and fairly become critical. Improper designations or inaccurate measures could mean that schools that are making progress receive intervention inappropriately, or that students in schools that need help may not get the assistance they require.
The most common method states and districts have used to determine adequate yearly progress is to set a goal for school performance, determine how long it will take to meet the goal, define progress toward the goal, and determine how school results will be structured so that the state or district could evaluate a school's rate of progress (Carlson, 1996). One of the best-known examples of this approach is the method used in Kentucky, which has been applied in some form in a number of other states and districts.
Under Kentucky's system, the state set the overall target for all schools at the level at which all students perform at the proficient level and called this level 100. They then determined each school's baseline performance, based on the results of the initial administration of the state test—giving greater weight to students at the proficient and distinguished (advanced) levels than to those at lower levels of performance—and subtracted that score from 100. They then set each school's two-year target at 10 percent of the difference between the initial score and 100. At that rate, state officials reasoned, every school would reach the target within 20 years.
This approach depends heavily on the quality of the measures of school performance. As noted in Chapter 4, using average scores to determine school performance can provide misleading inferences. (Although Kentucky uses a weighted average, assigning different values to students at different points on the distribution, it fails to disaggregate the results or to account in some other way for the student population in each school.) The risk of misleading inferences is significant in measures of growth. As Willms (1998) points out, schools with high initial test scores tend to grow at a faster rate than those with lower initial scores. In part this phenomenon reflects the fact that high performance tends to be associated with high levels of parental support, fewer disciplinary problems, and high teacher quality—all of which can contribute to performance improvement. At a minimum, this finding suggests, comparisons of growth rates that do not take into account the composition of the school's student body may be misleading.
A second factor in the “gap-closing” model, as the Kentucky approach is sometimes called, is a theory about the expected rate of growth. The Kentucky