> My sensibilities are much too delicate to go near the NewsHour. Fer one
> thang, I'd have to plug the teevee in. I know I could perfectly well Google
> this bitch's ass my own dang self, but what I am really looking for is the
> LBO no-holds-barred, fully editorialized bullet points. If you can bear to
> type without popping a major brain artery, could you maybe offer highlights?
> Either that or some insightful comments on your journey from arm-chomping
> third-grader to the articulate and opinionated soul you are today. See, I
> have a bored-ass nephew now wending his way through Our Public Schools. He
> seems to have teachers wise to some of the psychodynamics of his household,
> but I am still worried I am gonna be having to take him for rabies shots
> over some episode like one of yours.
As a part-time researcher in a little corner of the education bureaucracy in Australia I've looked a little bit into comparisons between education systems. One place the US really falls down is in relying on untrained teachers. In most of the countries at the top of the education league tables that edu-bureaucrats like to look at (e.g. Finland, South Korea and Hong Kong in the OECD PISA rankings) teachers are paid well above the average wage and tend to have post-graduate educations. In many states in the US, by contrast, a high proportion of teachers are hardly trained to teach at all, especially in the proverbial 'inner city schools'. Rather than agonising over phonics, higher stakes testing, etc., better pay and training for teachers might be a good place to start.
Here's a briefing I wrote last year when Joel Klein was over in Australia spruiking the Teach for America program, which aims to put 'high academic achievers' in those 'inner-city schools' at a fraction of the cost by appealing to their sense of charity and forgoing proper teacher training. (Alas, we now have Teach for Australia.)
-----------------------------
The Mathematica report
Paul T. Decker, Daniel P. Mayer and Steven Glazerman [2004]: “The effects of Teach for America on students: findings from a national evaluation”, Mathematica Policy Research report, http://www.mathematica-mpr.com/publications/pdfs/teach.pdf
This is the most positive research finding for the effect of TFA teachers on student achievement, and therefore often cited by supporters of TFA, and likely to be drawn upon by those who want to import a TFA-like scheme. It was subjected to criticism by Darling-Hammond et al [2005], to which Mathematica responded. In my view, the Darling-Hammond criticism sticks, especially if the Mathematica research is misinterpreted as providing support for a TFA scheme outside the American context, in which the alternative to TFA teachers is often not fully-trained and qualified teachers, but completely untrained teachers. The authors themselves stress this:
This report addresses the issue directly by answering the question, Do TFA teachers improve (or at least not harm) student outcomes relative to what would have happened in their absence? Our approach to addressing this question is to compare the outcomes among students taught by TFA teachers with the outcomes of students taught by other teachers in the same schools and at the same grades, whom we refer to as ‘control teachers’. [p. xi-xii]
Still, a big positive in the Mathematica report is the data it collects on that American context. (The section on how the TFA scheme works is also worth reading as an overview.) In the schools in the study (“broadly representative of the schools where TFA placed teachers at the time of the evaluation” [p. 8]) only 55 per cent of the non-TFA teachers had at least a Bachelors degree in Education, and a third were completely uncertified to teach. Of novice non-TFA teachers (i.e. those with less than three years’ experience) – arguably more representative of the alternative to TFA teachers – only a third had at least a Bachelors degree in Education, and almost two-thirds were completely uncertified. More than half of novice non-TFA teachers had no student-teaching exposure whatsoever before being put in front of their own class.
The difference in educational and ethnic background (and presumably socio-economic background, though this is not addressed) between TFA and non-TFA teachers is striking. 70 per cent of TFA teachers graduated from “a most, highly, or very competitive college or university” according to Barron’s rankings, compared to only 2.4 per cent of non-TFA teachers (a single teacher in the sample). Two-thirds of the TFA teachers were non-Hispanic whites, compared to only 11 percent of the non-TFA teachers. That is, the TFA phenomenon is about mainly-white Ivy Leaguers doing a two-year Peace Corps-like stint in ‘inner-city schools’ before going on to their real careers. (Only 11 per cent of TFA teachers reported planning to remain in teaching long-term, compared to 60 per cent of non-TFA teachers.)
The researchers found that over a year, students with TFA teachers improved in maths by 0.15 of a standard deviation compared to students with non-TFA teachers. Compared to novice non-TFA teachers, the difference was larger, 0.26 of a standard deviation. In reading, there was no significant effect, positive or negative. These effects are not massive, but seemed to show that TFA teachers would at least not be any worse than the alternative. Note also that other research favourable to TFA has generally found the effect on maths learning to be smaller than Mathematica. (e.g., Kane, Rockoff and Staiger [2006])
What explains the difference between this Mathematica study, and the negative findings of Darling-Hammond et al [2005]? Mainly, that the Darling-Hammond et al study compared average TFA teachers not with other teachers in general, but compared uncertified TFA teachers with equally-experienced, equally-educated certified teachers. Darling-Hammond et al concluded that
Controlling for teacher experience, degrees, and student characteristics, uncertified TFA recruits are less effective that certified teachers, and perform about as well as other uncertified teachers. TFA recruits who become certified after 2 or 3 years do about as well as other certified teachers in supporting student achievement gains; however, nearly all of them leave within three years. Teachers’ effectiveness appears strongly related to the preparation they have received for teaching. [pp. 1-2]
The Mathematica study did not control in this way. This is reflected particularly in the data on the characteristics of TFA teachers in their study: a quarter have bachelors or masters degrees in education, and half are certified teachers (more than the proportion of novice non-TFA teachers). This is because they have simultaneously studied teaching and/or achieved certification while teaching. In the Darling-Hammond et al study, such teachers would cross the line and be considered regular certified and trained teachers rather than TFA teachers, since they would have completed regular teacher training.
Mathematica defended their research against the Darling-Hammond criticisms in a website statement on the grounds that their purpose was explicitly to compare TFA teachers with teachers the students would otherwise have had. This may be a fair point, but the Darling-Hammond research is more relevant to an Australian context where the teaching corps does not include a large proportion of unqualified teachers.