The Heavens May Fall: School Dropouts, the Achievement Gap, and Statistical Bias

Abstract

In this article, the author examined the AG by noting that the NAEP reported math scores among 17-year-olds is upwardly biased when compared with the scores of this same cohort 4 years earlier when they were 13 years old. It is upwardly biased because an important event happens over the 4-year period between 8th and 12th grade—a large percentage of students leave school without graduating, and they tend to be the poorest performing students. The analyses indicate that there is indeed a considerable amount of bias in these test scores, for all three ethnic-racial groups and thus for the reported achievement gaps. There are two conclusions to be drawn from the analysis. First, it is inappropriate to compare the eighth-grade scores with those that occur 4 years later when they were 17 year old. Over the 4-year period, the population of high school students has changed as a result of significant events, such as dropout rates. These rates are especially significant among Black and Hispanic students. Second, because of such bias, the achievement gap, as reported, is misleading and adjustments need to be made if we are to get an accurate portrait of student achievement and the achievement gap.