue
- Used only when quantitative won't capture needed nuance

BIAS MITIGATION

Leading questions avoided:
- Instead of "How much do you love our product?" asking "How satisfied are you with [product]?"

Double-barreled questions avoided:
- Not asking "Is our product fast and reliable?" (two questions)

Neutral language:
- Not assuming positive or negative sentiment

Randomization:
- [If applicable] Question order randomized to reduce order bias
- [If applicable] Answer options randomized

SURVEY FLOW DIAGRAM

[Describe logical flow]
All respondents: Q1-Q4
If [condition]: Q5-Q7
If [condition]: Q8-Q10
All respondents: Q11-Q13 (demographics)

ANALYSIS PLAN

For quantitative questions:
- [How you'll analyze - e.g., "Calculate mean satisfaction score overall and by segment"]
- [Statistical tests if applicable]

For qualitative questions:
- [How you'll code and categorize responses]

Segmentation approach:
- [How you'll break down data - by user type, tenure, etc.]

Success criteria:
- [What findings would be significant]
- [Minimum response rate needed: X%]

SURVEY METADATA

Survey platform: [SurveyMonkey / Google Forms / Qualtrics / etc.]
Distribution method: [Email / In-app / Web link]
Incentive: [If offering - e.g., "$10 gift card" or "None"]
Launch date: [Date]
Close date: [Date]
Reminder cadence: [When to send reminders]

TESTING PLAN

Pilot group: [5-10 people from target audience]
Test for:
- Unclear questions
- Technical issues
- Time to complete
- Any missing response options

Revise based on: [Feedback from pilot]

SAMPLE SURVEY QUESTIONS (formatted for platform)

Q1: How often do you use [product]?
○ Daily
○ Weekly
○ Monthly
○ Rarely
○ Never

Q2: How satisfied are you with [product]?
Very dissatisfied  1  2  3  4  5  Very satisfied

Q3: What is the primary reason you use [product]? (Select one)
○ [Reason 1]
○ [Reason 2]
○ Other: __________

Q4: What could we improve? (Open-ended)
[Text box]

Objective: {OBJECTIVE}
Audience: {AUDIENCE}
Sample size: {SAMPLE_SIZE}