Reading Freakonomics (by Steven Leavitt and Stephen Dubner) was supposed to be a lark for me. You know, perusing fun ideas. There are certainly some of those. Who would believe, however, that the book was full of information about educational reform! Now the title of the book really makes me smile.
Some interesting data about Chicago schools may explain U.S. Secretary of Education Arne Duncan’s thinking. For some students, there really is a problem that no one can solve alone, and it is the culture of some schools that inhibits learning. Is it better to help a few to leave a bad situation than to do nothing? Or, do you take on the whole problem knowing you do not have the resources and capability to solve it? People will be hurt; children will be lost either way. What would you do? Leavitt’s data makes you think out of the box.
Leavitt is a University of Chicago economist who likes big, complicated data sets, and he found some. What he did with them was unusual. The methodology is as interesting as the results. He asked practical everyday questions and found intriguing answers. I have excerpted some of the findings the authors reported about teaching and learning. They remind us to constantly challenge our assumptions.
How Trusting Should We Be?
W.C. Fields’ quote “A thing worth having is a thing worth cheating for” is tied to high stakes testing. The authors report what happened when the IRS began to require social security numbers for children claimed as dependents; 7 million children disappeared! When Chicago began high stakes testing in 1996, schools and teachers whose students had low scores could be sanctioned. Students could be held back. A longitudinal data set from Chicago Public Schools revealed a dramatic rise in cheating on standardized tests that year. In 2002, Arne Duncan contacted the authors of the study. He was determined to do something about the cheating and devised a plan to retest some students. Yes, there was cheating just as there was in Atlanta a couple of years ago. The cheating declined significantly once people knew that the district was watching. Of course examples of cheating abound in other fields. But, in the current climate of high stakes for traditional public and charter schools, failure to control for cheating makes all data suspect even if most teachers do not cheat. There is skepticism about the validity of school and teacher evaluations whether state run or district run. What should be done?
Who Opted Out of Neighborhood Schools?
A telling set of data about school choice revealed which parents opted out of their local school. Using a lottery system, Chicago gave parents the choice of sending children to any traditional high school in the city. About one half of the parents opted out of their neighborhood high school, and those students had higher achievement scores than those who remained. In subsequent years, achievement scores did not change for either those who stayed or left. It wasn’t the school that made the difference; it was the people who left.
Is There an Achievement Gap or a School Gap?
The black-white income gap correlates highly with the black-white education gap. Why this is true interested Roland Fryer, a young black academician at Harvard. His work on cultural norms in some black neighborhoods found strong attitudes against being a good student. You were taunted as ‘trying to be white’ instead of being proud of your race. Another of Fryer’s papers, “Understanding the Black-White Test Score Gap in the First Two Years of School” sheds even more light on the push for alternatives to neighborhood schools. Using the data from the U.S. DOE ‘Early Childhood Longitudinal Study’, multiple variables describing 20,000 K-5 students were analyzed. Achievement of low income white or black students entering kindergarten is the same if you control for family income, parental education, and mother’s age at the time of the birth of her first child. This is not news. What is significant is that the achievement gap reappears within two years between these two groups even if you control for these factors. The data from 1000 schools show that schools do matter. There is no difference in academic achievement between black and white students who attend poor schools, both groups are equally behind. Black students in good schools outperform white students in bad schools. But, many more black children than white children attend highly segregated schools located in high crime neighborhoods, and these children do not do as well as comparable low income white children who tend to be in more diverse schools.
What is a ‘Good’ School?
So what is a “good” school? It is one where the social climate values learning. Arne Duncan advocates creating schools where it is possible to change the social and educational norms. He says this in his speeches. He tries to do this by closing schools and opening charters with different teachers and administrators. Thus far, this strategy may be more disruptive than constructive. There are a few ‘wins’ but on average not much improvement. Why? Will more access to preschool help? Not if they are not high quality programs, Leavitt asserts.
Roland Fryer’s current experiment in 30 of Houston’s public schools may shed some light on future directions. His approach is to provide more and longer school days with intensive tutoring in traditional public schools. His U.S. DOE grant, however, mandated that one-half of the teachers be replaced. See: The Five Tenets of Highly Effective NYC Charters . Is teacher resistance to change really that high, or is the driving force to hire inexperienced teachers at lower salaries to pay for tutors and extended time in school? This is a legitimate question. Chicago schools only resolved this issue after a bitter strike.
Back in my days as a psychometrician, we called these questions ‘ill structured problems’ for which there was no single best answer. Answers to these questions have consequences, some good, some not. Many consequences are totally unintended. There will be tradeoffs, but no group should consistently get the short end of the stick. We should look at our local schools. Are there large differences in funding, teacher qualifications, educational facilities, and materials? What are we willing to do about it?