Usability Tests vs. Validation
I recently asked Jonathan Avedikian, Certified Usability Analyst, Senior Software Analyst, Implantable Electronic Systems Division at St. Jude Medical in Plano, Texas, the following question:
What's the difference between usability tests and validation? Is validation just a final usability test at the end of the product development process?
Validation confirms by objective evidence that device specifications conform to the user needs and intended use(s). It demonstrates that users have the ability to safely operate and maintain the device, as well as, understand the accompanying labeling.
Validation activities must involve actual intended users. There are many user-centered design activities that happen throughout the design cycle. Usability testing is just one of them. Usability tests are conducted during the validation phase to prove the usability objectives have been met. Usability testing is a method of validation.
I validate the product by usability testing with users per the 62366:2007 standard. My software user interface design uses best practices but that does not guarantee it meets the user's needs, expectations, matches their mental model, or that the user understands the labeling.
FDA and TÜV requires us to prove we did that. So, I set usability goals and then usability test with users to make sure the device is safe, efficient, learnable and satisfactory to use. Meeting the user's expectations helps ensure positive user experience. I discover if I have done this during validation through usability testing.
The usability tests should NOT be conducted just at the end of the project. Tests should be run throughout the design cycle to ensure design goals can be met and uncover any safety issues. This will also allow changes to be made early while it is easy and less expensive to do so. These early validation activities are called formative testing. They are quick and dirty, informal, and use fewer users. Summative validation testing is performed at the end of the design cycle following verification. This testing is formal, has pass/fail criteria, at least 15 users per distinct group, protocols and reports.
Stephanie Wiseman, Community Editor, UBM