Follow me on Twitter at @AnthonyCody
A week ago I posted an interview with Teach For America’s head of research, Heather Harding. Ms. Harding answered some tough questions that have been raised in recent months here on this blog. Today, I am sharing some responses to her answers.
By way of context, I have come to believe that addressing teacher turnover is one of the linchpins of real reform in our struggling schools. Turnover is a key indicator of unhealthy working conditions for teachers -- and that tells us conditions for learning are poor as well. Programs such as Teach For America allow school districts to ignore these poor conditions, by providing a steady supply of novice teachers. Unfortunately, these novices turn over at a very high rate, and the schools must invest a lot of resources in their training -- which is lost when they leave.
There are a number of facts in dispute regarding Teach For America, so we need to look closely at the evidence in order to make sensible conclusions. Here are some of the questions Ms. Harding answered where the facts are in question, followed by responses from myself, and several readers with some expertise in this domain.
Question 2: Given that many Corps Members report feeling underprepared, is there any interest in expanding the training, or shifting to a more in depth preparation model, like the residency program some TFA alumni have suggested?
Heather Harding: New teachers, whether part of Teach For America or from other pathways, have one of the toughest jobs imaginable, particularly those working in high-needs schools. It is unlikely that any program fully prepares teachers for every challenge they'll encounter during their first year as teacher of record. However, we are committed to getting our teachers as prepared as possible, and then providing them with the ongoing support they need once they start teaching. In my own work as a teacher educator, both with Teach For America and in a more traditional program, my biggest goal was to support a candidates ability to self-diagnose and access to knowledge networks and ongoing feedback for continued development. We do this in a variety of ways.
We're encouraged by the fact that principals who employ our corps members overwhelming (87 percent) think that Teach For America's training is at least as effective as the training other beginning teachers get. Additionally, three recent studies conducted in Louisiana, Tennessee and North Carolina have shown that corps members often help their students achieve academic gains at rates equal to or larger than those for students of more veteran teachers.
I wrote a more in-depth piece on why Teach For America is not going to become a residency last summer for Rick Hess' Straight-Up blog. While the residency model provides a longer ramp for practice and may offer stronger alignment with local concerns such as curriculum, we don't yet have any evidence that the result is a more effective teacher.
Response from Anthony Cody:
With respect to residency programs, teacher effectiveness, and teacher retention: A recent evaluation of the Boston Residency Program found that resident teachers became more effective than other beginners over their first several years in the classroom. Furthermore, evaluations of the longest-standing residency programs in Boston, Chicago, and Denver have identified very strong retention rates for their graduates -- exceeding 80 percent after 4 or 5 years in the classroom. This compares to published retention rates for TFA graduates in New York and Houston of only 10 to 15 percent by year 4. These and other studies have found that the positive effect of a teacher with three or more years of experience is much greater than the effect of any entry program on student learning. Thus, programs that keep teachers in the profession have long-term effects on student achievement.
With respect to the effectiveness of TFA teachers, the New York and Houston studies cited above found that first and second year TFA recruits were less effective than those of fully trained and certified beginning teachers in both math and reading at the elementary level. The small number of TFA recruits who stayed in teaching after the third year became more effective than many other teachers in math (on some tests though not others), but did not fully recoup their lost ground in reading and were particularly ineffective for Spanish-speaking students in both reading and math (examined in the Houston study). A more recent Texas study also found that TFA recruits were less effective than other new teachers in teaching reading to Hispanic students.
This should not be surprising, given that the TFA 5-week entry model allows little time for learning to teach reading or to acquire the sophisticated techniques needed to teach English learners. The recent Texas-wide study also found that TFA recruits had left at rates more than twice those of other beginners by year three (TFA showed a 44-59% attrition rate as compared to 19 - 24% for other new teachers from other programs).
Studies that have found more positive influences of TFA teachers on student achievement have generally compared them to less well-qualified teachers, either by making comparisons to other teachers in the same under-resourced schools (who are less likely to be certified, experienced, or otherwise well-qualified than teachers in other schools), by not taking teacher certification into account, or by comparing TFA teachers to others after they have completed certification (generally by year three) rather than in their initial years. In states that have higher standards for most teachers, TFA teachers compare less well to others than in states that have lower standards for teachers generally, or for teachers allowed to teach in high-need schools.
With respect to the satisfaction of principals in schools that hire TFA teachers, such survey results come from a biased sample. Principals who have had poor experiences with TFA recruits do not continue to hire them, if they have decision-making control over staffing, and hence are not part of these surveys. Studies have confirmed that schools that hire TFA teachers are poorly staffed to begin with, so the principals of these schools are comparing TFA teachers to others who are often also inexperienced and underprepared. This does not mean that principals of schools with stronger staffs - who often refuse to hire TFA teachers - would compare them favorably to better-trained teachers.
Barbara Torre Veltri adds: I disagree on the evidence of residency not promoting more effective teachers. Look at the models of private day schools who insist upon their intern teachers completing two years as ‘assistant teachers’ working under a head teacher to see their retention as teachers and their long term effectiveness as well as the Reggio Emila approach to teacher interns learning the range of child development while immersed in a classroom working with credentialed Reggio teachers. I note this information as 12% to 19% of current corps are teaching Early Childhood Pre-K in several regions.
Jason O’Brien adds: Ms. Harding wrote “However, we are committed to getting our teachers as prepared as possible, and then providing them with the ongoing support they need once they start teaching.” Ms. Harding and I will have to agree to disagree on the notion of “as prepared as possible.” A “one size fits all” five week summer training institute with limited interactions with students is not preparing TFA members “as much as possible.” Anecdotal evidence suggests that most corps members have no idea as to which grade level they will be teaching until after their summer training. Would TFA be willing to honestly share the percentage of corps members who know what grade level or subject they will be teaching before they undergo summer training? Special populations such as Special Education students and English Language Learners are particularly vulnerable and TFA’s summer institute’s focus on “leadership” rather than differentiated instruction seems particularly troubling.
One of the most important aspects in traditional teacher education programs is the 16 week student teaching placement. These unpaid internships both provide real world experience which can’t be gleaned in a college classroom setting and also act as a safeguard to insure that novice teachers are ready for their own classrooms. In these placements, student teachers first observe then spend time teaching lessons under the guidance and watchful eyes of trained and certified teachers and university supervisors (who by definition must have at least 3 years of successful teaching experience in classrooms). TFA corps members spend the first several weeks “learning to teach” their students. Given that TFA members are often placed in Title I classrooms in which students are already demonstrably behind their peers, this is the single biggest flaw in TFA’s paradigm. Furthermore, the “ongoing support” is often provided by other TFA members who on average have between two and four years of classroom experience. In effect, there are undertrained and inexperienced TFA members being trained by other “experts” who have less experience than mentors from traditional teacher education programs.
Question 3: Given the strong evidence that staff stability has large effects on student achievement, how can TFA justify having teachers cycle in and out of teaching in the neediest schools, with most gone after 2 years and 80-90% gone by their 4th year? Doesn't this do a disservice to the students who most need a stable supply of expert teachers?
Heather Harding: District leaders and principals regularly comment on their satisfaction with our teachers. So again, the majority (85 percent) of principals who hire Teach For America teachers report that corps members have made a positive impact in their schools. Even more telling is the 89% of principals who indicate that corps members have become a part of the community.
It's a myth that Teach For America corps members have a significantly lower retention rate than other teachers. According to our internal statistics, in 2010, 88.6 percent of first-year corps members returned to teach a second year. In contrast, about 83 percent of new teachers in low-income communities (NCTAF No Dream Denied report, 2003). In the last five years, the percent of corps members returning for a second year has remained above 88% with some years as high as 92.4%. In the only external study done on Teach For America teachers' retention, Donaldson & Johnson (2010) report retention rates for three consecutive cohorts: 61% stay longer than two years, 44% persist into a third year, and 35% into a fourth year. These numbers are not much different than national trends that find nearly 50% of all new teachers leave the profession by year six.
Response by Anthony Cody: The Donaldson and Johnson study is not the only external study -- it is based on a survey with a relatively low response rate, also biased by the fact that those who have left the profession are more likely to be non-respondents. Based on actual record data from districts and states (cited above), TFA teachers leave teaching at rates of 80 to 90 percent by year 4, and other beginning teachers leave at rates of about 25 to 30% in that same time period.
The statistic of 50% of all new teachers leaving the profession is not accurate, as those studies show. National studies (the Schools and Staffing Surveys and the Baccalaureate and Beyond data base) find that for public school teachers, about 30% leave in the first five years. However, within that 30% there is a big difference among those who enter with full training and those who enter (like TFA) without full training. For example, Baccalaureate and Beyond data found that only 14% of fully trained teachers had left by year 5 as compared to 49% of those entering without training. Schools and Staffing followup study data found that teachers who entered without student teaching and key coursework (on child development, learning, curriculum, etc.) left at more than twice the rates as those who had this training.
This, to me, is the most critical issue. I worked alongside TFA corps members much of my career, and spent the past four years leading a program devoted to providing new science teachers with experienced mentors, in the hopes that we might retain more of them. Our experience in Oakland was that three years after they begin, 75% of TFA corps members have left. (The same holds true for interns from a program affiliated with the New Teacher Project.) A report prepared by the School Redesign Network at Stanford found the following:
The low productivity associated with first- and second-year teachers is of particular concern because their presence in Oakland Unified School District (OUSD) has increased consistently over the past few years... The proportion of novice teachers grew from 10% in 2004 to 24% by 2008, more than doubling in proportion over this time, with two-thirds of this group being in their first year.
Furthermore, ... schools with large concentrations of African American and Latino students were especially likely to have the greatest proportion of first- and second- year teachers, as are schools performing in the lowest tiers of OUSD's tiering system. For example, schools with 60% to 80% African American and/or Latino students had an average of 21% of their teachers in either their first or second year. Schools with 20% or fewer African American and/or Latino students had an average of only 5% of their teachers in their first or second year.
What is more, one of the studies cited by Ms. Harding bears this pattern out. The Louisiana study, found that:
Across samples, new TFA teachers' 5-year persistence in teaching ranged from 4% to 20%. In comparison, the 5-year persistence of new teachers with standard teaching certificates for these cohorts ranged from 62% to 65%.
This clearly creates substantial problems in an urban district like Oakland. It is expensive to train these new teachers, and the investment is lost when they depart. The inexperienced teachers are less effective, especially their first year, and if the turnover is high, a significant portion of them will be first year teachers. As this research shows, student achievement suffers when staff turnover occurs. These schools need stability, as do the students. It is difficult to establish and sustain a solid, positive school culture when turnover is this high. As a result, Oakland is now making a substantial shift in its hiring practices, away from programs such as Teach For America.
Jason O’Brien adds: One of the main criticisms of TFA is the lack of peer-reviewed research it provides to back up claims. This number is entirely misleading because we don’t know who these principals are who are making these claims. If TFA is allowed to self-select these principals, then this rate is going to be artificially high. This is why transparent, and rigorous research (which is peer-reviewed) is necessary. There are many researchers (including myself) who would love to conduct mixed methods research comparing the performance of TFA corps members with traditionally-trained teachers, but unfortunately TFA has been reticent to allow this type of research.
Ms. Harding wrote: “Principals and independent data report that our corps members are effective classroom teachers.” This statement is meaningless unless it is reported as research which was conducted under the auspices of IRB approval (to ensure anonymity and accuracy) and is vetted by researchers in the field. For instance, a 2010 study released by Project MATCH indicates that TFA teachers had an aggregate score of 32/100 on the measure used in their study (report available here). Guess which teachers did the best in this study conducted by Project MATCH? You guessed it, Project MATCH teachers. This is a non-peer reviewed research “brief” of the sort that TFA likes to post in support of their “success.” The point is that when individual “reports” are published, if they’re not rigorous and unbiased, the claims contained therein are highly suspect. This is why peer-reviewed research is paramount when making any claims in regards to what does and what does not work in schools.TFA has not reached out to allow those in the academic community to gather and report data on corps members’ performance.
Barbara Torre Veltri: This principal survey appears to be the Kane, Parsons data whose initial study was conducted some time ago 2004 and then was reexamined. This might not be as valid at this time when many accolades reported noted that the quote was attributed to ‘a former principal” or one that is a ‘former corps member” who serves as a current principal in a school that hires significant numbers of TFA teachers, such as the numbers of TFA who are teaching and leading the schools in the Recovery District of New Orleans, the KIPP Charter schools (who employ and recruit TFA alums into their programs), or the “2/3 of corps members who are in leadership positions in Oklahoma City,” a region that as recently as 2010 hired TFA.
This suggestion that TFA’s training is as effective as the training other beginning teachers get is misleading. Can a study be funded to check on the data here?
Currently, data from corps members across regions (from Memphis to Kansas City to Alabama and New York) report concerns with classroom management. How can TFA’s effectiveness be measured as successful, without application of a range of workable and realistic management plans? For the record, we are not advocating generic management applicable to corps who are placed in Special Education classrooms with very little preparation.
Question 4: Many urban districts that have concerns about the high turnover and frequently poor performance of new TFA members are nonetheless required to hire TFA recruits by external funders (supporters of TFA) who make this hiring a condition of their grants. Why are districts pressured into accepting TFA recruits when they don't want them? And why does TFA participate in this practice?
Heather Harding: The statements in this question are just not based in fact. We have 9,000 teachers in 34 states and more than 3,000 urban and rural schools across the country, and we are not aware of any teachers who were forced upon a school or principal because of a donor or grant. The decision to hire Teach For America corps members is made by school districts and individual principals, alone. Our corps members apply for open positions alongside other candidates, and principals determine who the best fit for their schools is. There is no one-size-fits-all solution for hiring; Teach For America is one of several options school districts are relying on for finding and hiring effective teachers.
Principals and independent data report that our corps members are effective classroom teachers. Ninety-two percent of principals who work with Teach For America teachers agree that our corps members are just as effective, if not more so, than other new teachers in overall performance. As I mentioned above, three recent studies conducted in Louisiana, Tennessee and North Carolina have shown that Teach For America is among the most effective sources of new teachers.
Current TFA Corps member Jameson Brewer responds:
The answer provided by TFA shows either a direct attempt to distort the facts or it sheds light into a breakdown of communication between TFA and the districts that they “serve.” As a traditionally trained educator who graduated at the beginning of the recession, I was only afforded two interviews over a two year period of time. The job prospect that was most promising ultimately led to a denial because my only teaching experience was student teaching. The principal felt inclined to hire someone else given their experience and his “dedication to hire seasoned teachers given the type of school (Title I).” However, the following year the principal hired approximately 7 corps members. Either this shows a change in his attitude of how inexperienced teachers would perform or it is the evidence of an invisible hand at work.
Out of desperation to secure a teaching job, I applied to TFA. Given that I acknowledged that I would be willing to teach special education on my TFA application, that was the position I was hired to do despite my degree and full certification to teach history. I petitioned TFA for countless weeks to switch my assignment to one in which I was “highly qualified.” During the course of my Institute “training,” I was offered a history position at a middle school outside of my affiliation with TFA. When I notified TFA of this offer, I was told, within that day, that there had been a history position opening and they were switching my assignment. I stayed in TFA.
The first day that I spoke with or met my new principal was the first day of pre-planning. I never interviewed for the position and he did not know my name. Certainly this is evidence of TFA placement by force. Prior to my arrival, there were four history teachers. I now know that the one I replaced was “leveled” the day before my arrival. That is, he was let go due to low student enrollment. After my first year, there was to be another “leveling” of staff. I was told by my principal that I should not worry as my position was protected by my TFA status. Another good teacher was let go. Further, corps members are told during Induction and Institute that the number of corps members each year is reflective of the “anticipated positions scheduled for TFA teachers.” When I ask my principal and other teachers at my school their thoughts on TFA teachers I’m told that, “it’s hit or miss, we never know who we are going to get.”
Jason O’Brien: In the Huntsville City School district, Ms. Harding’s statement is simply not true. The superintendent has contracted with TFA and has set aside a certain number of positions which will be filled with TFA corps members. Furthermore, in our “managed” (i.e., low SES) schools, interviews will not take place at schools with principals, but rather will require all new applicants to interview with the central office. In other systems, TFA members do not have to undergo an interview (Veltri, personal communication, 2012) and therefore have teaching positions held for them. Furthermore, because the TFA initiative has come from the superintendent’s office, how likely is it that individual principals in any district will publicly criticize these policies? Again, this highlights the need for rigorous research which guarantees the anonymity of the responses and open access to aggregate test scores.
Update, May 11, 2012: This Memorandum of Understanding between TFA and the Oakland Unified School District has just come to light. It documents the fact that 43 TFA corps members were reserved teaching positions for the 2006-2007 school year, in this agreement signed in April of 2006. The school district paid $1500 per TFA corps member, using Title II funding. I believe this was the practice for several years, but has recently been discontinued.
Question 7: Why does TFA not require corps to remain in their commitment for 3-5 years when they move beyond the novitiate state and advanced beginner stages of teaching?
Heather Harding: We've found that two-years is the right time commitment for motivating people's entry into this work. Given that 61% of our corps members teach beyond that initial commitment, and many more continue in the education sector for their professional career, we believe that most corps members retain what will be a lifetime commitment to improving education for kids growing up in poverty. Additionally, according to our annual alumni survey, 33.3 percent of all of our alumni (more than 20,000), are working full-time as preschool-12 classroom teachers today, even though less than 14 percent had considered a career in teaching when they first entered the corps.
In a recent study conducted by Harvard economist Roland Fryer, participation in Teach For America markedly affects corps members' beliefs. Fryer and co-author Will Dobbie indicate that the Teach For America experience strengthens participants' conviction in the academic potential of all children regardless of income level or race, and increases racial tolerance among participants across all racial groups. In addition, the experience increases the likelihood that participants will pursue a career in the education sector.
Anyone who has taught knows that teachers alone can't build the system all kids deserve. We believe it's critical that some of our corps members stay in the classroom beyond their two-year commitment, while others take on other leadership roles inside and outside of education. Teachers - and students - need and deserve committed principals, mentors, social workers, district leaders, public health workers, politicians and others working to provide every kid access to a great education.
Jason O’Brien: TFA’s focus on corps members leaving the classroom for “leadership” positions is particularly troubling for teacher educators who expect their students to spend their careers in the classroom honing the craft of teaching through content and pedagogic knowledge. Continuation in the “education sector” oftentimes translates into positions of power (administrative positions, lobbyist positions, working as legislative assistants). These TFA corps members then use these positions of power to advocate for more public funding for TFA and for the expansion of TFA into districts which were never intended to have TFA corps members. The perception from the outside is that TFA members are encouraged to take positions which will ensure its continued existence and continued amassing of wealth and resources.
Gary Rubinstein:
How can only something like 10% of TFAers stay for a fifth year and yet 33% become, as this number suggests, career teachers? I think the 33% of TFA alum are still teachers is very skewed since it doesn’t take into account the exponential growth of the size of the corps.
So if from 1990 to 2007 there were 20,000, while only about ten percent (2,000) are still teachers and then in 2008, alone, there are 3,500 alumni with about 25% (or about 1000) still teaching in their fourth (and generally, final, year), and then in 2009, there were about 6,000 teachers with 60% (or about 3000) still teaching in their third year. So, in total there are about 6,000 alumni still teaching, which might be close to 33% of all alumni, if you took the average for each cohort and weighted them equally, it would be much closer to 10%.
Incidentally, that 10% could be a number that they should be proud of since most of those alumni were not considering teaching as a career. But to inflate the number is not necessary or, in my opinion, honest.
Readers, what do you think of the debate here? How does this relate to your experiences?
Respondents:
Barbara Torre Veltri, Ed. D. is ssistant professor in the College of Education at Northern Arizona University, and the author of Learning on Other People’s Kids: Becoming a Teach For America Teacher (Information Age, 2010.)
Jameson Brewer is a traditionally trained educator (B.S.Ed. from Valdosta State University) who struggled to find a job teaching due to the recession. He is now a 2010 Metro-Atlanta corps member teaching high school social studies in the Atlanta Public Schools. He wrote about his experience with TFA here.
Dr. Jason O’Brien is an assistant professor of education at the University of Alabama, Huntsville.
Gary Rubinstein is a former Teach For America corps member who now hosts the Teach For Us blog.