Because the journalist, seeking to say what the journalist thinks needs to be said, always finds a way of saying it. Leave out a fact here, leave out context there, and suddenly the journalist is always a savior.
Take this:
"West Contra Costa Unified denied 2 Investigates’ request to speak to the district's superintendent Bruce Harter, so reporter Eric Rasmussen asked Walton if he thinks the district is identifying too many kids as "English Learners."
"Well, that's something the district doesn't have too much control over," Walton replied.
Experts such as Garcia Bedolla disagree.
Neither Bedolla nor the journalist explains what Walton could mean. And both rather like their whole-part fallacy because it makes them seem superior.
Texas has a similar kind of system. Most states do. They're ultimately based on SCOTUS decisions and federal law, with some local tweaks.
Some immigrant parents are convinced that their kids are fluent enough. They do not want their kids in ELL programs. They're asked to put their kids in an ELL program and they say no. This isn't all--which is where the journalist makes Bedolla's quote lead us to. But it is a non-zero set of people. I know some, and it puts teachers in a difficult position. Most teachers eventually figure out that they need to unofficially provide what services they can, even though it's a pain and time consuming, and often the kids are failing or nearly failing by that point. (The same is true with some SpEd kids.)
The schools have to have a way of testing without regard to what the parents think. That gives the teachers a heads up. The kids still get no official services, at least in Texas. Some nanny-states go further than Texas and put the kids into the programs anyway, usually with good intentions (but we make a point of denying good intentions to people we don't like; they're intentionally and self-assertingly evil). They want to protect the kids and help the kids. It's a social justice and social inequality thing. Often the teachers and administrators that go into ELL programs have precisely that view--they have to save these kids and as experts they know best. Not just "better." "Best."
The home language surveys are great. But these can be problematic. In many cases you're asking non-native speakers to assess their kids' native-speaker competence. Even if the kids are colloquial-fluent, they're still not fluent in the variety of styles, registers and contexts that constitutes real age-appropriate fluency for school.
The CELDT (like the Telpas, in Texas) has problems. I know monolingual kids who wouldn't pass the Telpas because they come from sufficiently low-SES backgrounds that they, too, haven't been exposed to English in the variety of styles, registers, and contexts that constitutes real age-appropriate fluency for school. Too many kids say things like "There's this thing that does what it does to this other thing." That could be a screwdriver screwing in a screw, a shark eating a fish, a sperm fertilizing an egg, or somebody rowing a boat. Lexically deficient, contextually dependent. This is bad when they're in high school. It's how L2 speakers get around being lexically deficient by relying on context.
That many monolingual kids fail it when tested makes sense. The ones most likely to be sent are going to have some kind of "marker"--surname + skin color, accent, geography--that'll mark them for testing. And since they're likely (statistically) to be low SES, they're (statistically) likely to fail.
I know in Texas the ELL label sticks with you until you graduate high school. However, it's federal law to phase you out of the ELL program or show evidence that you still require services. Once you're out of the program, you really get no services. And the label is utterly meaningless for most purposes. It does figure in for federal reporting purposes, though--part of "closing the achievement gap" involves tracking and looking at the data for ELLs. (This, too, is a form of nanny-statism, with a view to social justice and liquidating social inequality. ELLs, current and past, statistically are likely to struggle. Counterexamples are nice, but are counterexamples that test the generalization's applicability without seriously damaging the generalization's validity. In other words, they're exceptions that prove the rule.)
= new reply since forum marked as read
