Surveys and questionnaires have become part of our daily life. We are constantly asked to take a brief survey about a product we purchase or engage with customer service. But how often do you think about whether or not the questions you are asked make sense? Will they produce useful data? Can the questions actually be answered?
Here is an example of a question that I was asked several years ago about my son’s school: “Do you think your child has received a rigorous education at our school?” On the surface, this seems easy to answer. It’s a simple yes or no. But if you think about it, the question is really quite difficult. In some of his classes, my son did receive a rigorous education, but in other classes he didn’t. So how do I answer? It was presented as a yes/no question, but the answer really wasn’t ye or no.
One solution to this problem is to offer a Likert scale, which is a psychometric rating scale that we often see in questionnaires. When you have a statement–such as “My son received a rigorous education at our school”–that requires you to rate it with something like strongly agree, agree, no opinion, disagree, strongly disagree, you are using a Likert scale.
This example is that it points out how important it is to think deeply about the construction of questions for surveys/questionnaires and also to consider how the way you phrase questions will influence the analysis of the data you collect.
Below are links to pages that may help you in thinking through some of the issues that arise in writing questionnaires. This part of the site (like others) is undergoing regular revision and expansion, we will add more information in the future.