Norm-referenced tests compare a test-taker's performance to others in a group, with a goal of creating a normal distribution of scores. They are designed so that half of test-takers must pass and half must fail. Criterion-referenced tests compare performance to a set of objectives or mastery benchmarks that all test-takers can theoretically achieve at a 100% level. They provide feedback on specific course objectives rather than ranking students. Both test types have different purposes, with norm-referenced focusing on ranking along a continuum and criterion-referenced focusing on measuring mastery of objectives.