Market Research Survey

I was getting caught up on some reading over the weekend and came across a poorly designed matrix question posted on the Bad Research; No Biscuit. blog. The blogger explains during the market research survey, he was asked about different brands of cat food (14 brands of cat food). According to the post, he said he'd never heard of 5 of them, but then saw the survey question asking how he'd heard about all 14 brands.

Here's a thumbnail of the never-ending matrix, you can see the full matrix in his post, but I think this gives you the gist. Keep in mind it goes down for another 15 categories:


For the moment, I'm going to ignore how the survey writer is not following best practices for designing matrix survey questions and an appropriate length and focus on the fact that all the brands were shown after the survey respondent said they hadn't heard of all of them. This is exactly why we have online survey technology for complex survey questionnaires hosted online. If you're making it easy on the survey respondent to complete your survey - and the respondent experience should be a critical focus of your survey design - using survey logic is a no brainer.

If you're a Cvent client, you have the ability to prevent this survey design faux pas from happening in the first place. You can apply what we call Link Logic. Link logic allows the survey designer to customize the answer choices respondents see based on answers to previous survey questions. In this market research survey example, the survey writer would only have the 9 brands the survey respondent had heard of show up.