Monday, February 25, 2008

teaching to the test

Teach for Australia says this:
Teach for Australia would develop standardised literacy and numeracy tests for the Fellow to assess their students. These would be short, low-stakes tests conducted every month that track each student’s performance in the key skill areas

... we need a no-excuses, unrelenting focus on performance in Australia’s remote schools. Monthly tests in literacy and numeracy are a key mechanism to achieving this. Good teachers, of course, are likely to do daily or weekly mini-tests (in addition to the monthly tests) to gauge student progress and determine areas of weakness.

Will this mean that the Fellows will simply ‘teach to the test’? Quite possibly, but if the tests are well constructed and properly assess the knowledge that students are supposed to learn, then ‘teaching to the test’ presents no difficulties.
- Teach for Australia
I think this needs to be discussed and fleshed out a lot more. The main potential problem of teaching to the test is rote learning - knowing an answer is "correct" but not understanding why it is correct or how that knowledge might be applied in some sort of variation of the theme or "real life" situation.

Of course there is a huge literature on this and the curriculum wars rage on unabated. Here is one example, from many:
... many children who correctly answered pencil-and-paper fraction questions such as 5/11 x 792 = q could not pour out one-third of a glass of water, and of those who could, only a small proportion had any idea of what fraction of the original full glass of water remained.
- Fractions: A Weeping Sore in Mathematics Education
My view is that good teaching methodology is a continuum from constructionist to instructionist and teachers have to walk the walk along the whole of that continuum.

Related: Noel Pearson's "radical centre" concept applied to education


Mark Miller said...

I think the dream right now with the standardized testing is that everyone will know something about each of the covered subjects, even if the depth is shallow, as opposed to nothing. It turns out there is a risk to good schools, as I believe you are aware. There is a kind of reductionism about it that tends to pull everyone towards the "average", not in terms of score, but in terms of knowledge. I remember telling you before about the example of the science teacher who felt conflicted about teaching her students about the vents at the Mid-Atlantic Ridge. She feared they would get the answer wrong on the question, "What is the source of energy for the ocean's food chain?" The "correct" answer on the standardized test was the Sun, but the discovery of the vents has opened up that question. Perhaps it's chemosynthesis instead.

Perhaps the form of the questions doesn't fit the subject well. Science can't just be reduced down to pat answers. It's also about the scientific method; a spirit of discovery and analysis. For example, it might be better to ask analysis questions instead, like, "How did Galileo discover that everything falls to Earth at the same rate?", and then offer some possible experiments that might show this to be true (with one being the experiment Galileo used), and ask the student to pick one.

I think I remember my standardized tests for college entrance asked about my knowledge of chemistry, asking me to take two elements, combine them together, and answer with what the result was. That would also be a valid test of scientific knowledge.

Bill Kerr said...

hi mark,

Another point is that standardised tests are now being used to compare the education systems of countries - and then questions are asked like: Why is Finland number one?

This makes it even more important to evaluate what this whole process really means