Thursday, December 10, 2015

Obligatory Language Assessment and New Scots: An Examination of Test Validity

As Lyle Bachman didn't say to me, that evening we drank too much whisky in Boston, "It's all about validity, kid. The rest is just conversation." But it's a sentiment he might approve, if we mean consequential validity. We need to be sure about test fairness and reliability, too, but validity (construct as well as consequential) is The Father in that Trinity.

See, test validity is something people in education bang on about without really understanding it. They want valid tests, naturally, but don't want to invest in the time and independent research that's needed to validate a test. They just don't get it. On top of that, it's a complicated thing, not quite quantum mechanics, but not straightforward, either. Even people who specialise in Assessment, (I don't exclude myself) don't always know this essential concept back to front.

Yet it's vital. The "people in education", (for example, managers in a private English language teaching organisation) I referred to above will say. "We must have valid tests!" "But is it valid?" And they're right, a test which is not valid has been a waste of everyone's time, and a waste of money. Which has consequences for the organisation, and especially for the test taker, and with high stakes tests for society at large.

A test which can decide whether you can remain in a place where you, (and perhaps your family) have begun to put down roots, or whether you need to be shunted off to another place, and perhaps have to get to grips with a whole new testing regime, or even language, is very high stakes. If we're going to oblige folk to do such tests, then we must be belt-and-braces sure that the test is fair, reliable, and valid.

They are the three horses I have now put before my cart. That vehicle, we will go on to see, contains the other things I have begun to look at: emergent bilingualism; "native ability", "mother tongue", language ability and taxonomy as routes to "othering"; the process by which new Scots come to be tested (and come to be in Scotland, and where they come from, geographically, politically, and linguistically); the "ethically problematical" (Lyle really did say that) use of language testing as a means of deciding who can come to, and must go from, a country.

Voila. Eureka. And khalas. Well, not quite khalas. Everything I have read and written in the last few months must now be regarded as non-essential backgrounding. I now have to begin reading on test validity, big time, and re-write my 3 formative submissions on this module in line with this new framing.

[NB - I've started using labels again, and will continue from this post onward. I need to bear in mind that they will likely form the basis of keywords in my thesis].