Tags

, , , ,

As part of our class we are supposed to read Freakonomics, and we’ve been assigned chapter 3.  But I’ve always wanted to read the book, so I figured I’d just make it happen by going with the idea that it would be weird to read the middle of the book before the beginning.  I’m glad I did too, because the story of the cheating teachers in the first section was quite intriguing.  The Chicago Public School system had a ton of data on standardized tests, and a new CEO decided to pursue the reports that teachers were cheating.  With the help of economists the school district was able to pinpoint a handful of obviously cheating teachers and remove them from the school system.

The way in which this was performed is what I found most interesting.  Essentially, the team working on the case thought like cheaters.  How would a cheating teacher go about increasing scores without getting caught?  Well, the most obvious answer is altering the answers before they are submitted.  No witnesses, no risk of students informing the authorities.  Then,  since this is likely a big form of cheating and one which is definitively illegal and might actually be provable (versus teaching to the test), they set out to analyze test scores to find patterns.

Of course teachers wouldn’t change all the answers – the would not be enough time and it would be too obvious.  Nor would they randomly alter responses: it would be too difficult since the don’t have an answer key to work from.  Instead they would probably find a consecutive string to answer on multiple answer sheets.  The first questions are the easiest, so it is probably more productive to fix latter answers.

It was really interesting to read how an analysis found patterns in the data and pinpointed strings which were too similar to be a result of chance.  By picking up on strings of correct responses by a number of students, its easy to see this might be a result of teacher interference.  When the scores were considered in terms of past and future scores they proved higher than expected.  Many students had no correlation in answers between them and another student, until these selected questions.  And some students’ responses simply did not make sense: they failed all the early, easy questions but excelled at the later, more difficult ones.  Incredibly unlikely, or staged?

I’m excited to continue reading Freakonomics because it has a refreshingly down-to-Earth feel.  Most economics books put me to sleep, which is probably why I’ve read so few.  And I have already noticed some things about the writing that I don’t particularly like, such as the “we are smarter and notice what others miss” indication in some of the points.  Still, anytime a book can both educate and intoxicate is fine by me.

Advertisements