Designing and Using a Questionnaire

1. What makes a good questionnaire?

A questionnaire is a list of questions that we give to a group of people, called respondents, so that we can collect data in the same way from each of them. Because everyone answers the same questions, we can compare and count the answers easily.

A good questionnaire has these features:

• A title and a short introduction saying who you are and why you are asking.
• Short and relevant: ask only what you need to answer your main question. Eight to ten questions is enough for a class survey.
• Clear and simple language: one idea per question, with no hard words.
• Fair and neutral: the question must not push the respondent towards an answer.
• Easy to answer and to count: give answer boxes or blanks with units, such as GH₵ ______.
• Polite and private: thank the respondent and do not force anyone to give their name.

▶ Follow the five steps from the aim to the recorded results. Notice that the pilot test comes before the real survey.

2. Writing good questions

There are two main kinds of question. A closed question gives a fixed list of answers to tick, so it is quick to count. An open question lets the respondent answer in their own words or with a number, which gives more detail but takes longer to analyse.

Compare these pairs of questions. The first in each pair is poor; the second is better.

3. Administering the questionnaire

To administer a questionnaire means to give it out, explain it and collect the answers. How you do this is as important as how you wrote the questions.

1. 1Pilot test. Give the form to three or four learners. Ask them which questions were confusing. Fix them before you begin.
2. 2Choose a fair sample. For a class of 40 you could ask every learner. For a bigger group pick names by a rule, such as every 5th name on the register.
3. 3Explain and ask permission. Tell respondents the purpose and that answers are private.
4. 4Give the same instructions to everyone and allow enough time. Do not influence answers by smiling or frowning.
5. 5Collect all forms and check that each is complete. Number each form so that you do not lose or count one twice.

If you cannot reach every person, record how many forms you gave out and how many came back. If only a few come back, your results may not represent the whole group, so be careful with your conclusions.

4. Recording the results

After collecting the forms, copy every answer into one data table. Each row is one respondent and each column is one question. This is called a raw data table.

Next, for each question count how many respondents gave each answer. We use a tally: draw a stroke for every answer and cross the fourth with a fifth stroke to make a bundle of five, like ||||. Then write the total in a frequency column.

▶ The bar for Mathematics is the tallest, so it is the most popular subject in this survey.

Worked examples

Try each one yourself first, then open the solution and compare every step.