[snowwrestler]

As I understand it, the basis for the Gates Foundation position on education is that education produces measurable outcomes, so communities should actually measure those outcomes, and then use that data to improve education.

I think that point of view should be right at home here on HN. I bet most of us believe that measurement, testing, and iterative development are helpful in producing a good product or service. However, most of the U.S. education does not work that way, and parts of it (teachers' unions) are actively hostile to it.

Some schools produce outcomes that are so good that they are "visible to the naked eye". You don't need sophisticated testing regimes to know that Phillips Andover or Bronx Science produce strong graduates. It's no surprise that the richest man in the world has picked one of those schools for his children. You probably would too, if you could.

But, most people can't--they rely on public education. And in the U.S., most school districts produce outcomes that are not obviously good, or in some cases are obviously bad. In those situations, a clear set of standards and measurements seem to me like something worth trying.

Many teachers don't like it because it puts their jobs at risk, potentially based on outcomes that they have little control over (educational outcomes appear to depend heavily on outside factors). They are also worry that the curriculum will be developed by administrators who don't actually know anything about teaching.

Parents don't like it because new currulica don't match what they learned in school as kids. Forgive my bluntness, but I think that is stupid. If there is a better way to do things, we should change.

Students, many of them, don't like changes that make them work harder in school. Personally, I'm not very sympathetic to that either.

[rodgerd]

> I think that point of view should be right at home here on HN. I bet most of us believe that measurement, testing, and iterative development are helpful in producing a good product or service

Yeah, until we start trying to measure programmer output and then it's all on about how impossible special snowflakes are to measure.

[snowwrestler]

Actually that seems like a very insightful analogy to me.

[sheltgor]

One of the big issues with the testing argument is that, the way common core is implemented at least, it produces no real meaningful change from past failed testing initiatives like NCLB.

An increased emphasis on tests sadly tends to lead to teaching to the test. I experienced this in the early days of NCLB even in a gifted and talented program where several weeks leading up to the test ate up time outside of normal class in order to prepare for the battery of standardized tests that had little relevancy to much of what went on in the classroom. In schools where instruction is already poor, it tends to take even more focus away and devote it towards this specific niche.

The other big issue is more one with common core's implementation itself where Pearson has ultimate control over the incredibly lucrative market, and the various contractors they are working with tend to do a fairly shoddy job of it. Test question writing has devolved to such a poor point, due to endless cheapening and outsourcing in order to lower costs, that a question on a major test may net the writer a few dozen dollars. Forgive the anectodal evidence, but my mother worked for a long period as a freelance writer, and when she considered going back to it recently she discovered that many of the questions, ones that have an enormous impact on education, are going for a pittance. Further, in her current occupation teaching adult basic ed and GED prep (they have to work towards common core too), much of the official testing materials and prep questions for educators have been so poorly designed that neither the students nor the teachers can understand them.

So yes, quantitative measurements can certainly be a good thing in order to help support improvement, but the current testing model is horrendously broken and moving further towards it is not really the right way to go.

[snowwrestler]

Common Core actually does not stipulate any particular implementation in terms of curriculum or testing. It is simply a set of standards for educational achievement that the states have agreed upon. So for example it would say that a high school graduate should be capable of solving a particular equation. It is up to each state (or district) to figure out how to teach that knowledge, and how to measure it.

Most states have not even really started to do that yet. What we have now are a few commercial companies who are rushing shit products out the door with a "Common Core" sticker slapped on them, and the world is judging the entire concept based on these crappy products. See for example:

http://www.npr.org/blogs/ed/2014/06/03/318228023/the-common-...

> An increased emphasis on tests sadly tends to lead to teaching to the test.

This is not a dependent relationship; it is a choice. It is up to administrators and teachers to decide how curricula are designed and teachers teach.

The idea that setting standards automatically results in worse performance would not make sense in any other context; yet it's taken as a given in education.