When to Use MaxDiff Analysis
MaxDiff is a survey research technique for working out the relative preferences of respondents. It can be used to rank brands, products, attributes, and even people. Due to its versatility, MaxDiff is a popular tool among market researchers, political pollsters, and anyone looking to understand the tastes and preferences of a particular group of people.
Example of MaxDiff
A MaxDiff study presents a series of questions in which respondents are asked to pick a “best” and “worst” from a set of options. For example, consider the following question posed by a political pollster:
The question requires respondents to pick a “most appealing” and “least appealing” characteristic for the President of the United States and provides five options to choose from. Respondents may find all of these traits appealing (or unappealing), but they are forced to make a decision.
Advantages of MaxDiff
To understand when you should use MaxDiff, let’s consider some alternative surveying methods. The most obvious alternative is to ask respondents to rate each characteristic individually. Using the above example, that would involve taking each characteristic — “entertaining,” “concerned about poverty,” “good in a crisis,” etc. — and asking respondents to rate how appealing they are. This technique is only useful if the ratings for each category are significantly different. Respondents may rate almost every characteristic as “very appealing,” which would make it impossible to decipher their preferences.
Another method would be to list the characteristics and ask respondents to rank them from highest to lowest. This method would give you the complete preferences of your respondents, but the process can be painful and time-consuming, causing respondents to ditch the survey or answer randomly.
MaxDiff directly accounts for two significant shortcomings of these alternative techniques:
- Poor discrimination between alternatives: When asked to rate a list of exclusively positive or negative attributes, the majority of responses will be clustered together. This makes it difficult to rank the attributes. MaxDiff gets around this problem by explicitly asking respondents for their preferences.
- Yeah-saying bias: Many survey respondents are inclined to agree with the question or to indicate a positive connotation. Once again, MaxDiff avoids this problem by forcing respondents to choose a “best” and “worst” option.
Applications of MaxDiff
The primary use of MaxDiff is to determine the relative preferences from a set of options. In the example in this blog post, it would be voter preferences of presidential characteristics. A MaxDiff study can also be used to understand differences between people based on their preferences. Respondents can be divided into segments which may help you identify different voter groups.
Check out the Q Research Software eBook on MaxDiff
You’ve just uncovered the tip of the iceberg. Want to learn more about MaxDiff analysis?
Download the Q Research Software eBook on DIY MaxDiff.