Does Testing Produce Better Teachers–or Just More Contracts for Consulting About Them?

Now that PARCC testing is well underway in DC public schools, test scores and their correlation (or not) with student achievement and teacher quality, hiring, and turnover seem to be in the air we breathe.

In January, in a report titled “Teacher Turnover, Teacher Quality, and Student Achievement in DCPS,” researchers found that some relatively high recent teacher turnover in DCPS seemed to have a positive effect on student test scores. Earlier research showed that similarly high teacher turnover led to a decrease in student test scores.

Then, in March, other researchers discovered (in a report titled “Teacher Applicant Hiring and Teacher Performance: Evidence from DC Public Schools”) that among the measures of new recruits that might predict teacher effectiveness (undergraduate GPA and interview scores, to name two), few are currently being used by DCPS when it hires teachers. Weirdly, the study noted (in a footnote) that few of those measures are associated with the “probability of retention within DCPS.”

(So, high teacher turnover begets yet more high teacher turnover?)

With the imprimatur of the National Bureau of Economic Research, which published both, these studies lend an aura of credibility to the politically and financially fraught enterprise of correlating student achievement with teachers.

But as working papers, neither study was peer-reviewed, although the first study received wide press coverage. Moreover, whether one believes in the tenet that teachers make all the difference in student achievement (and its logical corollary, that such achievement is accurately reflected in test scores), these studies underscore the close and well-financed relationship between DCPS and private foundations and organizations that support education reform.

For instance, one of the co-authors of the first paper, Melinda Adnot, had been a staffer for The New Teacher Project (TNTP), founded by former DCPS chancellor Michelle Rhee (and also the former employer of current DCPS chancellor Kaya Henderson).

One of the authors of the second paper, Benjamin Lindy, is a staffer at Teach for America (TFA) and a former DCPS staffer working on teacher selection. Another author, Rachel Rosen, works for an organization funded by foundations that lobby for charter schools, in the name of education reform.

All this may seem like water under the bridge: after all, who cares where great teachers come from as long as they teach our kids?

But “great” is in the eye (or wallet) of the beholder.

The first study noted that 55% of DCPS’s replacement teachers come from outside DCPS. Since 2009, that hiring has been overseen by central DCPS staff (previously, principals at individual schools found, interviewed, and hired new teachers, rather than having them first vetted by DCPS).

Teacher recruitment and training are big business for both TNTP and TFA. Both use a multi-week boot camp approach to preparing college graduates, most of whom have not had any teaching experience, coursework, or credits, to teach in some of the highest poverty schools in the country. The idea is that such a new approach will tackle old problems that plague many poor urban schools.

Critics have warned that this results in a teacher corps that is distinctly not diverse; that it degrades the profession by presuming that 5 weeks of training is sufficient to tackle difficult school circumstances; that it harms some of the most vulnerable students by handing them off to inexperienced teachers; and that the high turnover among these new recruits ensures that they never reach their full potential as teachers, all the while leaving their students and schools—most of them high poverty–in a lurch.

Regardless, the relationship between DCPS and these organizations is close. Former DCPS chancellor Michelle Rhee founded TNTP after serving in TFA, and current chancellor Kaya Henderson served at both as well. Current director of teacher recruitment at DCPS, Brooke Miller, was a TNTP staffer as well.

The DCPS relationship with both organizations is also lucrative:

Right now, DCPS has an FY16 $300,000 sole source contract with TNTP for teacher recruitment, selection, and training. When asked about other such contracts with TNTP and other education reform organizations that train teachers (TFA and Education Pioneers), Ms. Miller responded that DCPS does not release contracts unless they are required to be posted and offered that she did not think DCPS has a current contract with Education Pioneers.

(Apparently, the requirement for posting contracts publicly is that a contract is sole source or no bid. I found no way to access easily the text of all DCPS contracts, although this website provides a listing of recent ones, along with this website.)

The former website shows that DCPS has at least two additional contracts in FY16 with TNTP besides that sole source one for $300,000, but these contracts are not specified as anything other than “professional services.” The total of those two contracts is $130,000. (There is also an FY16 contract with TNTP for $50,000, but it has the same contract number as that sole source contract, so may be related.) Additionally, DCPS has a contract in FY16 with TFA, for $175,000 for “professional services.”

All told, TNTP holds at least $430,000 of DCPS business in FY16, and the totals I could find for both organizations’ FY16 work with DCPS amount to more than $600,000.

There were other, apparently current, sole source contracts with TNTP that I also found online here, but that did not have contract numbers, including one for measuring teacher effectiveness in comparison to DC charter schools and another for producing an annual “labor report” that would “highlight the results of the district’s human capital reforms, aimed specifically at an external audience.”

The latter contract, for $30,000 that ended in September 2015, noted that “DCPS has already compiled extensive data on the reforms for internal audiences. With nearly a decade of experience producing acclaimed reports about critical human capital issues, TNTP can turn this collection of data into a polished story that paints a compelling picture of the district’s accomplishments.”

Even if what David Grosso said in the Post yesterday is true—that the chancellor has “no sway” over contracts held by DCPS—these contracts nonetheless represent a lot of business for two organizations that have connections to the chancellor and others connected with DCPS.

And they represent a lot of business for the education reform industry, a major objective of which has been to change the face of teaching by undercutting the power of teachers’ unions.

To be sure, that power in DC has appeared to wane of late, as departure rates of teachers within DCPS are very high.

Between 2009 and 2012, for instance, the average teacher attrition across DCPS was 18%, compared to attrition rates in 16 urban areas that averaged 13%. The first study found that so-called “low-performing” DCPS teachers (i.e., ones who had a low rating via IMPACT, the controversial teacher evaluation system in DCPS) had an attrition rate of 46% annually, presumably through forced departures as a result of IMPACT ratings.

But even the highest performing teachers in DCPS (who are rewarded financially under IMPACT) had high attrition, departing from high-poverty schools at a greater annual rate than from low-poverty schools (13% compared to 10%, respectively).

While it may seem like a no-brainer to shed bad employees in order to improve, teacher turnover in DCPS is not merely high by any account, but underscores the finding last year of the report on mayoral control of DC public schools, which was that schools with higher poverty students tend to have less qualified teachers.

But hey–the first study found that teacher turnover got better test scores, right?

Well, yeah: but only for students of the lowest rated teachers.

This is an issue for DCPS in general, because, according to math teacher David Tansey, IMPACT is good for identifying outliers in performance, whether the very best or the very worst. But most DCPS teachers fall somewhere outside the very worst or very best IMPACT ratings.

The first study also showed a negative effect on test scores as a result of the departure of highly effective teachers in DCPS, who were the majority of teachers who left DCPS in any one year, according to the Washington Post.

David Tansey additionally pointed out that the assumptions underlying this study were considerable. For one, it noted that there were “not important unobserved factors changing at the school or grade level that influence student achievement and that are also correlated with turnover (e.g., increasingly effective principals).”

But for Tansey, new leadership at a school could be a hugely important factor in performance of both students and teachers. He felt that this could have been controlled for by comparing changes in one class’s test scores with those of the rest of the school. If the entirety is rising, he noted, it is less likely the result of a change of a single teacher.

Tansey also observed that during the time period studied, half of any teacher’s IMPACT rating was tied to student test scores, so the higher the student scores, the higher the IMPACT score. As a result, teachers rated more highly would, as the PERAA report showed, be more likely found in schools with students of higher socioeconomic backgrounds, who for the most part get higher test scores than students who are more impoverished.

(Next school year, after a brief hiatus for the new PARCC results, DCPS is going back to using student test scores to evaluate teachers—one of several forthcoming changes to IMPACT, including using student surveys and eliminating master educators as teacher evaluators, in favor of teaching coaches and observations by principals.)

Here in DC, an additional problem with test scores tied to teachers’ evaluations is that no one knows how student mobility affects test scores, since we track data by school, not by students.

And student mobility in DC is among the highest in the nation. As Tansey testified last year before the council, a third of his students in any given year churn in or out of his high school. The effect of this is profound: those kids have been moved on from one school to another as blank slates, resulting in educators and administrators struggling to figure out what each student needs and how to get it for them, all the while providing for other students within the time frame of less than an entire school year.

And of course, mirroring this student churn is the churn of teachers.

So what percentage of teachers within DCPS are new recruits of the TNTP/TFA ilk, given those FY16 contracts?

I could not find an answer to this. Although Daisa Gainey, an analyst in DCPS’s Office of Talent and Culture, noted that TNTP has since 2001 trained more than 1300 teaching fellows to go into DCPS classrooms, neither she nor Brooke Miller had any response on how many are still teaching in DCPS—nor the recruitment and retention rates for teachers from TFA and Education Pioneers.

The second study provided a bit more information about such recruitment, when it stated (p. 12) that TNTP and TFA brought in about 160 new teachers between 2011 and 2013 and that about 190 DCPS teachers are hired every year (p. 39, footnote). I found no information on retention rates.

The second study also noted (p. 22) that about a third of all teacher applicants in DCPS have no full-time teaching experience, although those with no teaching experience are less likely to be hired. (p. 26) It found that DCPS applicants without any teaching experience who ended up being hired anyway did not perform “significantly worse” that those with more experience–and even received higher performance evaluations than those who had more than a decade of teaching experience. (p. 30), although it did acknowledge “some evidence” that experience can be a strong predictor of performance in middle school, high school, and high-poverty schools (p. 34).

So, in conclusion:

We the public do not know everything clearly and unambiguously about DCPS contracts for teacher recruitment and training (and messaging!), although the chair of the council’s education committee says that the chancellor has no “sway” over these contracts. And yet, high teacher turnover in DCPS has some benefits and lack of experience in new DCPS teachers is not entirely detrimental–at least, according to studies created by those associated with organizations that both stand to gain monetarily from teacher churn in DCPS through those contracts AND have ties to the DCPS chancellor beyond the contracts themselves.

Hmm: Maybe David Grosso has a different definition of “sway” than the rest of us, since the council approves these contracts?

2 thoughts on “Does Testing Produce Better Teachers–or Just More Contracts for Consulting About Them?

  1. Mathematica is another outfit that does a lot of research on DCPS HR policies. DCPS also contracts them to rank the teachers by test scores. All the algorithms are in these reports.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s