Skip to main content

Frequently Asked Questions

Should I ask for the respondent's name or an ID?

There are several reasons for tracking the identity of your respondents. First, you may wish to track who responded so that you can send reminders and follow-ups only to those who have not yet responded. This is an admirable goal, but you should weigh the burden of an unnecessary contact with respondents against the discomfort they may feel at your having an identifier.

You might track respondents if you have additional individual information about them that you'd like to be able to link with their survey responses. For example, College wide surveys sometimes track respondents so that we can learn how opinions may relate to institutional data such as respondents' participation in majors or minors.

Another reason to track respondents is if you have offered an incentive for responding that you will need to award. In this case, however, you may be able to find a way to have their response status tracked without actually linking their identity to their survey.

In any of these situations you should be clear in the survey invitation whether and how tracking will be used.

Should I have a neutral point on a scale that I am using for a question's response options?

You can find experts arguing on both sides of this question. The neutral point allows the respondent to... well, to be neutral! It is the 2 in a 3-point scale, or the 3 in a 5-point scale. It is the choice to neither agree nor disagree, or to be neither satisfied nor dissatisfied. A scale that does not provide a neutral point essentially forces the respondent to give an opinion. Some audiences don't respond well to being forced (Swatties, perhaps?), and some issues may be complex enough that a neutral option is warranted. Again you should balance the value of having a clear opinion against the accuracy of your data and comfort of your respondents.

When should I offer a response option such as "Not Applicable" for a question?

When it is possible for a question to not be applicable to everyone. It is useful to know whether people didn't respond to a question because they skipped it (who knows why) or because it didn't apply to them. You're not learning an opinion, but you are learning some important information. Just be sure to handle those responses properly in the analysis stage.

Should my scales go from negative to positive or positive to negative?

You should consider the layout of your questions in deciding the direction. If the response options are laid out left to right, it may make intuitive sense to put the positive options toward the right, like this:

Strongly Disagree Disagree Agree Strongly Agree

But if the response option go from top to bottom, you may wish for the respondents to see the positive options first, like this:

Very Satisfied
Somewhat Satisfied
Somewhat Dissatisfied
Very Dissatisfied

What is really more important is that whichever way your scales go, they all go the same way, so that your respondents don't miss a change of direction, or get confused.