(Guest Post by Matthew Ladner)
…but other than that, it’s swell!
Let’s start with ineffective. Super chart! below from the Brookings Institute, you will recall, shows average gain scores in mathematics for three different groups of teachers: traditionally certified, uncertified and alternatively certified:
The differences between the three groups are far, far, far, far smaller than differences within each of the three groups. Schools keeping out the uncertified and alternative certified teachers on the right side of the curve are doing a grave diservice to their students.
Next, let’s look at evidence from Paul Peterson and Daniel Nadler’s recent Education Next piece on certification.
There is some suggestive evidence that alternative certification programs help state teaching forces look more like the ethnic composition of their states. Another way to put this might be to say that requiring people to take 30 hours of course work for no apparent reason (see Super Chart! above) may have a disproportionate impact on minority students. Or, put another way, genuine alternative certification seems to provide more opportunity for minorities to enter the teaching profession.
Now “racist” is a tricky term. Some argue that the SAT exam is racist, as it has a disproportionate impact on minorities. From what I understand, the SAT does successfully predict college success to a large degree, while teacher certfication does not predict student gains (see Super Chart! above). If so, by my way of thinking, the SAT is not racist, while teacher certification may be, de facto.
Let’s put it another way: if SAT scores don’t predict college success any better than certfication predicts successful teaching, I’ll happily join the chorus calling to eliminate the exam.
Finally, Peterson and Nadler show that the 21 states who have done more than symbolic alternative teacher certfication have made larger than average gains on NAEP.
This of course does not prove that alternative certification caused the faster gains, but they certainly didn’t prevent these gains. Florida, a leader in alternative certfication, has about half of their new teachers coming from alternative routes. As you can see, it doesn’t seem to be hurting their academic achievement.
As the figure below shows, Florida’s free and reduced lunch eligible students now outperform the statewide average in my home state of Arizona.
As one of Jay’s neighbors once said, it’s time for a change in how we train, recruit and compensate teachers.
Jay P. Greene's Blog 
So the SATs probably aren’t anymore racist than the successfully completed educations they predict. So if you think the test is racist you might want to look at institutions that don’t use the SAT as an acceptance criteria.
What I was trying to get at is that a screen with an disproportionate racial impact is especially suspect if it serves no valid purpose. If the SAT serves to reduce the number of students washing out of colleges and/or helps sort students into institutions which are better suited to their level of preparation, well, we can debate that. If teacher certification fails to improve student learning but keeps a large number of highly effective teachers out of the classroom, then it should hit the ash heap of history.
I get what you’re doing with this comparison to the SATs, Matt, but I do think there’s an important difference worth considering. In the case of teacher certification, we have a strong external validity test – although many deny it, performance on basic skills tests is a good, objective indicator of the success of education in basic skills. However, in the case of college, the external validity test is less strong. Grades in college classes aren’t very meaningful. The college dropout rate is a better indicator, but even that isn’t as strong as basic skills tests (for one thing, it’s binary and thus provides less precision).
But your main point, I take it, is simply this: If we conclude that teacher certification requirements are de facto racist, this does not imply that the SAT is also racist. And on that you’re clearly right.
The question of whether anything can be “de facto” racist is also worth noting. In my view, the term “racist” ought to be reserved for 1) racist states of mind, whether conscious or unconscious, and 2) actions accompanied by such states. You ought to be able to show what the lawyers call mens rea before laying down the R word.
Greg-
So if I follow your first paragraph, one might infer that the SAT is a better predictor of teaching effectiveness than certification?
Some foreign education systems may have figured out ways to predict teaching effectiveness, but from what I understand, our ability to do it in the United States is very limited. The logical thing to do is to monitor progress after teachers get in the classroom, reward the good ones, remediation for the mediocre, and let go the ones on the far left end of the bell curve.
I’ll have to think about your point on racism. I guess I do believe in “accidental racism” but I’ll give it some thought.
My point was meant as an aside. I probably should have used “one” in the second sentence instead of “you”.
Well, I don’t know whether you can infer that the SAT is a better predictor of teaching effectiveness than certification from what I wrote. But it’s true, whether you can infer it from what I wrote or not. In fact, studies have found strong links between a variety of undergraduate-level acadmeic measures (SAT scores, selectiveness of the college, etc.) and teacher effectiveness.
On the subject of racism, I’ll go this far with you: I’ll admit that an abnormally strong indifference to racial injustice could be a form of racism without actually involving a mental state as concrete as the mens rea I called for in my last comment. Perhaps we could extend my legal metaphor and distinguish between “first degree racism,” that is, racism with both mens rea and malice aforethought (i.e. conscious malice); “second degree racism,” that is, racism with mens rea but without malice aforethought (i.e. unconscious racism); and “raceslaughter,” that is, an abnormally strong indifference to racial injustice without the mens rea of concrete racism.
The idea that traditional certification is largely ineffective, and likely counterproductive, is certainly not new, but data to show that is always welcome. I would hesitate to use the term “racist”, as that can easily be taken to include racist intent. But “differential impact” does not share that drawback. It seems to me that common sense, if not the data, would support the idea that traditional certification requirements would have a differential impact on minorities wanting to become teachers.
Traditional certification requirements ought to be severely curtailed or eliminated. We’ve known that for decades. The situation has hardly changed since I went through the teacher training curriculum at the University of Missouri in the early sixties. But where would change come from? Can a politician seeking state office gain any votes by attacking traditional certification requirements? In my experience I can’t recall anyone ever trying that, though it would certainly attract my vote. I don’t know much about the politics of teacher certification, but I would guess that there is an “education establishment” with some political clout and savvy on one side, and a great mass of apathy on the other. John Q. Public, I suppose, passively accepts the general principle that occupational licensing laws are for the benefit of the public. End of story.
So your thoughts and data are certainly welcome, and I hope beneficial, but where do we go from here to affect needed change? I haven’t the slightest idea.
Mr. Rude-
I think the demographic changes on the way will strongly encourage states to follow the lead of states like Florida, and other changes as well.
Oh and also- I agree with you that there was no racist intent- which is part of the reason I described teacher certification as “possibly racist.” No one was out to stick it to African American and Hispanic children when they created teacher certification.
Apparently, from what we can tell however, they did in fact do so. Further, those who continue to oppose reform despite this growing body of evidence…
[…] Ladner presents Teacher Certification: Ineffective, Counterproductive and Possibly Racist… « Jay P. Gree… posted at Jay P. Greene’s […]
One big problem is that we cannot reliably MEASURE teaching effectiveness, much less predict it.
I’m not sure whether teacher certification is effective, but the data you present, although suggestive, does not prove it is ineffective, unless there is no significant difference among the students; an alternative explanation (which, in an ideal world, would be true) is that the certified teachers get students that are harder to teach, and so may be more effective and still not be able to improve the students’ rank.
The way I see it, the problem with measuring effectiveness is that we would need to measure three things:
– Knowledge of students
– The rate at which they can learn (sorta kinda IQ, but also related to current knowledge and environment)
– The rate at which they can become smarter (I don’t see ‘IQ’ as fixed)
Knowledge is (relatively) easy to measure, rate to learn is harder (IQ kinda works, but …), rate to become smarter is much harder (do we have any measures ?)
Unless we can measure those things (and the effects of the kids environment),comparisons are highly suspect.
[…] for greater accountability? The effectiveness of teacher certification on student learning is very much in doubt, to put it kindly. Let’s hope the DPS board is willing to take such a small chance to help […]
You have to also consider the demographics of students. Some students come from literate families while others do not. This greatly affects learning potential. Some are educated prior to coming to the United States, others are not. Some languages are pre-literate, and students speaking these languages usually take longer to learn English. Wisconsin requires English Language Learners to take state assessments, (I assume Florida does, too) however the assessments are accompanied by a translation document with a dialect that few can understand, that even aides can’t translate and that students can’t read, rendering the test unreliable.
You have some good points, but as with most things, you can’t only look at test scores. We need qualitative assessments, too.
The alternative certification may have something to do with the gains, but I doubt it is the whole reason for them.
The point I was trying to make above is that there are factors in test-taking that can’t be controlled for, which I’m sure you’re aware of.