Introduction
Format of questions
Open and closed questions
Analysing the result
A quick guide
Format of a Questionnaire
You need to decide how to introduce your questionnaire and support it with some introductory text because most people will want to know something about the survey before they agree to answer your questions. Your introduction should be polite and should clearly explain the purpose of your research. You should then take care to reassure people about the confidentiality of their answers. Finally, make an effort to thank people for their time and effort in helping you.
Instructions
You need to give clear and simple instructions so that the questionnaire can be filled in properly. If your questionnaire is being used as the basis of an interview (i.e. Someone is reading the questions to the respondents and recording their answers) then the person who is asking the questions may also need to be given instructions.
Opening Questions
Opening or warm-up questions should be simple, closed questions that are easy to answer as they will help the respondent to relax.
Main Questions
A good questionnaire has a coherent structure where questions are grouped together under clear subject headings and developed in a logical order. You should avoid leaving important questions to the end of questionnaire and try not to ask questions that require the respondent to turn the page before answering. The main questions of your research may be more difficult but they should still be written simply, avoiding long words and technical terms. Bear in mind the target clients' abilities and make it as easy for the respondent as you can by asking for as little time and effort as possible.
Remember K.I.S.S. - Keep It Short and Simple.
Length of Questionnaire
The appropriate length of your questionnaire depends upon the purpose of your research, the type of respondents targeted and the type of questions. Generally speaking keep your questionnaire as short as possible; remember that people get bored and may not answer questions properly or may not complete all of the questionnaire. The form should not take more than 10 minutes to complete (many textbooks say you should limit it to between 10 and 20 questions). Finally, the questionnaire should be well presented and look attractive.
What Happens Next?
Having decided upon questions, length, layout and so on produce bulk copies of your questionnaire and hand them out to your sample of respondents. You cannot possibly hand out questionnaires to the whole of the population that you are interested in (all car drivers, all fans of a particular pop group, all users of a particular child minders etc.) so you choose a representative sample. The sample should reflect the population as a whole, in other words it should be a cross-section.
Choosing a cross-section allows you to say that results obtained from your sample can be applied to the population as a whole; but getting a representative cross-section can be tricky. So how do you choose?
Choosing a Sample
Random Sample
This is the least scientific method being exactly what it says it is. You just ask whoever comes along next with no prior thought or planning. Whilst this is extremely easy it also means that the results are totally random and therefore almost meaningless.
Quota Sample
This is the most widely used method of sampling and involves you observing and recording the details of your target population throughout a set time span. Imagine that you were going to sample the customers of a particular shop; observe the customers at a number of different times of day and record basic information about them such as approximate age, gender etc. If 54% of the customers were female then 54% of your sample should be female; if 43% were children then 43% of your sample should be children and so on.
Stratified Random Sample
If you can obtain access to a list of customers, clients, members or whatever you use the list to select equal numbers of people from each sex or from specific age groups and you then take a random sample from these selections.
Systematic Sampling
Another simple method in which you choose every third, fourth, fifth person or whatever and ask them to complete the questionnaire. Again the simplicity of this method leaves the results lacking in conviction.
Whatever method you opt for (remember that tutors may require you to use a specific method) print out the questionnaires and have them completed by your respondents. Gather in the completed papers and you can then begin analysing your raw data and turning it into useful statistics.