Because the purpose of a usability report is to bring to light the issues that came out during the test, the things you report may seem more important than they actually are. By simply including an observation in the report, you are already inflating its severity. Someone who reads your report can
easily seize upon an observation that "sounds right" to them - or worse, confirms a bad design idea they thought of themselves - and blow it out of proportion.
Assigning a severity rating to observations can keep this inflation in check. The more objective you can make the ratings, the better. Of course, the severity of a usability problem is a totally subjective metric, buy you can set some criteria for what counts as a major problem. the most typical definitions are followed:
High severity
- Prevents the user from completing the task.
- For example, User could not find the submit button and believed the task was complete even
though it was not.
Medium severity
- Does not prevent the user from completing the task, but caused a significant distraction that the
user noticed.
- For example, The system threw an error when the user clicked submit. Upon clicking back to the
form, the user realized that he missed a checkbox that was required before moving forward.
Low severity
- Does not prevent the user from completing the tasks, but caused a delay or an error that the user
did not recognize.
- For example, The user knew he needed to click a submit button, though he could not immediately
find it.
