Fattening the Pig: Testing Versus Teaching

A colleague of mine returned recently from New Zealand, where he had taught for several years. ‘The kids there don’t write tests’, he told me. ‘Just an examination at the end of each term.’

That news fits in with the reading I have been doing over the past 10 years regarding the testing versus teaching debate.

In Finland, schoolchildren also write very few tests, and hardly any standardised ones. When formative assessments are used, these are often not numerically graded, but rather annotated with comments such as ‘good work’ or ‘needs more practice’. The results are often not shared with students or parents, but are used by the teacher for planning.

Yet when Finnish students write their finals, they outcompete South Africa by miles. This is evident in the results of the annual TIMSS and PIRLS, where Finland routinely comes in the top five countries in the world, and we come last or second-last.

There are other top countries that test regularly – Singapore and South Korea come to mind. But there, the pressure to perform is so intense that suicide is an unfortunate feature of the school-going community. Is that what we want?

In much of the Western world, assessments (as we understand them in South Africa) are few and far between. The reasons are obvious. Firstly, every time you test, you are not teaching new content. Preparing for the test, writing the actual test, and going through the test afterwards, all take time. In contexts where less time is spent on tests there has to be more time for teaching new content.

Secondly, many people believe that tests ‘don’t work’. This casual phrase needs some unpacking. In conference after conference, I have heard international speakers using it. What do they mean by this? I have certainly found tests useful in sorting out who in my class understands the work, and who doesn’t.

Tests, marks and performance

Perhaps what people using the phrase mean is, ‘tests don’t improve marks’. This certainly seems to be true. While a learner’s marks can improve over time, other interventions, rather than assessments, are usually the cause.

Evidence for this is found in Hattie’s well-known ‘effect sizes’, where testing ranks as only the 66th most effective way of raising student achievement. So, perhaps, one or two tests during the year might suffice to separate the slackers from the achievers?

Why, then, does South Africa not follow international best practice? The answer may be in habit, in tradition: if a school were to stop testing, you can bet that many of the parents would be up in arms. These parents may not be aware of international trends, or our standing in the world rankings. Zimbabwe, Botswana, Zambia, and most other African countries outperform us in mathematics, science and English.

Our top independent schools may think this does not apply to them, yet the Human Sciences Research Council in Pretoria has determined that our very best students are only as good as the average students in the world’s top six countries.

There are several ways to fatten a pig, but repeatedly weighing it is not one of them.

This term alone, I have had to set 17 assessments. Each of these has to be moderated three times: once by a subject specialist and twice by a language specialist (we are a bilingual school). Each one then undergoes post-moderation. Each moderator has to sign (pre and post) as does the Further Education and Training head. That is over 100 signatures I have to collect, in one term.

This is not an efficient system. It is a system almost designed to keep the marks low. My focus is certainly not on planning creative lessons.

Testing versus teaching: what is the correct balance?

Testing versus teaching: we need to find a better way

Testing merely monitors the situation. Similarly, repeated class tests do not improve marks – they tell us what we already know. What type of system might work better than our current one? How about just examining the students at the end of each term?

Of course, a teacher may do as many informal, formative assessments as he or she likes, and choose whether or not to share the marks – or mark them non-numerically, as in Finland. But these assessments of the teacher’s choosing should not be standardised or specified by management. They should be tailored (by the teacher) to each class’s needs.

The counter-argument to the above, one I have heard for three decades now, is that you need ‘enough’ marks to get a fair reflection of the student’s progress each term. To this I would answer: if the student fumbles their first term exam, they have a chance to fix it in term two. And in term three. And in term four.

If you are a three-term school, you have longer terms and can perhaps fit in a ‘mini’ assessment half-way along. But again, many overseas schools do not bother with this. It is only in South Africa that we are so obsessed with marks.

Is it not perhaps time to modernise – to use peer-to-peer teaching and other innovative interventions, as opposed to testing as the primary intervention? We need to realise that South Africa sits at the bottom of the pile and over-testing may be one of the reasons for this.