Iterate on Evidence
Rank what you found by severity, fix the worst problems first, and test again.
What you will learn in this lesson
You will turn test notes into a ranked fix list. Then you will improve your prototype and test again.
List every problem
Go through your notes from all tests. Write each problem as one line. Note how many testers hit it.
- "4 of 5 did not see the 'What to bring' button."
- "2 of 5 thought 'Proof of residency' meant a passport."
- "1 of 5 wanted the hours on the first screen."
Rank by severity
Severity means how bad a problem is. Use three levels:
| Level | Meaning | Example |
|---|---|---|
| High | People can't finish the task, or they could be harmed | Can't find what papers to bring. Data shared without consent. |
| Medium | People finish, but with trouble or errors | Confused by a label, but figured it out |
| Low | Small annoyance | Wanted the hours higher up the page |
Fix every high item. Fix medium items if you can. Note low items for later.
Accessibility problems that block people are always high. A form a screen-reader user can't finish is a task failure, not a style note.
Change, then test again
Make your fixes. Then test with two or three new people. Did the high problems go away? Did the fixes create new problems? This loop of test, fix, and test again is called iteration.
Judgment check: what you won't fix
You can't fix everything. Say what you left out and why. That honesty is a strength in a case study. It shows you made choices on purpose.