MaxDiff

One of the most common tasks in market research is measuring the preference/importance of multiple attributes.  Commonly, measuring attributes is done by having the respondent rate attributes on a scale (For example, a 1 to 10 scale where 1 would be a negative impression and 10 would be a positive impression).  More often than not, when a respondent uses a scale, they will tend to group their ratings together in the same area on the scale.  So, if price, color, and style are all rated a 7 on a scale of 1 to 10, which is the most important attribute?

MaxDiff removes the guess work.  This approach presents the respondent with a series competitive sets of attributes wherein they select their favorite and least favorite attribute over a series of questions.

For instance, let's say you were the maker of a yogurt and you were interested in seeing if price, flavor, or texture was the most important to your customers.  If you were to ask a customer how important price is on a 1 to 10 scale you would find that price is always important.  You would also find that flavor is always important as well (After all, who would want to eat a bland, expensive yogurt.)  However, if you were to use a MaxDiff exercise you may find that when faced with a choice between price and taste, taste comes out many factors of magnitude more important.

The MaxDiff technique accurately measures attribute importance by mimicking shopping behavior.  This methodology offers increased accuracy over traditional ratings scales.  Here are some additional materials about MaxDiff exercises:

• MaxDiff Sample Survey- What it looks like to the respondent
• MaxDiff - Example outputs
• MaxDiff - The technical paper

Contact one of Chamberlain's research experts for more information on what MaxDiff can do for your research!