OK, constructive criticism time:
The hardest but most necessary thing for children and young people to grasp is the 'dialect' or 'register' of written English. The main reason for that is that we don't usually speak in sentences and paragraphs. We break off in the middle of what we're saying, we correct ourselves and track back, we use a lot more pronouns and stand-in phrases (ie lots of 'him' 'it' 'there' 'here' etc) because it's obvious from the context (eg through gesture or what someone said earlier), what we mean. We interrupt each other a lot, finish what the other person is saying, we break off and then start again...and so on.
Most writing (apart from speech in fiction and plays, and for certain effects in poetry) is not like that.
So, it seems to me that the key thing we can do to help children and young people is to get them to a) make transcripts of their own speech b) make transcripts of other people's speech c) compare these with various kinds of formal prose taken from newspapers, magazines, novels, instructions or whatever.
Then, just as people do in learning a foreign language, attempt 'translations' of the informal speech into formal prose, looking at why and how things are changing to go from one to the other.
So, rather than fatuous decontextualised sentences, some of which don't make sense and couldn't ever have been said or written, tests of any kind could follow this model of working between speech and writing, comparing the two, coming up with written forms of the spoken and so on. There would be of course various possibilities in terms of 'correct answer' - but then that's real language in use. There are always acceptable alternatives.
The 'tailback' in terms of teaching would be much more interesting and exciting than the tailback implied by this awful and silly test. Making transcripts of real speech brings you face to face with real language in context and raises all kinds of questions about whether people know what they're saying, whether it matters if they've made errors or said things that are ambiguous and so on. Trying to turn this into written language is hard but it's the key challenge for all children and young people. The moment you try, you come up against how to construct formal sentences. That's the challenge, so why not cut to the chase and get children and young people in the laboratory of real language in use trying to do it. This new DfE test will simply create a vast industry of exercises, cloze procedures, newsagents guides to grammar and the rest. For many children it will be an occult art, for the simple reason they don't read enough books for pleasure to assimilate most of this stuff through the texts they're reading.
By the way, the great grammarian and linguist M.A.K. Halliday nearly always (or always!) used language examples that had either been said or written. He didn't make up these spurious decontextualised ones. Real grammarians know that they have to start from real language in use.
