There is a fine line between requiring respondents to answer a question thus avoiding missing data, and pushing them to a place where they decide to exit the survey before completion. As the survey author you have the ability to specify whether or not a question is ‘required’ or alternatively a respondent can skip the question and continue with the survey.
When a respondent skips a question they are qualified to answer this creates missing data. This causes further problems when you enter the survey data analysis phase, especially if you are using multivariate techniques such as regression and discriminant analysis. Missing data can be imputed using other variables and procedures such as mean value substitution. However the depth of this topic requires its own post.
Requiring a respondent to answer a question minimizes the risk of missing data. However, it can set up a dangerous feeling within the respondent that ultimately drives them to terminate the survey before completion, if they believe it is taking too long to complete the task. Cvent, and most online survey platforms, offer reporting tools which highlight questions that lead to a disproportionate increase in abandonment.
Should you decide to set a question to be ‘required’ it can be done during the question creation phase and can be easily edited post-facto. The graphic below highlights this process.
In order to reduce the desire to exit the survey, it is best to focus your efforts on improving survey design, with the goal of making the survey engaging, yet respectful of the respondent’s time. This process includes ensuring proper screening, writing clear and concise questions, including ‘none’ or ‘N/A’ options, and keeping the survey on task (avoiding question creep that expands the size of the survey).
The fine line is between requiring a question to be answered and allowing a respondent to skip a question thus creating missing data. This issue can be resolved through proper questionnaire design.