Tests test what they test, not necessarily what is important

Source: The Economist

An article in The Economist raises the question of whether the tests we use to measure student mastery of mathematics really measure what people need to know to use mathematics effectively in their lives and work. It is an interesting question.

I suspect that a lot of professional mathematicians are not too good at arithmetic, while the ability to add, subtract, multiply and divide is pretty useful to most people as they shop, and for many people in their work. 

As an engineering student I was told to learn how to estimate the general range of a quantity, since that kind of intuition helps to prevent the kinds of errors that result in bridges falling down or buildings collapsing.

As I learned operations research I came to have an intuition about how to approach modeling, and how detailed a model had to be to provide the ideas needed to understand a problem.

So people in different situations will want different mathematical abilities. I am not sure that the people who set school tests fully understand those abilities.

You would hate to tell another Einstein that he was not good at Maths because he didn't do well on a test that tested the wrong thing. He might  believe you!