A Blueprint for How Not to Measure Confusion: Coffee Case Study