(I’m going to start this post by pre-empting some criticism: Of course not all assessment is machine markable, or externally markable. Of course not. But some of it is. We are currently in the frankly unbelievable situation where some of the most skilled professionals in our whole society — teachers — are spending over ten hours a week doing work that can be done instantly and easily by machines, and I don’t know why more people aren’t enraged by that.)
With that out of the way — consider the following two content objects:
1. A blank test on amphibians
2. Vivian’s completed test on amphibians
There is a fundamental difference between these two entities: 1 is replicable, and 2 is non-replicable. 2 is unique.
As a matter of fact, if 1 is *not* perfectly replicable, 1 fails — everyone in Vivian’s class must receive an identical blank test for the results to be valid. So replicability is a necessary feature of any assessment object.
(With my UX hat on, I’d be inclined to extend that criteria to *easy* replicability — and the easiest option right now is still paper. The sequence of print > photocopy > distribute results in a pile of physical objects that are very difficult to corrupt or lose — it’s a hard model to beat for simplicity. By the time you’ve found, powered up, and logged into a computer, you’ve already got a longer and more complex sequence running.)
So: for publishers who charge for content, this presents a problem. Your content must be replicable, and the harder it is to replicate, the less likely it is to be used. But if people replicate your content for themselves instead of paying for it, you lose revenue. The only way to stop the would-be replicators (or at least, slow them down) is to make the content harder to replicate, which is to say, harder to use, which is to say, less likely to be used, which will also reduce your revenue. Dilemma!
Here’s another option:
Don’t charge for replicable content.
Charge for unique content.
The blank amphibians test is replicable, but Vivian’s completed is unique. You don’t have to worry about replication, because there’s no reason to replicate it.
If, rather than charging for distribution of content, you charge for providing a service to content — marking completed assessments — the problem of replication vanishes instantly.
Looking backwards to my initial caveat — no, this model most definitely does not apply to all assessment. There is no external or machine based assessment mechanism that can measure effort, for example, or listening skills, or group work — to name but the tip of the iceberg. But it could apply to some.
In this post I’ve been concentrating on the benefits to publishers, but there are benefits to educators too. If, by redeploying a similar amount of money to what is currently spent on content to marking services, you could remove even 10% of the assessment burden from teachers, wouldn’t that be a good thing?
Of course it would.
(This post was originally posted on my old blog, at zoero.se. Comments may have been lost.)