A Statistical Analysis of the O E O Experiment in Educational Performance Contracting

Private educational firms claimed to be able to substantially increase the reading and mathematics achievement test scores of disadvantaged students. The OEO Performance Contracting Experiment tested this claim. Although analysis of experimental effects is complicated by imperfect matching of experimental and control students and by measurement error in pretest scores, there are statistical techniques for dealing with these problems. Our results indicate that the private companies which participated in the experiment did not have the capability of bringing about any great improvement in the educational status of disadvantaged students.