THOMPSON: Counting What Really Counts
Two cheers for Schoolboy's reminder that we need early warning systems for potential dropouts. As early as 6th grade, we may be able to identify with 75% accuracy the individuals who are at most risk. We still need to consider whether a student with low attendance needs different interventions than a student with behavior problems or low grades, and we can not forget that dropout prevention has not had a long record of success.
Schoolboy, however, used a poor metaphor when he asked whether indicators were like cholesterol tests which can clearly improve outcomes or prostate cancer tests. If we have no treatments and early diagnosis can do little more than upset patients, then early warning is of minimal value. But prostrate cancer is not contagious. Classroom disruptions by failing students are.
We should also remember that two "treatments" show promise. Lynn Canady explains how we can easily identify students who are failing by October, but the system just lets them fail until June. The answer, says the famed "smart ass," is to make them "fail faster." Canady has a number of intriguing alternative schedules that allow for intensive instruction in reading and math that would allow students to rejoin the "regular" populations after a semester. The What Works Clearinghouse reports that Accelerated Middle Schools can increase the rate of "staying in school" by 18 percentile points and of "progressing in school" by 35 points. I am not sure what is the difference between Accelerated Middle Schools and high-quality alternative schools, but if a name change creates an opportunity for open-minded discussion then it’s all good.
Update. A new tirade regarding graduation rate accounting can be found here. Someday, the theorists should attempt a thesis sentence, such as "I believe that better graduation accountability models will help increase the number of qualified high school graduates because ..." John Thompson