Julian Vasquez Heilig, who recently moved from the University of Texas to California State University at Sacramento, is one of the nation’s leading authorities on Teach for America. He has studied their performance over time (see here and here), and he is not a fan. When Mathematica released its latest study of TFA, Heilig read it closely and analyzed the findings. TFA boasted that the study showed that its teachers were just as good as those who had studied education and intended to be career teachers. Some readers gleaned from this finding that “anyone can teach, no professional preparation needed,” that is, if they graduate from a highly selective college and are admitted to TFA.
Heilig digs deeper and has a different take on the study. The main finding, he says, is that Mathematica found no statistically significant differences in the groups of teachers they studied. However, he points out, the TFA teachers were overwhelmingly white, and few had any intention of staying in teaching as a career.
He notes that the test of “effectiveness” in pre-K-grade 2 is a five minute test:
Equally effective at what?…Mathematica utilized performance on the Woodcock Johnson III for the Pre-K-2 results— which takes 5 minutes to administer. Thus, the effectiveness of TFA teachers compared to Pre-K – 2nd grade teachers is based on a five minute administration to capture letter-word identification (Pre-K – 2) and applied problems for mathematics Pre-K – 2). Furthermore, one of the more egregious issues in the study is the aggregation of grades is that of the states that have Pre-K programs, more than half of states do not even require Pre-K teachers to have a bachelor’s degree. The report does not state that lack of a degree was an exclusion criteria and it is explicit that community preschools were included, so it appears than an aggregate that includes not only alternatively certified but also non-degreed teachers worked to TFA’s advantage. Should we really be impressed that TFA teachers outperformed a group that could have included non-degreed teachers? And they do it twice: with kindergarten and with grades K, 1, and 2.
What are the lessons of the study? Heilig writes:
So the [TFA] teachers were— on average— young, White, and from selective colleges. They had not studied early childhood in college and had very little teaching experience. They reported a similar amount of “pedagogy” (primarily the 60 hours from the five week Summer Institute), and more professional development (as we discussed above, they viewed it not very valuable). TFA teachers also reported less student teaching experience before they entered the classroom. They also were more likely to be working with a formal mentor (I mentioned David Greene’s point about the drain on mentors due to the constant carousel of Teach For America teachers in and out of schools here). As new teachers, they spent more time planning their own lessons, but were less likely to to help other teachers. Finally, TFA teachers were less satisfied “with many aspects of teaching” and less likely to “plan to spend the rest of the career as a classroom teacher….”
In conclusion, read at face value, here is the message Mathematica appears to promulgate with the report:
We do not need experienced (read: more expensive) teachers when non-experienced, less expensive teachers get the “same” —though not statistically significant— outcomes.
We do not need a more diverse workforce of teachers, again, because TFA teachers, who are overwhelmingly white, get the same outcomes.
Is TFA really in alignment with a vision for providing every student a high quality teacher? Or do they, Mathematica et al. just keep telling us that they are?
For myself, I have read many times that Teach for America invites young people to “make history” by serving for two years. And Wendy Kopp has frequently said that “One day,” all children in America will have an excellent teacher. I have a hard time understanding the logic of these claims. If the TFA teachers get the same results as current teachers, how is that “making history”? If most TFA recruits leave after two years, how does that lead to the conclusion that one day all children will have an excellent teacher? If TFA persuades policymakers that teachers can do a good enough job with no professional preparation, doesn’t that decimate the idea of teaching as a profession? If anyone can teach so long as they went to a selective college, how does that raise the standard for teachers? If our policymakers prefer churn, with teachers leaving every two or three years to find their real career, how is that good for students? How does TFA improve the profession? It doesn’t. It eliminates it.
For his fearlessness, for his willingness to stand up to those with money and power, for his willingness to present the evidence as he finds it without fear or favor, I place Julian Vasquez Heilig on the honor roll of this blog. He is an example to all researchers of the ethics of his profession. To be an outstanding researcher requires years of study, scholarship, discipline, dedication, and experience. Sort of like being a great teacher.
Thanks Julian for sharing . We need to share this one all all or teacher friends . TFA does not work .
Just a note. My sense of the Woodcock Johnson is that it is not reflective of the kind of work we require in the classroom today, given the curriculum under the common core. When we have had students assessed using that instrument, they generally do well, even when they have difficulty, often with many supports, in the classroom. Let me know if I am incorrect, but might that also be a reason why there was little or no difference found between TFA and regular teachers (since the skill of teaching is more apparent when remediation or other strategies are required to help a struggling student reach a higher standard)?
Just a note–there is no “University of California, Sacramento”. Dr. Heilig is at the California State University, Sacramento–an entirely different university system.
Beverly, thanks for catching my error. I fixed it!
Thank you!
The WJ2 reading and math achievement subsection scores are based on a very small number of items that require small, deskilled rote learning or incidental learning from home and community. So we are not testing much when we say the child scored X on the WJ2. A bigger issue is that the confidence band on the subtests is huge, thus making any use of the raw scores, grade equivalents, or the percentiles meaningless. What a silly way to try to figure out if a teacher is “effective” or “adds value”. How bankrupt this study is; and kudos to Julian for his public critique.
The measure of effectiveness is, as always, the Achilles heel in these studies. Mathematica uses standardized tests. Tests only measure what is on the test, nothing more. To blindly assume causation is a major mistake. Shame on Mathematica, your bias is showing. Plus TFA teachers, like all new teachers, depend heavily on veteran educators who must take time out of their day to mentor and support the newbies for the good of the overall profession and long term success of schools. Thank you, my great mentors for your patience in my early career (and for not laughing).
This silliness insisting collaboration, practice, and training mean nothing shows how statisticians can focus on flawed models while diving down a rabbit hole. If that is true, let’s just abolish education and everyone can just do whatever they feel like tomorrow – no experience necessary. Wait, that might be exactly what The Reformers want….
TFA=Teach For All funded by jet set and billionaire corps who hire a swarm of young, innocent, naïve college grads with a lower pay.
Kind of philanthropy business welfare they are doing to feed temps for 2-3 years, like JET settlers (who are also hired as ALTs in a non-English speaking country)
I wonder how Lakeside, Sidwell, or Chicago Lab School parents would take it if their young’uns had a string of two-and-done TFAers getting them ready for Princeton or Stanford at 40K per year?
They might not be as concerned by it as you think, assuming that they could exhibit a strong knowledge of content in the area where they were being hired to teach. In fact, 20% of the faculty at the Lab School, 21% at Lakeside, and almost a third at Sidwell Friends do not possess an advanced degree of any kind. Many of these are indeed younger teachers who have not decided on whether they will pursue teaching as a career.
Are they TFAers with just 5 weeks of training? Are they TFAers who have no doubt they will be out in two years max? And if your stats are good, 8 out of ten at elite private schools are committed. well trained teachers. Doesn’t sound like the parents are clamoring for TFAers so they can cut down on those pricey tuitions. My guess is that these well educated parents with extremely high expectations for their kids fully appreciate the value of teaching commitment and experience.
As anyone who has administered the Woodcock Johnson III (I had administered it hundreds of time), and based on Buros’ Mental Measurements Yearbook, this test is highly inflated at the early grade level–especially kindergarten. If a student can get one or two letter recognition items correct and has one/one correspondence to five, the test puts them on grade level with standard scores over 100. If at that age, they can recognize one or two words, then the test puts them close to the second grade level. Is this the curriculum? No. It is especially not the infamous Common Core curriculum. Therefore, conclusions are drawn using an invalid and unreliable test that was standardized back in the late 1980s.
It seems very odd to use a five minute test to differentiate between the teaching ability of their teachers. We used to use the WJ to identify potential weaknesses as part of the diagnostic testing for IEP reviews.
You are 100% correct. I was an Educational Evaluator for a Committee on Special Education child study team in NYC for 19 years. The WJ III is a diagnostic normed-referenced assessment that is supposed to be used to determine the strengths and weaknesses of LD students. It is NOT supposed to be used to determine progress and growth to determine teacher effectiveness. You only use it to measure the progress of a child for either an annual review or a mandated three year evaluation to determine special educational services. What these folks are doing is a MISUSE of the test. Think of it. They are drawing conclusions about teacher effectiveness based on the inappropriate use of an assessment instrument standardized in the late 1980s. Obviously, it does not even come close to measuring today’s curriculum. But guess what, it is not supposed to.
I am planning to write these researchers and blast them for using such a test. We must spread all over the Internet information about their flawed research. Isn’t it obvious why they would use an inflated test? They wanted their numbers to look better. This makes me so angry.
TFA=Totally Freaking Arrogant.