Questionnaires

Types of questions

Questions largely fall into two types and you may decide to use a mixture of closed and open questions in your questionnaire

Closed questions (or fixed response).

Fixed responses may be in the form of

• 'yes', 'no' or 'don't know',
• marks against items in a list, for example, 'tick all of the books you used at least once during this module';
• a rating scale where respondents have to say how far they agree or disagree with a statement, for example:

How likely are you to recommend a friend to take this course?

Highly likely
Possibly
Not sure
Not likely

(Note, it is usually wise to give an even number of choices in this type of question to prevent people opting for the middle of three or five)

Analysis of fixed responses

Fixed responses can be counted, so for example, you can say that X out of Y respondents were 'highly likely' to recommend the course, but W out of Y were 'not likely' to and Z were 'not sure' if they would.

In the case of ticks against a list of items you would be able to say, for example, that X out of Y respondents used a particular book at least once.

Questions and responses of this type could be useful in doing market research or surveys. What we don't know however is whether a respondent 'hovered' between two responses and finally, for speed, ticked a particular response. There is no opportunity for respondents to say 'it depends' or to qualify their responses in any way.

Open questions

Open questions allow respondents to express individual differences and their particular thoughts on a question. So, for example, after asking people to say how likely they were to recommend the course to a friend, they could be additionally asked 'Please give reasons for your choice'. Or they could be asked, 'What is/are your main reason(s) for wanting to take this module?'

Analysis of open responses

Open-ended items impose fewer constraints on the responses that can be given but they can be challenging to analyse. How do you differentiate between a response written like an essay and a brief one-word response? Do you give them equal weight? There are many ways to analyse such qualitative data, for example,

• you may count the number of times a particular word or idea occurs and provide numerical data in your 'results',
• you may look for themes which you decided on before the questionnaires were distributed,
• you may wait until they are completed and see whether themes emerge.