Portland Marketing Analytics (PortMA) | Portland, Maine

Treating Likert Scales in Event Marketing Research

Treating Likert Scales in Event Marketing ResearchLikert scales are a great tool for event marketing research. Giving a range of response options lets you probe into exactly how people feel about your product. That being said, Likert scales do have their troubles and drawbacks to take into account. So long as you remain aware of these and set them up properly, you will find them a valuable analysis tool.

Likert scale basics

The typical Likert scale takes on a form like so:

  • Strongly agree
  • Somewhat agree
  • Neither agree nor disagree
  • Somewhat disagree
  • Strongly Disagree

The most important aspect of this is the symmetry, with the number and phrasing of the negative responses mirroring that of the positive responses. The neutral middle is technically optional, but I recommend it in most event marketing research, as it is easy for consumers to be ambivalent to a product. On more politicized questions, it also gives a safe response option to people who have no opinion or are unaware of the topic at hand.

Debates over analysis and bias

When dealing with Likert scales, always be aware of the debate over whether a Likert scale is interval based, or simply ordered. The central point of this argument is whether you can consider the difference between answers on a Likert scale the same, regardless of which consecutive answers you are comparing. For example, is the difference between “strongly disagree” and “somewhat disagree” the same as the difference between “neither agree nor disagree” and “somewhat agree”? The most important thing is simply to be aware that there is a controversy around this, and so while some people may be fine with you considering a Likert scale to be interval based, others may disagree, and you should be prepared to speak to why you made the decision you did.

While there are multiple causes that can distort responses to a Likert scale, the one you should most take into account is central tendency bias (many others have balancing factors that limit their impact, while central tendency bias does not). Some survey respondents will be reluctant to give the most extreme response option to avoid being seen as having extremist views (as well as a few other reasons). This can be true in event marketing surveys, with consumers not wanting to commit to saying they are extremely likely to buy something, and instead hedge by simply saying they are somewhat likely.

There are other aspects of Likert scales you should consider, so please look for my follow-up post on Likert scales, including a few additional potential biases to be aware of in responses as well as handling multiple Likert scales in analysis.

Share Your Thoughts...
Tell us what you're thinking...
and oh, if you want a pic to show with your comment, go get a gravatar for free here!
Comments are moderated and may not appear immediately...