Your test should explain what it's testing. Via it's name or some other mechanism. Otherwise how do you know what's gone wrong when it fails?