Questionnaire design
- Concept
How to write, order and test questions so that everyone understands them the same way and the answers can be analysed.
What it is
A questionnaire is a measuring instrument: each question should measure one thing and be understood the same way by everyone. It can be self-completed (on paper or digitally) or given by an interviewer (in person or by telephone), with closed or open questions or scales such as the Likert scale. Wording, order and layout change the answers, so a validated instrument is preferred and, if a new one is written, it is piloted before the study.
Key facts
- Recoding a reverse item on a 1-to-5 scale: new value = 6 − original value.
- Rescaling a score to 0–100: (score − minimum possible) / (maximum − minimum) × 100.
Worked example
- Made-up 5-item adherence scale, from 1 (never) to 5 (always). Item 3 is reverse-worded.
- One person's answers: 4, 5, 2, 4, 3. Item 3 is reversed: 6 − 2 = 4.
- Score: 4 + 5 + 4 + 4 + 3 = 20, on a possible range of 5 to 25.
- On a 0–100 scale: (20 − 5) / (25 − 5) × 100 = 75 points.
- Pilot with 15 people: 4 leave item 3 blank (4 / 15 ≈ 27 %) and 6 say ‘sometimes’ and ‘almost always’ sound the same.
- Item 3 is rewritten without the double negative and the options are clarified; in a second pilot nobody leaves it blank.
- Reading: the score makes sense only if reverse items are recoded, and the pilot found the problems before the questionnaire reached everyone.
Types of question
| Type | Advantage | Disadvantage |
|---|---|---|
| Closed (yes/no, multiple choice) | Quick to answer and analyse | May not include the true answer |
| Open | Captures nuance and unexpected answers | Hard to code and compare |
| Likert scale | Measures attitudes and frequencies in degrees | Tendency to answer the same way or in the socially acceptable way |
| Visual analogue scale | Continuous measure (pain 0 to 100) | Needs clear instructions; hard by telephone |
Markers
- Likert
- VAS
- Pilot
In practice
- An interviewer-administered questionnaire gets more complete answers; a self-completed one gets more honesty on sensitive topics.
- Before writing a new questionnaire, look for one already validated in Spanish and in a similar population.
Common pitfalls
- Double-barrelled questions (‘do you smoke and drink alcohol?’) that one option cannot answer.
- Summing a scale without recoding the reverse items.
- Skipping the pilot to save time.
Test yourself
Think of the answer, then tap the question to see it.
What is wrong with the item ‘You do not fail to take your medicines’ as a question?
It has a double negative and leads the answer; it is rewritten in a direct, neutral way.
On a 1-to-5 scale, how is a 1 on a reverse item recoded?
6 − 1 = 5.
A 10-item scale scored 1 to 4 gives 25 points. What is that on a 0–100 scale?
Range 10 to 40: (25 − 10) / 30 × 100 = 50 points.
What is a pilot test for?
To find confusing questions, missing options, broken skips and the real completion time before the study.
Connections
Educational content on epidemiological and statistical methods; the examples use made-up numbers to teach the calculation. Not a clinical or legal guide: current Colombian rules are in the official sources.
Sources: Collecting Data (Centers for Disease Control and Prevention (The CDC Field Epidemiology Manual), 2024); Recall bias | Catalog of Bias (Catalogue of Bias Collaboration, 2017).
Spotted a mistake?We are always improving for the community. Tell us what is wrong and our team will review it.Report a mistake
Representative model, for educational purposes only.