For assessment to be meaningful, it must be formative rather than summative — that is, those who are being assessed must benefit directly from the results of the scoring. When students know ahead of time what is expected –– and can even participate in developing the scoring guide –– they will know how they measure up against rubrics or standards. They will get an opportunity to reflect upon and discuss how they would change what they have done, in order to deepen their understanding of the problem or task at hand, after their work is completed. Students may be given an opportunity to redo or take their work to the next step, demonstrating a higher level of competence. Traditional standardized high-stakes tests fail on all aspects of this scenario. If any change in instruction occurs as a result of the test scores, it will benefit those who sit in the student desks or take the test the next year. The most powerful measures of understanding are built into student work as part of a continual feedback loop.
   -- from Assessment Outside the Box by Sara Armstrong CUE Newsletter, Oct. 2001