Carrying out interview surveys
To collect primary data or to investigate people's perceptions about an issue, project or development.
Interview surveys can be more detailed and more flexible than questionnaires, with open ended (rather than closed) questions and the opportunity for respondents to give their opinions without being pigeon-holed by option boxes. Interviews can also be less structured than questionnaires, so particularly interesting points can be pursued and you can adapt to follow the flow of the conversation.
- Research the topic thoroughly
- Decide who you need to interview
- Try to find out the names of specific people (use personal contacts if possible) and write a formal letter to them - in plenty of time and explaining who you are, what you want and why you want it, and asking for them to contact you to arrange a mutually convenient time. They may want to know the type of questions you'll be asking
- Decide on exactly what it is you want to find out from each person
- Decide on the sampling strategy if you are planning to interview large numbers of people, for example if interviewing visitors to a country park
Structuring your interview
- Start with a few factual, easy and short-response questions to put the person at ease
- Think about the sequence of your questions. They should follow a logical sequence and the deeper more searching questions requiring a more thoughtful response should come later in the interview
- Consider the length of time it might take - aim for no more than 30 minutes
- Ensure that your questions are pertinent and show your understanding of the topic; think about each question beforehand and consider how you would explain your meaning if the respondent is not fully clear what you're getting at
- Ensure the questions encourage more than just a yes or a no answer. Be prepared to prompt for more detail
- Think about the wording of questions. Your questions should be clear and unambiguous, without any ‘conditioning' in the way that the question is being asked (for example, questions starting with ‘do you agree...' encourage a ‘yes' answer). Think also about the tone of the question as you ask it
- Be aware of potentially sensitive or provocative questions
- Conduct a pilot test and refine your questions accordingly
Conducting the interview
- Dress smartly
- Be punctual and polite - introduce yourself confidently
- Use a digital voice recorder to record the interview if possible, but make sure that you ask for permission first
- Explain the purpose and the basic content of the interview. Try to make some ‘small talk' to put them at ease
- Talk clearly and be prepared to clarify or explain your questions
- Adapt to the person and try to build a rapport with them - you don't have to stick rigidly to the questions if they make an interesting point you'd like to follow up
- Use a note-book to jot down notes for each question, and any other information they give you which is not directly related to any of your questions - it can be tidied up later
- Obtain permission to use quotes from them if possible
- End as you began: politely thanking them for their time
Using the data within an investigation
- Be selective with what you use: it's impractical to try to include everything in your report
- Think about the best ways to represent your main points, for example a table summarising positive and negative comments or a comparison of the opinions of different respondents, for example a developer with a local farmer
- Include other primary or secondary data where relevant to reinforce and back-up your information
- Visual representations of the key points can be effective, for example speech bubbles around pictures or a central map
Considerations and possible limitations
- It may be difficult to arrange an interview - you want to aim to interview the ‘top' people, for example the developer or site manager and these are busy people! You may not even get a response and have to re-think your tactics
- Take care if the issue you are investigating is sensitive or controversial. You must strike a balance between getting someone's opinions and going too far! You do not want to cause offence
- Be prepared to have certain questions or requests refused
- Interviews are time consuming; both in terms of preparation, which needs to be carefully done, and the interviews themselves
- Quantitative results which can be statistically tested are hard to obtain; qualitative results need careful thought with regard to how to use constructively in a way which will add real value to your study
Interview of a developer of a new housing estate:
- Will you explain and justify the choice of location please?
- Will you please summarise the main outcomes of the EIA into this development?
- In what ways have local people been involved in the decision-making process?
- In what ways have the findings of the EIA and local people's opinions influenced the plans for this development?
- a.) What impacts do you anticipate during the construction phase of this development, both to the environment and to local people? b.) How will you minimise these impacts?
- How will this development help to meet the housing needs of the local area, for all social sectors of the community?
Use within an investigation
- Questionnaires are a very common feature of investigations where the opinions of a group of people or different groups of people are relevant
- They can be used to obtain information about the people themselves (for example a social survey to investigate the characteristics of the population being surveyed), information about patterns and processes (for example the origin of visitors to an amenity or area), or information about opinion, preferences and behaviour. Some questionnaires may involve gathering information on all of these aspects
- The data collected will not be available from any other source - it is primary data unique to your investigation
- Questionnaires can be used to assess the level of opposition or support for a development, and to compare different groups of people or different areas around a development such as a quarry
- They should be used in conjunction with other primary and secondary data rather than in isolation
There are a variety of ways in which the person can mark their choices on a questionnaire - circling, ticking, and ranking. Think about the best method to use for each question; a mixture may make the questionnaire more interesting
Think about the length of the questionnaire and the sample size and make-up...
- Think carefully about the overall aims of your investigation and how you want your questionnaires to fit in with these aims. Roughly draft some questions and play around with them until you are satisfied. The questions you design should follow some basic rules:
- Start off with a few general, easy questions to put people at ease. Ask them things like how far they have travelled, or how they have travelled
- Beware of personal or sensitive questions. You may want find out how old your respondents are as it may be interesting to analyse whether different age groups have different opinions. Age is tricky though - some people are sensitive about their age. The best thing to do is to use categories, for example under 20, 21-30, 31-40 and so on. That way, people don't have to give their exact age, and can just point to the appropriate category
- Names and addresses are usually not important, and are not relevant to your study, so leave them out
- Do not put in embarrassing or irrelevant questions
- Open questions are those which do not give a choice of answers, so people are free to say what they like in response. For example, questions which ask for an opinion
- Closed questions are those which have a set choice of answer options for people to select from. People can still be asked for their opinion but must choose from a pre-arranged set of options. This makes it a lot easier to analyse the results
- Leading questions. It is easy to phrase a question so that it leads the respondent towards a certain response. For example, Do you agree that graffiti is a problem in the park? People are led towards saying yes because of the way the question is worded and phrased. This makes the results biased, and you must therefore think carefully about how you phrase questions
- Double questions. Avoid asking two in one questions, like Do you think that there should be more litter bins and dog waste bins in the park? What about the person who does not own a dog and is not bothered about dog-waste bins, but does think there is a litter problem? Similarly, avoid double negatives in your questions
- Keep your questionnaire simple: short and clear questions without ambiguity
Length - You need to find out certain things, but a clear, well-designed questionnaire with unambiguous (mostly) closed questions should not take up too much of peoples' time. Aim to get what you need to know in no more than 10 questions and try to keep it to one side of A4 if possible. Sample size - Too few respondents and your questionnaire won't be representative, too many and it's unrealistic. Try to strike a balance - aim for a 10% sample of the population. The make-up of the sample is an important consideration too. A sampling strategy will need to be identified based on systematic, random or even stratified sampling (if the composition of the population is known and you are interested in examining the results by age or gender)
Method of delivery - Is the questionnaire going to be face to face or leave and collect
Pilot study - Once you have designed your questionnaire, carry out a pilot study. Ask respondents to tell you if any questions are unclear, leading or biased, sensitive, or if you have missed anything. Think about how easy it will be for you to analyse the results. Make any changes you need before you undertake the real investigation
Carrying out your questionnaire survey
Start each questionnaire by introducing yourself, saying where you are from and what you are doing. Ask people politely if they could spare a few moments to answer your short questionnaire.
While you are asking your questions, remember to:
- Listen to what they are saying
- Look interested
- Respond to comments and be prepared to explain questions if they do not understand
- Ask for clarification if you do not understand something they have said
Finally, do not be offended or put off if people do not want to stop and answer your questionnaire - people are busy.
Considerations: Limitations and validity
- Carrying out face-to-face questionnaires can be time consuming
- Questionnaire fatigue - people are fed up with being stopped in the street
- People may not be completely honest in their responses - they may be tempted to say what they think sounds good or what they think you want to hear
- People may not actually know the answer to some questions, for example the distance they have travelled
- The sampling strategy may be hard to decide upon and implement in practice