Chapter 2: What Quality Is and Why It Matters
Looking back on their own time in school, most people can recall extraordinary teachers. They can usually recall some lousy ones as well. Yet experts have often tied themselves in knots trying to define an effective teacher.
Somewhere between the meaningless credentials demanded by box-checking bureaucracies and the crude common-sense standard we sometimes feel in our guts (“I know one when I see one”), there are hard indicators we can refer to. If the goal of a school system is to create educated citizens, then for years Chicago Public Schools have been badly staffed. In 2011, only 21 percent of Chicago eighth graders were proficient or better on national reading standards, and only 20 percent were proficient or better at math. Yet under their longstanding rating system, 93 percent of Chicago teachers earned “superior” or “excellent” ratings. This defies common sense. Says Howard Paley, chief operating officer of the Rodel Foundation: “If students are not learning in class, that teacher can’t be considered a great teacher.” As a former educator himself, he notes that the job is not just about showing up and delivering material. “Our job is to ensure students learn.”
The traditional approach to defining and rewarding teaching quality, however, hasn’t looked at whether students are learning. Instead, districts have generally looked at input factors. Inputs are easy to see and measure. So most states put a heavy emphasis on teachers being certified. Prospective teachers generally become certified by studying at an approved university’s school of education. But a teacher’s performance, never mind the performance of his students, turns out to have little connection to whether a certificate was obtained.
Districts generally compensate teachers based on years of experience. This is easy to measure, and almost all teachers improve during their first few years on the job. But then many level off—yet continue to earn pay increases as they age. Veteran teachers can easily be earning twice as much as entry-level teachers (well into six figures, in larger districts), though there’s no evidence they achieve twice the results.
Many districts also define a high-quality teacher as one who has a master’s degree. Districts in the U.S. spent about $15 billion in additional compensation for master’s degrees in the 2007-2008 school year, though in most subjects there is no link between a graduate degree and student achievement. The funds tied up in master’s degree bonuses represent money that can’t be used for other things. Automatic compensation for master’s degrees amounted to $446 per pupil in Illinois, for instance, in 2007-2008. That could cover 20 hours of individual tutoring, or the cost of a tablet computer, for every single student.
According to studies, one of the few inputs that might matter is a teacher’s own academic skills, as assessed by college grades, standardized test scores, and the selectivity of college attended. These are proxies for intelligence. As Jessica Lahey, a teacher who blogs for the New York Times, puts it, “I think a lot of it comes down to being able to think really fast on your feet.” The children aren’t grasping your original approach, so can you improvise another way of reaching the same objective? Some states require a test of basic competence for teacher certification, but the bar isn’t particularly high.
The reality is that most current systems for accrediting and paying teachers make no meaningful distinctions between good instructors and poor ones. An important 2009 report called “The Widget Effect,” published by TNTP, used data to illustrate how teachers are effectively treated as interchangeable units by school districts today, rather than as individual talents. Almost all teachers are rated good or great by today’s ranking systems. Less than 1 percent of teachers, the report showed, receive unsatisfactory ratings at present—even in schools where students fail to meet basic academic standards year after year.
As a result, excellence goes unrecognized. When everyone is given top ratings, truly exceptional teachers cannot be identified. Nor can they be compensated, promoted, or retained. In this environment, few teachers develop as professionals as fully as they should. In their last evaluation, almost three out of four U.S. teachers received no specific feedback on improving their performance. Novice teachers are also being neglected. Low expectations and a toothless tenure process mean that beginning teachers receive neither prods nor support that push them toward excellence. And perhaps worst of all, poor performance goes unaddressed. Half of the districts studied had not dismissed a single tenured teacher for poor performance in the previous five years. None dismissed more than a handful each year. Students suffer grievously from the failure of our schools to differentiate among teachers, and no significant improvements in teacher quality are likely to take place until more honest and accurate systems of assessment and advancement are put into place.
Figuring out teacher quality is worth the trouble
So our public education system has generally rewarded inputs that don’t matter, and ignored the few that might. This is frustrating, given how much teacher quality matters. A growing research consensus finds that teacher quality (of which principal quality plays a part—a story we’ll get to later in this guidebook) is the most important school variable in student achievement. It is more important than factors that school authorities obsess over (and pour huge amounts of money into, over recent years) like class size.
Economist Eric Hanushek of Stanford University has studied teacher effectiveness for more years and in more depth than any other academic today. He recently made a simple calculation: What if we stopped arguing over what it is that makes some teachers effective and simply accept the reality that certain instructors are more effective at getting students to produce test scores? His results are powerful.
Replacing the least effective 7-12 percent of all teachers with teachers who are merely average “would bring the United States to the level of the highest-performing countries in the world, such as Finland.” This would in turn have an energizing effect on our economy, as children whose brains had been somewhat more fully stretched applied their skills to solving national problems. Approaching Finland’s educational achievement would, “by the historical patterns of economic growth, yield a gain in present value of more than $100 trillion over 80 years,” according to Hanushek.
That chunk of change is surely worth the trouble of figuring out more accurate measures of teacher quality—and then acting on what we discover. Fortunately, in the past few years, education has seen a great expansion of the available metrics—in many cases spearheaded by philanthropy. Improved ways of measuring are starting to change what we know about teacher quality, moving the definition away from crude inputs and gut feelings to something more accurate and rigorous.
One of the most intriguing changes is the growth in value-added tests. The Measures of Academic Progress (MAP) exam, which is now a staple of charter schools and used by a growing number of conventional district schools as well, assesses students at the beginning and end of the year. Because it is a computerized assessment, the MAP is able to adjust to each student. Whether a child is advanced, average, or lagging, the test will give a fine-grained measure of what she knows, and what she’s learned over the course of her time with a particular teacher.
These cumulations of how much new knowledge a child absorbs during a school year are referred to as “value added.” When you look at an entire classroom of students and determine how far they move during a school year from where they started, you have a very relevant and valuable measure of a teacher’s effectiveness. Technology is enabling a culture of more accurate and timely assessment. Schools that work with the Achievement Network, for instance, a nonprofit supported by the Michael & Susan Dell Foundation, the Walton Family Foundation, and other donors, get valuable data and assessments on the progress of their students every 6-8 weeks. They are taught how to value this data and use it to improve student instruction.
There is often wide variance in what different instructors achieve. Teachers in the top 20-25 percent of the distribution can produce 1.5 years of academic gains in a year; those in the bottom 20-25 percent average only half a year’s student progress during a school year. In other words, the effective teachers produce literally three times the gains of ineffective teachers.
Picture two children who start first grade with the exact same academic skills. One is assigned to three top teachers in a row. He or she would start fourth grade with the knowledge of a mid-year fifth grader—a level that would qualify the child for gifted programs in some schools. Meanwhile, a similar child subjected to three low-quality teachers in a row would be performing like a second grader as he or she entered fourth grade.
For children who start school behind their peers, even average teachers (those in the 25th to 75th percentile, who produce about a year of gains in a year of time) will leave them behind the pack. Only excellent teachers can close achievement gaps. The first crucial step, therefore, is to use value-added tests to reveal which teachers obtain superior, average, and poor results with their students.
Teaching must be taught
Once you know who achieves greater student growth, you can work backward to try to understand and improve performances. What do superior teachers do differently? Are there common techniques and approaches that let them take children from all walks of life and move them forward toward rigorous standards?
Doug Lemov and Uncommon Schools have become famous in education circles over the last few years for their study of this question. Uncommon Schools—a network of high-performing charter schools in Newark, Boston, Brooklyn, and elsewhere—systematically evaluated its best-performing teachers to study their methods. Lemov’s team analyzed hours of videotape—as football coaches might do—to see how teachers kept students engaged, on task, and thinking hard. Lemov’s resulting 2010 book, Teach Like a Champion, isolated 49 specific skills that excellent teachers use. Among them:
They “cold call” on students. The old tradition of hand raising is ingrained in schools, but why should only children who want to be called on get engaged? In one video clip on Lemov’s site, second-grade teacher Hannah Lofthus cold calls on the same girl twice in a row. The point is clear: I control this class, and you’d better pay attention.
They circulate. Moving around the room forces children to pay attention, and keeps people in the back from disengaging.They narrate positive behavior. (“I see Daniel thinking. Carlos made a reasonable guess...”).
They make maximum use of classroom time. Losing just five minutes a day in paper shuffling or cajoling children to gather in a circle wastes 15 hours in the course of a school year.
They stretch students to add depth to their answers. Requiring students to take their thoughts one step further cements and extends understanding.
Rather than presenting predigested information, they pull kids into the process of instruction. Great teachers know that they’re in a different line of work than preachers and motivational speakers. They avoid straight lectures and make their kids do much of the cognitive work in class.
They plan lessons intensively—not only what they’ll do, but what the students will do. Teachers prepared for things that could come up are able to make interesting diversions even as the original objectives get met. Planning doesn’t squash spontaneity, it enables it.
Lemov found that these techniques and dozens of others he has identified correspond with student gains. He and Uncommon Schools have now instructed more than 10,000 teachers on how to use them. They will train the teacher-trainers of any interested district. In about a dozen workshops per year, Lemov reports, “we do a two-day overview of some part of the content (classroom culture; reading; high academic expectations; pacing; etc.). We model what training looks like and make them experience it as participants. Then they step back and reflect on how to use and adapt these things in their setting. They leave with 50-75 videos and an electronic binder of ready-to-use materials so they can lead workshops for teachers in their own districts.” Funders have given grants to school districts and charter operators so they can go through this program.
Another very important effort to bring research rigor to the question of what good teaching looks like is the Measures of Effective Teaching project spearheaded by the Bill & Melinda Gates Foundation. Researchers studied 3,000 teachers in cities including Charlotte, Memphis, and Pittsburgh. They used student surveys and student gains on achievement tests. They relied on multiple trained observers to assess educators on how they managed behavior, created a culture of respect, engaged students in learning, and other factors. These measures were combined to create an overall measure of effectiveness for each teacher.
Then researchers assigned the teachers to new classrooms to see if their MET scores would accurately predict how effective they would be with a different, randomly assigned group of students. They did. The study concluded that “it is possible to identify great teaching by combining three types of measures: classroom observations, student surveys, and student achievement gains.”
Improving teaching requires work
Clearly, these methods could be invaluable to schools serious about improving their teaching. At the same time, hard results like these can be threatening to some incumbent teachers. To try to soften resistance to scientific assessments, the MET researchers noted that their system can be used to improve feedback and support, as well as for sorting instructors. “While some teachers’ low performance will require administrative action on behalf of students,” the measures also “provide rich information to help teachers improve their practice…. Many of the teachers who participated in the MET project video study told us that seeing themselves teach was one of their most valuable professional development experiences.”
The Gates Foundation is now focusing many of its grants around MET principles: “It’s a big thing for us,” says program officer Ebony Lee. “A lot of what we do over the next couple of years will look at what systems can do to accelerate implementation of those.” How can districts get hiring and compensation right? How can they build professional development that can help instructors improve? Most schools are just beginning to “hold teachers accountable and give them data that has consequences,” says Lee. And “some places are more hostile to change than others.”
It should be encouraging to see that the crucial elements of effective teaching are for the most part skills that can be learned. If teachers were “good” or “bad” based on some inherent qualities of personality, then options for improving performance would be limited. Ineffective teachers would be reduced to keeping their heads down and just hanging on until retirement. All school leaders could do would be to try to fire poor teachers, which can be almost impossible in school districts governed by union agreements. But if good teaching is a skill, then this opens up more positive prospects, if parties will commit to making necessary improvements. Skills can be learned and practiced, even if you’ve already been on the job for a while.
As economists like Eric Hanushek have shown (see the beginning of this chapter), removing the bottom 10 percent of teachers would be a big win for the students who’d otherwise be consigned to their classes. It would also give an upward jolt to the productivity of the country as a whole. But it won’t overcome today’s shortage of excellent teachers. We don’t just need to get rid of bad teachers; we need a significantly larger number of good teachers. It is good teachers who will overcome the achievement gap at the bottom of our schools.
Larger numbers of effective teachers should be developed in the same way that talent is created in other sectors. Recruit from pools of smart people, train them intensely in practical skills, hire carefully, conduct careful ongoing observation and measurement of results, offer feedback and practice for improvement, reward success, and point those who remain ineffective toward some different career. “Teaching isn’t that different than lots of other fields,” says Norm Atkins, founder of the Relay Graduate School of Education.
“People are beginning to rethink the staffing model of schools in a way that could dramatically improve the education kids are getting, and also transform the profession into something that’s much more attractive to high caliber people,” says educational researcher Bryan Hassel. The next chapter looks at how good teachers can be recruited, trained, retained, and evaluated, so that quality isn’t just a side effect of extraordinary personal initiative, but is built into our educational structures and incentives in a way that leads to consistent results.