Random Variation in Student Performance by Class Size: Implications of NCLB in Rural Pennsylvania

Schools that fail to make "adequate yearly progress" under NCLB face sanctions and may lose students to other schools. In smaller schools, random yearly variation in innate student ability and behavior can cause changes in scores that are beyond the influence of teachers. This study examines changes in reading and math scores across Pennsylvania's schools over time. There is no evidence that rural or smaller schools are systematically disadvantaged by NCLB. In smaller schools, 80% of the increase in scores is estimated to be caused by factors that generally cannot be influenced by school staff. Some schools likely should have received an award in a given year, but failed to earn one because nonpersistent factors reduced their scores to a level that disqualified them. Poverty depresses the gains achieved by schools, but small schools are able to offset the negative effect of poverty on changes in scores over time. Also, increases in poverty in a district reduce the gains in scores that can be achieved.