I believe that we should get away altogether from tests and correlations among tests, and look instead at more naturalistic sources of information about how peoples around the world develop skills important to their way of life.

—Howard Gardner

The kinds of changes in instructional practice described in the previous nine chapters require an equivalent adjustment in the manner of assessment used to evaluate learning progress. It would certainly be the height of hypocrisy to ask students to participate in a wide range of multispectrum experiences in all eight intelligences and then require them to show what they've learned through standardized tests that focus narrowly on linguistic or logical-mathematical intelligences. Educators would clearly be sending a double message to students and to the wider community: "Learning in eight ways is fun, but when it comes to our bottom line—evaluating students' learning progress—we've got to get serious again and test the way we've always tested." Thus, MI theory proposes a fundamental restructuring of the way in which educators assess their students' learning progress. It suggests a system that relies far less on formal standardized or norm-referenced tests and much more on authentic measures that are criterion-referenced, benchmarked, or ipsative (i.e., that compare a student to his or her own past performances).