This week’s “Fix-It Friday” looks at survey branching. The question below was taken from a survey about events and activities at a particular high school. The purpose was to survey parents to see which activities and events they had attended and get feedback on their experience at those activities and events.
PROBLEM: For each of the five types of events and activities, the respondent has to consider EIGHT possible responses. In addition, the responses are inconsistent throughout the main categories of No, I’d like to, and Yes.
The matrix layout saves space and looks nice but in this particular example it is too confusing. Respondents will be overwhelmed by each of the 40 scenarios they will need to look at and consider. Once respondents feel overwhelmed, they tend to either not provide answers or drop out of the survey altogether.
SOLUTION: A better way to construct this survey would be to use branching. Branching narrows down the number of scenarios each respondent will need to consider in a question and it only shows questions that are relevant based on a previous answer. Another advantage to using branching techniques is the initial question is usually very simple.
For example, in this scenario, respondents would be asked if they had attended each of the activities or events separately. Respondents will only need to think about two options: Yes or No.
This question is straightforward and does not take much brain power to answer. (Note: To make the above question even easier, you could add in a time frame, e.g., “Have you attended an open house at ABC High School in the last year?”). If a respondent answered “Yes” to the above event, then he would be directed to a set of questions perhaps asking about his experience. If a respondent answered “No” to the above question, then he would be directed to a different set of questions perhaps asking the reasons behind his lack of attendance.
Let’s assume the respondent reply “No” to question one. Now he gets skipped over to question two that asks about his lack of attendance.
Once again, the respondent is answering a very simple question and not distracted by answers that are not relevant. Including an “Other” option allows you to capture any possible explanations you did not think of when creating the survey.
As the survey developer, setting up your survey in this way might take you more time as there are additional questions to program along with the branching pattern. However, your respondents will find the survey easier to understand and answer, which makes them more likely to complete the survey. The additional investment in developing and programming the survey will lead to better data collection. Better data collection leads to better insight.
Have a bad survey question that needs some improvement? Send it to FixIt@SaquiResearch.com. Your survey question could be the one we fix in our weekly Fix-It Friday series.
