Image Image  member login  |  privacy policy  |  contact us
    Home  |  FAQ  |  Awards  |  Reports  |  Forum  |  Jobs  |  Chapters  |  Join  |  Order
  Forum Questions
Have a question?
Send it to
          Jane Farber.

    ASPonline.com  >  ASP Forum

ASP Forum


Do survey scales affect the accuracy of customer satisfaction scores?

"Our current customer satisfaction questionnaire asks people to rate us on a five-point scale--5 is the best, and 1 is the worst. A friend tells me we'd get more accurate results if we used either a ten-point scale or school-style grading--A, B, C, etc. Would this really make a difference?"

—Mavis from Mattapan                           



Dear Mavis:

We asked Fred Van Bennekom, a survey design expert and author of "Customer Surveying: A Guidebook for Service Managers", for his advice:

"The traditional argument for using 10-point scales is that you get finer measurement. However, I'm skeptical about this theory. My experience is that few respondents ever select 6, 7, or 8 as answers, because they get hung up trying to pick the answer that accurately describes their feelings. Finally, they just gravitate toward the end points, usually 9 or 10. In other words, more scale points can lead to worse measurement! On a paper or Web form survey, moreover, the 10 points take up a lot of territory, leading to more line wraps for the questions and the appearance of a longer survey. (The alternative is to have respondents write in the answer on a paper survey or use a pull-down box in a Web survey, both of which I dislike.)

"The 'School Grade Scale' is an interesting one. I've tested various scale types in a telephone survey, and I've found that this scale is easy to use and promotes a good dispersion of responses, unlike the 1-to-10 scale. Respondents have a lot of experience with how this scale measures performance, moreover, so they seem to have an easier time translating their feelings into points on the survey scale.

"Here's another option: a 6-point scale. This forces respondents to commit to one side or the other, no fence-straddling allowed. One of the hardest problems in survey design is figuring out how to describe a completely neutral midpoint anchor. But with a six-point scale, you don't have to agonize over this question. That's one reason I often lean toward this scale type."

—Fred Van Bennekom  fred@greatbrook.com
    Great Brook Consulting  www.greatbrook.com
    978/779-6312




[Other comments and suggestions about this topic? Send me an email and we'll post your feedback.]