|
ANY professional marketing researcher will surely agree that the ordering of questions is a critical consideration in the construction of a questionnaire. While none of us would ever consider asking an aided question prior to an unaided one, the placement of what we at Q & A Research call "overalls" is still contested.
Overalls are questions that are broad in scope. For instance...
What, if anything, do you like about XYZ brand ice cream?
or
How satisfied are you overall with the service you received from XYZ?
When accompanied by attribute ratings, some researchers have debated the ideal placement for these overalls. Should they be on the front end or the back end of the attributes? At Q & A we don't know what all the fuss is about. We have a definite opinion that the overalls should always precede the attributes and here's why.
Suppose we ask a series of attributes about key product features. Then we ask what the respondent likes about the product. I shudder when the respondent's answer includes the very features we just asked about.
Take another example. Imagine that we ask how satisfied a respondent is with several key aspects of a service. Then we ask the overall satisfaction question. Our experience is that the overall tends to be nothing more than an average of all previous attribute ratings given. We have in effect told the respondent that we consider these to be the most important criteria by which to make one's overall evaluation. And if this is the case, why ask it?
When placing the overall question in front of the attributes you will most assuredly avoid any potential order bias.
Have you seen my overalls? If you have, you've undoubtedly seen them right where they belong.
|
