Friday, November 18, 2011

Test scores are low, make them go up

“Errors, like straws, upon the surface flow;
He who would search for pearls must dive below.”

~ John Dryden

Some people make the case that good teaching and real learning are the best ways to achieve high test scores. The argument goes something like this: Help your students become thinkers and learners and the scores will take care of themselves.

Let's for the moment ignore the evidence that suggests this might not be true. 

There is a growing body of evidence that shows you can improve test scores without improving learning. Some of these strategies range from illegal cheating such as changing students answers -- to legalized cheating, otherwise known as teaching to the test

Cheating has always been frowned upon, and at one time so was teaching to the test, but in response to the ever increasing demands for higher and higher test scores, teaching to the test has gone from something even the test-makers vehemently opposed to something that is encouraged as best practice. And when the pressure to raise scores threaten the very existence of schools and the employment of educators, system-wide forms of cheating are encouraged from as high up as administrators and superintendents.

In reaction to this systemic pressure, many educators now experience professional development as nothing more than "test scores are low, make them go up."

Anyone who has played the game of school as a student knows that there are ingenius ways to achieve high grades without learning too much and there are also ways to learn a lot and still earn low grades.

Here's what I mean. Here are two students:

One student has a teacher that artfully guides them to play an active role in constructing and discovering a love for learning; the other has a teacher that focuses on scripted test preparation and direct instruction that has this student play a passive role in reproducing right answers on tests. The former experiences school as something done by them -- the latter as something done to them.

Here are their testsandgrades data:

So which student is which? Which one sees learning as a means to an end and which sees it as an end in and of itself? Which one has a teacher that artfully guides them and which has a teacher that does test prep?

The information provided about Liam and Samantha is as common-place as it is limited, which means we can't possibly know which student is which based on the (lack of) information provided --and when I say we need more information, I don't mean we need more testsandgrades.

The ultimate point here is that inferences made from testsandgrades can be undermined leaving the successful and superficial students indistinguishable. The reductionist data gleamed from testsandgrades is by definition necessarily incomplete -- making the use of this limited and narrow kinds of data to make any kind of decisions or judgements about a student at best unhelpful and at worst harmful.

Perhaps the most disturbing part of this whole debacle that I call chasing "testsandgrades and calling it learning" is that there is a whole generation of beginning teachers who have known no other education than one that makes testsandgrades both the means and end to learning. The distinction many people make between real learning and testsandgrades makes no sense to them because they have been taught from the beginning that testsandgrades are good teaching and learning; they simply do not have any other model to draw from.

The mantra of test scores are low make them go up and the cult following we have in our culture of education for testsandgrades may prove to do more long-lasting harm to our education systems than we can fathom.


  1. Thank you so much! We are forced to spend hours looking at data that MEAN NOTHING. Without knowing the specific questions students are consistently missing and which ones they're getting right, there's no way to even begin to see if there's a pattern to it. It doesn't do me any good to know that my students' area of weakness is "literary response and analysis" if I don't know how that is assessed each time. Low test scores almost always boil down to inadequate literacy, and there is no test-prep curriculum existent that addresses that problem.

  2. Incisively argued. Thanks.

    Heads and teachers appear trapped in a game that is too dangerous to question. What can they do that will not jeopardise their jobs?

    Centralised force from democratically elected governments is hard to counter, even if it becomes anti-democratic.

    There are similarities with health and nutrition - statins, for example, are prescribed at ever low cholesterol levels, simply to make pharmaceutical industry more money. We become caught in a game run by power-mongers.


Follow by Email