Regressions and Growth
One of the strongest critiques of No Child Left Behind’s annual testing and proficiency targets was that they didn’t sufficiently account for students who made significant progress but still remained below. A student who started at far below grade level could substantial academic gains but still fall short. The school would not be given credit for this progress, or growth, and could face sanctions because of it (the law does contain a “safe harbor” provision for sub-groups that make significant annual gains, but that tracks the progress of schools and sub-groups within schools, not individual students).
In response, the Department of Education launched a pilot growth model program that allowed states to come up with their own definitions of growth. States like Iowa created a relatively simple model that essentially created more levels than just the one existing proficiency bar. If students jumped from one level to a higher one, they would be considered proficient, even if they remained short of the eventual target. They had three years to complete their growth and be on grade level. The image below comes from an evaluation of the pilot program (.pdf), and it shows how Iowa’s tier system works (click on it to make it larger).