What’s a natural experiment? No Child Left Behind edition


In a particular school district where NCLB dictated better information about schools for parents and allowed them to choose schools, some kids got randomly assigned to better schools. Lo, they performed better.

3 thoughts on “What’s a natural experiment? No Child Left Behind edition”

  1. Standard question: how do they define performance?

    The main criticism of NCLB is that it forces teachers to explicitly teach students how to pass the tests, as opposed to teaching useful skills like critical reasoning, self instruction, and problem solving. Judging a school based solely on its NCLB testing performance is dangerous. Does that mean that the school presents a well-rounded education, with teaching methodology suited for your kid, or does it mean that the school concentrates on testing at the expense of other subjects?

    Look at the SATs, for example. College bound students and their parents are frightened of the SATs. We see SAT prep classes, SAT camps, SAT study programs, SAT guides, SAT study groups, SAT instructional videos, SAT breakfast cereal, practice SAT sessions, etc. Students take the test over and over to try to maximize their scores. Are they learning anything about the subjects presented on the test, or are they cramming facts into short term memory to artificially inflate their benchmarks? Do students retain much information after the tests?

    Would you choose a doctor based solely on their MCAT score?

    The premise is that the testing methods influence the results, often in undesirable ways. In this case, we might have some useful information about NCLB’s influence when the first batches of these kids reaches the workforce.

  2. I think they have several measures of performance, e.g. truancy.

    The point, though, is that whatever you think about NCLB, it provided a natural experiment, randomly assigning kids to good and bad schools.

Comments are closed.