Freakonomics and education

This post began as a comment on Joe Bower’s blog, for the love of learning, in response to his post, “Shame on you Steven Levitt.”

I’ve been listening to Steven Levitt’s Freakonomics the past week while commuting. While I find the perspective and insight fascinating, I truly detested Levitt’s handling of the “cheating teachers” studies in Chicago.

His perspective on education is obvious when he talks about “bad teachers” and “good teachers” (as defined by standardized test scores). The disdain with which he talks about the teachers supposedly proven to be cheating the tests is also disappointing. The real problem, though, is that he asked the wrong question. The question shouldn’t have been, “do teachers cheat?” Sure, some teachers cheat – so do some people in every profession in the history of the world. Teachers are human too.

The real question should have been, “why are these teachers cheating and what does it tell us about the educational system?” Or, better still, “are standardized test valid measures of teaching and learning?” Instead, huge assumptions were made:

  1. Teachers are cheating to make themselves look good and to get pay raises (or avoid being fired) – while this may be entirely true, individual teachers’ motivations may have been drastically different;
  2. Standardized tests are valid measures of teaching and learning – obviously I don’t believe this at all;
  3. Test scores should be relatively stable from year to year – anyone who has pored over data from frequently administered tests knows this to be untrue. I’ve seen kids’ scores on the MAP test go up and down dramatically within a year

No debate as to the validity, reliability or value of standardized tests is mentioned. No concern about cohort effects is addressed.

Why should we expect an economist to understand the full scope of the issue?

We shouldn’t.

And yet, considering Levitt’s premise of using data and statistics to tell the story behind the story (I found his analysis of cheating in sumo wresting fascinating) I think he missed the true opportunity here.

The story Steven Levitt should be telling is the correlation between family income and test scores and what this indicates about our educational system and the future of our nation.

(Dead) end of course exams

I used to think that end of course exams would be a good thing in Algebra, Geometry, and Biology. Now I see the error of that way of thinking.


Right now, my state’s science exam takes place after 10th grade.  It covers everything students should have learned since 8th grade.  Many of our state standards are about inquiry, systems, and application.  That leaves me lots of latitude to teach Biology content in 9th grade in an inquiry-based, real-world relevant, project learning style.

Instituting a Biology end of course assessment would make much of that impossible.

Dead End

It would bring much more pressure to teach very specific Biology content to prepare them for the state exam.  It would probably force us to move Biology to strictly a 10th grade class.  It would leave little room for interdisciplinary projects.

It would be a shame.