In an experiment, same group of participants are required to test (low and high) visual quality of a user interface. The goal is to find whether perception of usability is affected by any of the low and high visual aesthetic conditions.

I am using a questionnaire to measure "perception of usability" (dependent var).

(i) Do I need to measure "perception of aesthetics" of the participants as well to find out if aesthetics has an impact on usability?

2 Answers 2


Is it important to you to find out whether perceived aesthetics affects perceived usability? If yes, then ask the question and do a mediation analysis. If no, then you don't need to ask about aesthetics -- you're already measuring the thing you care about.

Also note that asking users to test both versions of the interface is not a common setup for A/B testing. They're likely to rate the two relative to each other -- e.g., users might think the low-quality version is okay if they see it on its own, but if they just saw the high-quality version then it might seem ugly. There's no way to stop people from doing comparisons, except to randomly assign each person to see only one of the interfaces. If you have few enough people that you absolutely have to double up, remember to randomize the order they see them in.


The answer is Yes. It is important to find out whether perceived aesthetics affects perceived usability. Thanks for direction of pursuing a mediation analysis. Can I say in this case that there is one independent var (low and high aesthetics) and two dependent var (perceived aesthetics and usabili

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.