Crafting Effective Survey Questions: Tips for Accurate and Unbiased Data (Full Transcript)

Speaker 1: Good survey questions. Hi, I'm Dr Nick and I'm going to talk about the questions you ask in a questionnaire or survey. This video follows on from another video, Designing a Questionnaire. These ideas are also useful for critiquing a questionnaire or survey. It is very important that the questions in a questionnaire are well written to find out what you want to know and to avoid bias or non-sampling error. There are different types of questions. Open questions let the respondent say what they like, while closed questions give respondents a choice of response. Open questions are useful when you do not know how people are likely to respond or when you want to avoid influencing their response. But open questions can be very difficult to analyse as there are so many different ways that they can be answered. You might think that it is easy to ask good questions in a questionnaire. But it isn't. You want to make it as easy as possible for the respondent to give you the information you want. To start with, the questions should be as short as possible. Questions need to have correct spelling and grammar. Questions should use language appropriate to the subject and the people you are asking. If you are asking experts you can use technical language, but generally you need to keep the language simple. It is usually better to ask what a person did in a certain instance or over a certain period of time than for them to recall in general what happens. But you need to be realistic about what you can expect people to remember. And you need to define your terms. It is important to have just one idea in a question. If there are two ideas, use two questions. It is better to ask questions in positive terms if possible. Avoid a double negative as that can be very confusing. In closed questions you need to be careful about the alternatives you provide. You need to make it clear whether the person is supposed to choose one alternative or as many as they wish. And you need to make sure that the alternative choices do not overlap and that they cover the whole range of alternatives. Often a category, other, can be used for alternatives you might have missed. You want to avoid influencing people's response which can lead to non-sampling error. So you need to word the question as neutrally as possible. You need to be careful if you are asking a question that may be personal or upsetting. As we described in our previous video, it is very important to check your questions with a pilot survey as this will identify potential problems. The more effort you put into devising the questions, the better quality data you will get. This video was brought to you by Statistics Learning Centre. Visit our website for more resources to help you learn.