Statistics, probability and measurement

I just finished reading The Drunkard’s Walk: How Randomness Rules Our Lives by Leonard Mlodinow. It gives a history of statistics, probability and measurement, and how they are used currently to inform and how easily they can misinform when used improperly. Also in the book he highlights Washington’s Bill Gates for special mention about the misuse of information when he sought to “fix” public education by deciding that small schools were the solution, the silver bullet, to make all perfect with public education. After disastrous results, he regrouped and is now pushing charter schools and teacher pay tied to test scores as the panacea, even though there is no research which shows this will do any better than what is happening right now. But no fear, facts will not stop this effort; he has the money to burn to push his notion of public education. After all, if you had any good ideas, you would be as rich as he is.

This tied right in with a radio interview I heard recently on NPR with Rep. Lamar Alexander. He was outraged that the U.S. government now had a significant investment in General Motors and “535 members of Congress will be trying to tell auto manufacturers how to make automobiles.” I appreciate his concern; after all, Congress, state Legislatures, and influential, I mean rich, members of the public all want to exclude educators while they tinker with public education, and all the while blaming educators for being less than perfect. As an engineer who became a teacher, I can state with total conviction that it is far easier to design and build a quality automobile than it is to educate a child. And in auto manufacturing, if you repeat your process, you repeat your results. As any parent of multiple children can attest, repeating the process on raising a child can have dramatically different results. As Nobel laureate Max Born once said, “Chance is a more fundamental concept than is causality.”

Mlodinow also dove into the misuse of testing in education. Testing serves a useful purpose in education, but as soon as something is measured, it is subject to misuse by people who not only do not understand what measurement really is, but also that all measurements have a variability, or error. A good example he uses is the SAT scores which have an error of ± 50 points. So if one person scores 20 points higher than another on the SAT, they are really tied because you cannot tell who actually had the higher score. He also described a similar occurrence with the first election of our current governor. The candidates tied because the difference between the two candidates’ vote totals was less than the error of measurement. They could have flipped a coin to determine a winner, but our state has a more complicated system to break the tie. People who don’t understand measurement would be outraged by deciding our governor by flipping a coin, so we have a much more expensive but equally random method for breaking the tie.

Thus we have these real life math problems that don’t go away just because you are a billionaire who doesn’t comprehend the research in a field you know nothing about. Very real problems for someone who, because of a societal lack of understanding regarding randomness, reliability, and measurement, is condemned to a life of lost opportunities and lower earnings due to the denial of a high school diploma because of the misuse of a high-stakes test.