The test that is not a test

Another old post I didn't get round to finishing! This all still appears to be true, so I decided to finish and post it.

One of the things that comes up in class regularly is short tests. I was initially quite perplexed by them, and have eventually, I think, worked out what's going on.

GenkiJACS daily and weekly tests don't seem especially good as tests go. Why? Because the teacher goes through the vocabulary items from the test, then does the test. As a testing strategy to establish how much you've genuinely retained, that's got some obvious problems. But I don't think they are actually intended to work as "tests" in the normal sense at all - on reflection, I suspect they are actually a learning technique rather than a test of knowledge, although they do check what you've learned to some extent.

Basically you get exposed to some language, use it a bit in class, get told a day or so later that there's a test, go through the vocabulary for the test, then do the actual test, then go through it immediately. I think what it does it set up a situation where remembering the vocabulary feels important, so when you go through it you do cling onto it. You also combine multiple skills in quick succession:

• try to remember and say it when a cue card comes up
• see it and hear it from the teacher
• see the cue card again and write it in silence for the test
• get the card again and be prompted for the answer during marking
• check and correct your neighbour's answer
• if you got it wrong, get feedback on what was wrong
• hand tests in and get written feedback later from the teacher

You combine several attempts to recall the term, as well as spoken recall, listening, writing and reading out. The only thing you don't really get is reading comprehension (or use in context, but that's a bit different). Combining these different modes makes it a bit more likely you will remember the term. It's essentially like a live interactive version of flashcards.

My impression, which is all I can give, is that this actually works reasonably well as a vocabulary learning technique, while being considerably more interesting than going through drills to get the same amount of exposure. However, I think it will only work with a limited amount of language, so I don't think it's something that would scale up too well.

I should note that there are also more traditional tests every few weeks, covering everything from vocabulary to grammar and contextual use. These act like tests usually do, with no pre-drilling or anything, and are used to check whether you're ready to progress. The difference is significant.