Questionnaire Design: Types, Principles & Best Practices

What is a Questionnaire?

A questionnaire is a structured set of questions designed to collect data from respondents systematically. It is a primary data collection tool used in surveys, interviews, and experiments to gather quantitative or qualitative information.

Why is Questionnaire Design Critical?

• Poorly designed questionnaires lead to biased, incomplete, or unusable data.
• A well-designed questionnaire ensures accuracy, reliability, and validity of research findings.
• It helps in standardizing data collection, making comparisons easier.

Types of Questionnaires

Questionnaires can be classified based on structure, administration method, and purpose. Below is a comparison table:

Key Principles of Questionnaire Design

A well-designed questionnaire follows these 7 golden rules:

1. Clarity & Simplicity

   • Use simple, unambiguous language (avoid jargon).
   • Example: Instead of "To what extent do you agree with our service’s efficiency?", ask: "How satisfied are you with our delivery speed?" (Scale: 1-5).

2. Relevance

   • Every question should directly relate to research objectives.
   • Example: If studying customer retention, avoid questions about product packaging unless relevant.

3. Avoid Bias

   • Leading questions (e.g., "Don’t you think our service is the best?") skew responses.
   • Double-barreled questions (asking two things at once) confuse respondents.
     • ❌ "Do you like our app’s speed and design?"
     • ✅ "How satisfied are you with our app’s speed?" (1-5) + "How satisfied are you with our app’s design?" (1-5)

4. Logical Flow

   • Start with easy, general questions, then move to specific/demographic details.
   • Example:
     1. "How often do you use Pathao?" (Daily/Weekly/Monthly)
     2. "What features do you use most?" (Check all that apply)
     3. "What improvements would you suggest?" (Open-ended)

5. Response Options

   • Closed-ended questions (multiple-choice, Likert scale) are easier to analyze.
   • Open-ended questions provide qualitative insights but require manual coding.
   • Example of a Likert scale (1-5): "How likely are you to recommend Khalti to a friend?" 1 (Not at all) → 5 (Extremely likely)

6. Avoid Overloading

   • 10-15 minutes max for online surveys; shorter for phone/paper.
   • Example: A Daraz order tracking survey should not exceed 8 questions.

7. Pilot Testing

   • Test the questionnaire on a small group before full deployment.
   • Example: Before launching a NTC customer satisfaction survey, test it with 10 employees for clarity.

Question Formats & Their Uses

1. Open-Ended Questions

• Definition: Respondents answer in their own words.
• When to Use:
  • Exploring new topics (e.g., "What challenges do you face using eSewa?").
  • Gathering qualitative data (e.g., "Why did you switch from Ncell to NTC?").
• Example: "Describe your experience with Daraz’s return policy."

2. Closed-Ended Questions

• Definition: Fixed response options (multiple-choice, yes/no, rating scales).
• When to Use:
  • Quantitative data (e.g., "How often do you use WhatsApp for business?").
  • Large-scale surveys (e.g., "Are you satisfied with Nabil Bank’s loan process?" Yes/No/Maybe).
• Subtypes:
  • Multiple Choice: Only one correct answer allowed.
    • Example: "Which payment method do you prefer?" (Khalti/eSewa/Cash/Card)
  • Check-all-that-apply: Multiple answers allowed.
    • Example: "Which features do you use in Google Maps?" (Check all that apply: Directions/Routes/Traffic/Reviews)
  • Likert Scale: Measures degree of agreement (1-5 or 1-7).
    • Example: "How would you rate Pathao’s driver reliability?" (1=Poor, 5=Excellent)

3. Dichotomous Questions

• Definition: Only two response options (Yes/No, Agree/Disagree).
• When to Use:
  • Quick screening (e.g., "Have you ever used a digital wallet?" Yes/No).
  • Binary data (e.g., "Do you have a bank account?" Yes/No).
• Example: "Have you faced any issues with NEPSE’s trading platform?" (Yes/No)

4. Ranking Questions

• Definition: Respondents order options by preference.
• When to Use:
  • Prioritizing features (e.g., "Rank these Daraz delivery options from fastest to slowest").
  • Market research (e.g., "Rank these phone brands by importance: Ncell, Smart, Nepal Telecom").
• Example: "Rank the following eSewa services by how often you use them:"
  1. Bill Payment
  2. Money Transfer
  3. Recharge

5. Scaling Questions

• Definition: Measures intensity of response (e.g., satisfaction, agreement).
• Types:
  • Likert Scale (1-5 or 1-7).
  • Semantic Differential Scale (e.g., "Fast" (1) ———— "Slow" (5)).
• Example (Semantic Differential): "How would you describe Daraz’s customer service?" Poor (1) 2 3 4 5 Excellent

Common Mistakes in Questionnaire Design

Step-by-Step Questionnaire Design Process

flowchart TD
    A["Define Research Objectives"] --> B["Review Literature"]
    B --> C["Develop Questions"]
    C --> D["Pretest Questionnaire"]
    D --> E["Revise Based on Feedback"]
    E --> F["Finalize & Deploy"]
    F --> G["Collect & Analyze Data"]

Step 1: Define Research Objectives

• What problem are you solving?
  • Example: "How can eSewa improve user experience for bill payments?"
• What data do you need?
  • Quantitative (numbers) or qualitative (opinions)?

Step 2: Review Literature

• Check if similar questions have been asked before.
• Example: If studying customer loyalty, review past surveys from Nabil Bank or Global IME.

Step 3: Develop Questions

• Start with broad questions, then narrow down.
• Example for Daraz delivery delays:
  1. "How often do you experience delayed deliveries?" (Never/Rarely/Often)
  2. "What are the main reasons for delays?" (Open-ended)
  3. "Would you switch to another platform if delays persist?" (Yes/No)

Step 4: Pretest the Questionnaire

• Test with 5-10 respondents (colleagues, friends, or a small sample).
• Ask:
  • "Was any question confusing?"
  • "Did you skip any questions? Why?"

Step 5: Revise & Finalize

• Remove ambiguous or redundant questions.
• Ensure logical flow (easy → difficult, general → specific).

Step 6: Deploy & Analyze

• Distribute via online tools (Google Forms), paper, or interviews.
• Example: NEPSE investor survey → Deploy via email with a Google Form link.

Real-World Applications in Nepal

1. eSewa: Improving User Experience

• Problem: High customer complaints about failed transactions.
• Solution: Designed a structured questionnaire with:
  • Closed-ended: "How often do transactions fail?" (Never/1-2 times/3+ times).
  • Open-ended: "What steps should eSewa take to reduce failures?"
• Outcome: Identified bank integration issues as the main problem → led to API improvements.

2. Daraz: Customer Satisfaction Survey

• Problem: High cart abandonment rate.
• Questionnaire Design:
  • Likert Scale: "How satisfied are you with checkout speed?" (1-5).
  • Ranking: "Rank these checkout issues by importance" (Payment options, Loading speed, Security).
• Result: Found slow loading speed was the top issue → optimized server response time.

3. Ncell: Network Quality Feedback

• Problem: Customers complaining about dropped calls.
• Questionnaire Approach:
  • Dichotomous: "Do you experience dropped calls?" (Yes/No).
  • Open-ended: "Describe the situation when calls drop."
• Insight: Most drops occurred in hilly areas → led to tower upgrades in remote regions.

4. Nabil Bank: Loan Repayment Behavior

• Problem: High default rates on personal loans.
• Research Questions:
  • "What is your primary reason for taking a loan?" (Education/House/Business).
  • "How confident are you in repaying?" (1-5 Likert scale).
• Finding: Lack of financial literacy was a key factor → led to mandatory repayment counseling.

Case Study: Chaudhary Group’s Market Research

Problem: Declining sales in retail stores despite high footfall. Solution: Designed a mixed-method questionnaire for customers.

Questionnaire Structure

Key Findings & Actions

• 60% of customers wanted more organic products → Led to expanded organic section.
• 30% complained about long queues → Introduced self-checkout kiosks.
• Low satisfaction with online orders → Improved e-commerce logistics.

Exam Tip: How to Score Full Marks

1. Understand the Marking Scheme

• Definition (5 marks): Clearly explain what a questionnaire is and its purpose.
• Types (10 marks): Compare structured vs. unstructured with real examples (e.g., eSewa vs. Daraz feedback).
• Question Formats (15 marks): Know when to use open-ended vs. closed-ended with examples.
• Design Principles (10 marks): List and explain clarity, relevance, no bias, logical flow.
• Case Study (10 marks): Apply concepts to a real Nepali business (e.g., NTC, Nabil Bank).

2. Common Exam Questions & How to Answer

3. Model Answer Structure (10 Marks)

Question: "Design a questionnaire to assess customer satisfaction with Pathao’s ride-hailing service."

Answer:

1. Introduction (1 mark)

   • "A well-structured questionnaire ensures accurate data collection for improving Pathao’s services."

2. Research Objectives (2 marks)

   • "Assess satisfaction with driver behavior, app usability, and pricing."

3. Questionnaire Design (5 marks)

   • Section 1: Demographics (Age, location, frequency of use).
   • Section 2: Driver Experience (Likert scale: "How satisfied are you with driver punctuality?").
   • Section 3: App Usability (Ranking: "Which feature do you use most?").
   • Section 4: Pricing (Open-ended: "Do you find fares reasonable? Why/why not?").
   • Section 5: Feedback (Open-ended: "What improvements would you suggest?").

4. Pilot Testing (1 mark)

   • "Test with 10 regular Pathao users to check clarity and response time."

5. Conclusion (1 mark)

   • "This questionnaire will help Pathao identify key pain points and enhance customer experience."

Final Checklist Before Submission

• All question types (open/closed, Likert, ranking) are justified.
• Real-world examples (eSewa, Daraz, Ncell) are integrated.
• Avoids bias, ambiguity, and leading questions.
• Follows logical flow (easy → difficult, general → specific).
• Includes a pilot testing plan.

How to design a 5-point Likert scale for satisfaction (Image: Nicholas Smith vectorization: Own work, CC BY-SA 3.0, via Wikimedia Commons)