Drawing Conclusions from Educational Experiments
Under increasing pressure to produce evidence of greater student achievement, schools are eager to adopt practices that have a proven track record. This has led the Houston public schools to engage in a novel experiment called Apollo 20 ("A School District Mimics Charters, Hoping Success Will Follow," The New York Times, Sept. 6).
By implementing five key tenets of successful charter schools (such as the Knowledge is Power Program) at nine secondary schools last year, the district hoped to be able to post similar results. One high school reported gains that were enough to move the school to acceptable after many years of being unacceptable. But four remained unacceptable, and the rest saw the percentage of passing test scores drop. Undeterred, Apollo 20 will be extended to include 11 elementary schools this school year.
Now here's where the matter gets interesting. Roland Fryer, an economist and head of Harvard's EdLabs, intends to compare the results of the Apollo 20 schools (the experimental group) with those of other Houston schools with similar demographics (the control group). The problem is that even if the Apollo 20 schools outperform the other schools in the district, it will be impossible to know for certain which factors were responsible. Was it (a) longer day and year (b) more rigorous hiring (c) frequent testing (d) "no excuses" culture (e) all of the above?
I say that because one vital factor is given short shrift: Parents who choose to enroll their children in Apollo 20 schools signal they are involved in their children's education. Time and again, research has demonstrated the key role that this factor alone plays in student achievement. To put the matter another way, if all parents whose children attend traditional public schools in Houston were required to sign a contract spelling out their responsibilities, I believe that this prerequisite would likely outweigh (a) through (e).
This point was underscored by "What Teachers Wish You Knew" by Lisa Collier Cool (msn. Lifestyle, Sept. 6). Nevertheless, the results of the experiment will be picked up by the media as "proof" that charter schools are superior to traditional schools. Only sophisticated taxpayers will know how to correctly interpret the results.