Garbage In, Garbage Out On Studies Concerning Which Countries 'Lead' In Education
from the does-it-matter dept
We’ve been discussing how bad metrics can lead not only to bad conclusions, but also to a tendency to optimize for those bad metrics. That seems to apply to some of the recent news concerning “lagging” by the US in the OECD’s education metrics. However, as Greg Ferenstein highlights in a piece over at TechCrunch, these stats may be meaningless. Not only has the US pretty much always lagged since these stats were first calculated, there’s little indication that the stats being measured are representative of anything actually important.
Ferenstein’s article is well worth reading, but it highlights a few key points. First of all, what the OECD is measuring might not be particularly meaningful:
However, the report implies that education translates into gainful market skills, an assumption not found in the research. For instance, while Chinese students, on average, have twice the number of instructional hours as Americans, both countries have identical scores on tests of scientific reasoning.
Second, whenever you’re talking about aggregate numbers, a lot of important nuance can be lost in the mess. For example, there is evidence to suggest that while the US may not be good at educating everyone, it does seem to do quite a good job with taking its best students and preparing them for the world.
Most importantly, the innovators at the helm of an economy come from the top quarter of students. While the United States has a dismal track-record of inequality, we treat our brightest minds quite well. The “average test scores are mostly irrelevant as a measure of economic potential,” write Hal Salzman & Lindsay Lowell in the prestigious journal, Nature, “To produce leading-edge technology, one could argue that it is the numbers of high-performing students that is most important in the global economy.”
The United States, they find, has among the highest percentage of top-performing students in the world.
Now, there are reasonable policy arguments to be made about this, and about what it means for everyone who is not included in those “brightest minds.” You can reasonably argue that we should be seeking ways to move more of the “non-brightest minds” students into the “brightest minds” category. But we shouldn’t take the aggregate data of ill-performing education metrics and from it assume that the entire system is broken. When you look for an across-the-board solution to metrics like that, you very often end up wiping out the good stuff (such as how we handle top performers) in the process — potentially making the whole thing worse.
Ferenstein also points out a second point on the “economic impact” of our education system: historically, we’ve tended to fix that economically by attracting the best and brightest from other countries to come to the US for higher education and for work as well. This is something that we’re doing less and less of these days, due to ridiculous and reactionary laws on immigration and civil liberties.
None of this is to suggest that we shouldn’t figure out ways to improve our education system. It does remain pretty clear that good education is important to other aspects of the economy. But we need to be quite careful in understanding what we’re really measuring and what it’s tied to, rather than just accepting that if one report says we’re “lagging,” it must really be an actual problem.