How to Avoid Bias in Card Sorting

When conducting a concert, minor changes to the cards you include can sometimes have unexpected effects on your results. Card sorting is a user research technique in which you find out how users expect your content to be presented by giving them cards that represent your content and asking them to group those into meaningful categories. Oftentimes, we can’t include all of the content in a card sort. After all, research participants are only human and they can’t be expected to sort through hundreds or thousands of items at a time. Instead, it’s common to choose a sample of your content to represent what the offerings are. And have users sought just that subset? But the sample you select can have a big influence on the groupings that your research participants create.

For example, if you’re content with types of food and you selected for your card stored items apples, bread, broccoli and muffins, you’d probably end up with people creating a category for produce and a category for bakery. But just a slight change in the cards you include could lead to a very different result. For example, if you included bananas and carrots, then many people would group your fruit into one category and vegetables into a separate category, leaving you with three.

Overall, the more items you include from a specific area, the more likely people are to make that a standalone category when they’re creating groups. So to avoid biasing the results of your card sort, make sure to include cards that proportionally represent your content. Don’t overdo it and include so many cards from one area that people make it a special category when in the grand scheme of things, it wouldn’t really make sense. It’s also a good idea to pilot test your card sort with just a few participants so that you can find out early on if they’re making very narrow categories. And if so, you might want to go back and revisit what cards you’ve selected finally. Keep in mind that even choosing cards that are proportionally representative is based on your preconceived notion of what types of content you have. That’s why it’s always a good idea to follow up a card sort with a tree test where you ask users to actually use the categories that you’ve selected to find the content on your site.

Leave a comment

Your email address will not be published. Required fields are marked *