McKinsey & Co recently published a compelling report on the strategies that high-performing nations use to recruit teacher talent from the top third of college graduates. In the US, the report points out, only 23% of teachers entering the workforce are among the top third, and the proportion is even lower (14%) among those taking jobs in high-poverty schools.
In Finland, South Korea, and Singapore, by contrast, 100% of teachers are recruited from the top third of college graduates. While academic talent alone doesn’t fully predict teacher performance, it certainly can’t hurt to have the best and brightest teaching the nation’s students.
In the US, only Teach for America is so selective and successful in attracting top talent. TfA admits high-achieving elite students, often from the Ivies, and admission carries serious prestige. However, TfA graduates often have no intention of staying in teaching more than a few years, and are still subject to the same labor market conditions that drive top talent into other professions. Even as TfA continues to rise in prestige, teaching itself remains a lower-status job in the US than in top-performing nations.
Most teacher preparation programs in the US screen only for basic academic and personal qualifications; recruiting “the best of the best” is difficult for a variety of reasons:
- Starting and maximum teacher salaries are often lower than those in other professions, and are based on seniority rather than performance
- Teacher-bashing has become a national pastime, and the prestige of the profession seems to be on the decline
- Teacher education classes are often made very easy, to ensure that everyone succeeds; consequently, high-achieving students choose majors that are more engaging and challenging
On this last point, I can speak from personal experience. Many of my friends left the teacher education program at my university after sitting through a 3-credit course on how to use an overhead projector and a photocopier. They had no interest in having their intelligence insulted, nor in being surrounded by the university’s worst students for four years. My interest was sustained only by my much more rigorous science coursework. I graduated first in my class in the physical sciences department, yet I was the only physical science major to also earn teacher certification. Clearly, my university had failed to attract many of its best students into the teacher education program.
Finland, South Korea, and Singapore use a combination of salary, prestige, and labor-market responsiveness to ensure that teaching is a career to which the best and brightest aspire. Critically, these strategies are used with great purposefulness on the national level, whereas in the US, there is no national strategy for developing a strong teacher workforce.
While these nations’ high performance is due to a variety of factors, teacher quality is certainly a centerpiece of their national strategies for excellence:
When asked what’s accounted for the extraordinary rise in the nation’s educational performance in recent decades, Korean officials put the matter simply: “The quality of an education system cannot exceed the quality of its teachers.”
It’s inspiring to think what could happen in US schools if we developed a national strategy for attracting and retaining top talent in teaching. It’d be nice to know that the only people accepted into teacher certification programs are top-third students, but obviously that’s a long way off.
I hope that as a nation, we can recognize the logic of the McKinsey report’s recommendations and start to take steps to attract high performers to our nation’s classrooms. In addition to identifying how Finland, Singapore, and South Korea attract top talent, McKinsey conducted market research to determine the approximate cost of developing a comparable strategy in the US. It wouldn’t be cheap, but at a time when we’re spending billions to improve our schools, making teaching a more attractive profession to our top college graduates is certainly a worthwhile investment.
Ignoring these nations’ examples would be to stake America’s future on the idea that the U.S., alone among nations, can prepare its children to thrive in a global economy while relying on lower-achieving graduates to teach them.
One perplexing finding: A large proportion of teachers who rate highly one year fall to the bottom of the charts the next year. For example, in a group of elementary-school math teachers who ranked in the top 20% in five Florida counties early last decade, more than three in five didn’t stay in the top quintile the following year, according to a study published last year in the journal Education Finance and Policy.
Meanwhile, the District of Columbia began evaluating teachers based on test scores last school year, and fired more than 150 teachers after the school year because of poor performance. Test scores count for 50% of teacher ratings in subjects that are tested.
A report from the Department of Education released last month shows that even with three years of data, one in four teachers is likely to be misclassified because unrelated variables creep in.
Even with these questions, relying on student test scores to create a quantitative assessment of teachers might be better than the current standard practice. At many schools, principals grade teachers based on a few minutes of classroom observation (and then give most of them high scores).
–Carl Bialik, Needs Improvement: Where Teacher Report Cards Fall Short (WSJ)
One of the most visually arresting elements of the article is this chart that shows the instability of value-added ratings from year to year:
As you can see, the idea that excellence in teaching (as defined by impact on student learning) is a stable construct is not supported by the data from this study. Value-added is not ready for prime time.
But is it better than what we currently use for teacher evaluation?
But even skeptics of test-score-based evaluations acknowledge that a uniform, data-based approach for ranking teachers could be superior to subjective methods—such as principals’ observations—that still predominate in schools. “Damn near anything is going to be an improvement on the status quo,” says Daniel Willingham, a cognitive psychologist at the University of Virginia.
Indeed, the sad state of teacher evaluation is not due to a lack of data, but a lack of diligence on the part of school administrators. But I would disagree sharply with Willingham’s prescription for change. Far too much energy has been devoted in recent months to making a show of doing something – anything – to address issues of teacher performance. We seem to be enamored of anything that uses data and promises accountability, even if the results are patent nonsense, as in the illustration above.
We like this approach because it’s neat and quantitative, even if it’s dead wrong. Truly improving instruction is a lot more complicated. In order to do a better job of identifying effective and ineffective teachers, and ensuring that the latter make the necessary improvements, principals will have to make time to observe instruction, provide feedback, and take responsibility for the quality of teaching in their schools.
Daniel Duke & Edward Iwanicki wrote a brief article in 1985 discussing the idea of “fit” between a principal and the school, community, and district context. They explored 9 cases of principals who were fired or reassigned due to (according to their superiors) a lack of fit rather than poor performance.
In one case, a principal was removed due to a perceived inability to handle the workload of leading a large school (p. 29). In other cases, the removal was due to communication, decision-making, handling change, or other factors.
How can principals avoid problems with “fit”? The authors suggest that principals must filter the expectations of their various constituencies, developing their “received role,” or the set of expectations they accept:
The earlier principals share the job expectations that they believe are real and important to fulfill, the easier it is to deal with any potential problems of fit. p. 30
In other words, principals must clearly understand what’s expected of them, and strive to fulfill the most important expectations of their critical constituencies. Of course, discerning which constituencies and which expectations to attend to, and how to fulfill them, is no small matter.
Duke, D. L. and R. J. Stiggins (1985). “Evaluating the Performance of Principals: A Descriptive Study.” Educational Administration Quarterly 21(4): 71-98.
Is education a science, and if so, what kind of science? What implications does this have for instructional leadership?
We typically think of physics as the ideal science – it is consistent, universal, and predictable. An experiment conducted in France can be replicated in Mexico or the United States, and the same results can be expected. If teaching is a science, why isn’t there a similar level of predictability?
This question has enormous implications for both instructional leaders and for policymakers. In the October 2009 issue of Educational Researcher (AERA), Pamela A. Moss, D. C. Phillips, Frederick D. Erickson, Robert E. Floden, Patti A. Lather, and Barbara L. Schneider take up the question of quality in education research in their article “Learning From Our Differences: A Dialogue Across Perspectives on Quality in Education Research” (38: 501-517).
Erickson explains why educational research is constantly (and unfruitfully) compared with natural science research:
The reason social science has suffered from physics envy is the assumption that the social world is basically like the natural world. What makes physics and chemistry work is an assumption of the uniformity of nature—a unit of force, or of heat, or a chemical element is the same in Britain as it is in France or on the face of the moon or in the most far-flung galaxy.
In the 19th century, as the social sciences were developing (looking over their shoulders at the mathematical physics of Galileo and Newton), there was a serious argument over whether social sciences should model themselves after the natural sciences or try for something else.
Adherents of what became the meaning-oriented approaches to social inquiry, the hermeneutical approaches described by Moss (2005b), took a position that meaning differences made such a difference between one social setting and another that there was in effect a nonuniformity of nature in social life (as I called it in my 1986 article on qualitative research on teaching; see Erickson, 1986). The notion was that it is local meaning that is causal in social life, and local meaning varies fundamentally (albeit sometimes subtly) from one setting to another. One of the consequences of this notion is deep distrust of the possibility of any generalization at all in social research…
Close descriptive study of a setting, based on extended participant observation and interviewing, doesn’t try to generalize directly from that setting to others…what happens in Miss Smith’s first grade is fundamentally different as a local ecology (subtly different, despite surface appearances of similarity) from what happens in Miss Jones’s room across the hall in the same school building. (Parents know this—that’s why they fight to get their kids into Miss Smith’s room, away from Miss Jones.) Nor is what happens in Miss Smith’s room quite the same as what happens in Miss Robinson’s room in the next school district. It follows that policy evidence for “scaling up”—trying to get everybody to adopt “best practices”—no matter how well produced technically—just doesn’t tell us what we need to know as educators. Best practices, as specific behaviors, don’t travel intact across the hall in one school building, let alone across the country. (p. 508, emphasis added)
Erickson’s extended argument implies what we, as instructional leaders, have long known: good teaching can’t be measured simply by checklists of “best practices.” Some of our best teachers don’t use the best practice du jour, and some of our most compliant adopters of new best practices are unable to pull everything together to create powerful and coherent learning experiences for students.
This suggests that instructional leadership is going to remain a labor-intensive, and inherently local, endeavor. If we want to improve the quality of teaching and learning in every classroom, we will need to be in every classroom. We will need to know the research, but the research will not save us. It may give us direction and help us understand what is taking place in our classrooms, but it does not (and cannot) provide a recipe for high-quality instruction.
In order to understand what is happening in a classroom and whether it’s good for kids, we need to adopt what Elmore (in Instructional Rounds) calls a descriptive-analytical-predictive approach. Briefly, we must ask three questions:
- What is taking place in this classroom?
- What dynamics does this create?
- What learning do we expect this set of dynamics to cause?
After asking these questions, we can consider what next steps will improve the teaching and learning taking place in the classroom.
How do you see social science research influencing your work as an instructional leader?
One set of findings is of particular relevance to principals. The study divided teachers into three groups – contented, disheartened, and idealists. While surveys of this type show only correlation (not causation), teacher responses about the support they get from principals are revealing:
In short, there is a strong relationship between job satisfaction (the category in which the teacher is placed, based on responses to other survey questions) and the leadership and support provided by the principal.
Today I read a quantitative study of class-size effects, conducted in Tennessee around 1990. This study (Finn & Achilles, 1990) randomly assigned teachers and students to one of three conditions – small class size, regular class size, or regular class size with aide.
The findings are interesting, but I’d like to point out something unusual about this piece of research: it actually compared different policy options in real school settings to see which was better.
It’s remarkable how rare this type of experimentation is in education. Perhaps it’s the ethical difficulties, which are numerous.
First, we must do no harm to students in conducting research. Second, if it turns out that one option is superior to the other, how do we make it up to the students who received the inferior treatment? Or, at the very least, how do we justify to parents and other stakeholders that the research did no harm?
The Huffington Post uses A/B testing to see which of two headlines gets more clicks. After a few minutes of testing, the superior headline is declared the winner, and is the only one used from that point forward.
How much improvement could we see in education if we were able to regularly employ this type of “which is better?” experimentation?
There is no lack of research for those who seek to promote discussion of effective teaching. The issue is whether or not educators are prepared to accept their responsibilities to work together to become proficient consumers of that research. A professional learning community will fulfill that responsibility by ensuring that frequent and focused discussions on teaching and learning are the standard practice in its school.
DuFour & Eaker, Professional Learning Communities at Work, p. 226
A study by Texas Appleseed, a “non-profit, public interest law organization,” reports that many students are being sent to alternative schools for minor, nonviolent infractions. The report, entitled Keeping Schools Safe While Reducing Dropouts: Breaking the School-to-Prison Pipeline, asserts that zero-tolerance policies are inappropriately funneling students into alternative programs, which have a dropout rate five times that of a traditional school.
As the report’s title indicates, there is a strong correlation between being sent to an alternative school for disciplinary reasons and being sent to prison. However, school administrators have significant discretion in sending students with nonviolent behaviors to these programs. As the following chart shows, the majority of Disciplinary Educational Alternative Program (DAEP) referrals are for nonviolent offenses:
According to the Houston Chronicle, referrals to DAEPs disproportionately affect students of color and students in special education programs. In some districts such as Pasadena (a suburb of Houston), students in kindergarten or first grade are sent to alternative placements:
Suburban districts tend to be the most the most punitive, said Deborah Fowler, Texas Appleseed legal director.
Of these, the Pasadena Independent School District was one of the harshest.
It is the only district in Texas to appear on the report’s Top 10 districts with the highest alternative school referral rates each of the last five years. Houston-area school districts that made multiple Top 10 appearances in the past five years also include Galveston (twice), Spring Branch (twice) and Katy (three times).
“These numbers indicate that it is not the behavior that determines whether a student is referred to a (district alternative education program),” the report’s authors wrote. But rather, “it is, in large part, the district where the child goes to school.”
Other districts were spotlighted for sending black or special education students to alternative schools for nonviolent, noncriminal offenses at at least twice the rate of other students for the past five years. They include Humble, Klein, Katy, Pearland, Tomball, Goose Creek, College Station, Bryan and Huntsville. link
While clearly this is an embarrassment for the districts named in the Chronicle article, there are several implications for school leaders:
- We must do whatever we can to reduce racial and socioeconomic disproportionality in discipline referrals
- We need to create schools where students are set up for success and safety, not assume that kids will need to be sent to alternative schools
- We need to create systems and structures to promote positive behavior, not just systems for punishing negative behavior efficiently
- For the small number of children who cannot be successful or safe in a regular school environment, we need to create alternative schools where students can become successful, not take the next step down the path to prison
What do you think? Leave a comment.
ASCD reports on a study that may indicate that self-discipline is a stronger indicator of school success than IQ. The ResearchBrief summarizes a study focusing on 8th graders in a magnet school in the Northeast. The ResearchBrief lists both implications of the study and caveats that should be taken into consideration.
Today’s SmartBrief introduces the article by saying “Some research suggests that students who are highly self-disciplined may be able to better focus on long-term goals and make better choices related to academic engagement.” But the study measured self-discipline by giving a questionnaire to students, teachers, and parents. When using a questionnaire and measuring something as nebulous as self-discipline, it is likely that measuring self-discipline is simply another way of measuring school success. Inasmuch as the factors leading to school success are well-known among students, teachers, and parents, it is meaningless to say that better students have these characteristics. Of course they do, and everyone knows it.
The study is also based on a single school, a magnet school with selective admission practices. The representativeness of this population is not established in the review, though this is pointed out in the caveats section.
This leads to crucial questions regarding summaries of research intended for educators:
- When is it misleading to summarize a study?
- Do people hear the caveats appended to research as clearly as they here the apparently dramatic findings?
- Do people really understand the principle that correlation does not equal causation?
Educators are pressed for time, and research of this type is not intended for a broad audience of classroom teachers and school administrators. However, it is critical that all educators understand how research is done, so they are able to determine the implications and significance of a study like this one. Otherwise, statements like “Self-discipline outdoes IQ in predicting academic performance of adolescents,” the title of the original study, will be accepted as un-nuanced gospel.