"It's about instruction, stupid." That 2001 headline, which appeared in Catalyst, the magazine of record about Chicago school reform, says a lot about what has happened—and what hasn't—in the first decade of reforms. Despite some notable successes at individual schools, there is this striking fact: one-third of schools have improved, one-third are treading water, and another third appear dead in the water. Why such disparities? Books could—and will—be written dissecting that question. But one answer heard more often than not these days is "instruction."

In the last issue of the Harvard Education Letter, we highlighted research showing the importance of strong, trusting relationships in schools—the power of social trust as an agent of school reform. In successful schools, the development of trust between administrators and teachers —and among teachers themselves—depends in large part on the amount of respect they have in each others' instructional abilities. Trust begins first and foremost with the question, Is this person committed to and capable of high-quality teaching?

The Consortium on Chicago School Research—a group of researchers from local universities, community groups, and the school system—has conducted a number of studies in an effort to identify why some Chicago schools are improving while most are not. Much of that work has focused on instruction. A survey of Consortium reports reveals some of the earmarks of improving schools:

1. Improving schools have a coherent instructional program.
This requires a common framework for learning—literacy is one possibility—that gives shape to curriculum, instruction, and assessment. It also requires that principals organize personnel and resources only in ways that support and advance those core goals. In short, the school day revolves around instruction.

Researchers measured the coherence of school programs based on these characteristics through a variety of data, including surveys of more than 1,000 teachers throughout the system and field studies from 11 high-poverty elementary schools with a variety of instructional approaches (see the report entitled "School Instructional Program Coherence: Benefits and Challenges"). From 1993 to 1997, schools with coherent instruction had a 12 percent increase in scores on the Iowa Test of Basic Skills; schools without a coherent plan showed no improvement and in some cases saw their scores drop.

Coherence shouldn't be confused with programming, the researchers note. Some schools try a number of initiatives—all of which may have merit and meet real student needs—and still be plagued by low student achievement. Why? In many cases because, as the report says, the principal and the teachers "find themselves faced with a large and fragmented array of school improvement grants, programs, and partnerships that rarely afford them the time or support to adopt and master practices that may improve student learning." In successful schools, focus takes precedence over flair.

The report also notes that coherence doesn't imply heavy-handed leadership or inflexibility. In fact, it is more likely to result from a mix of top-down and bottom-up strategies—strong leadership working in tandem with teachers who have a stake in the work of the whole school and some decisionmaking role. For instructional coherence to work, a school needs its teachers to be invested in making the strategy work.

2. Improving schools offer challenging instruction.
In studying the impact of the $150 million Chicago Annenberg Challenge, the Consortium has produced a series of reports examining the intellectual demands of classroom assignments. The researchers favor the framework of high quality or "authentic intellectual work," which they define as involving the "construction of knowledge" through "the use of disciplined inquiry" that leads to "discourse, products, or performances that have value beyond school."

Using work samples and assignments from the 3rd, 6th, and 8th grades, the initial report ("The Quality of Intellectual Work in Chicago Schools: A Baseline Report"), published in 1998, showed that most Chicago students got assignments that emphasized rote learning, work that can be valuable for developing a base of knowledge but does not necessarily give students the opportunity to develop and demonstrate interpretive abilities, organizational skills, a deeper understanding of concepts and how they connect, or other "higher thinking" skills—that is, the kind of intellectual work that may become essential as the so-called knowledge economy grows. The study found that when high-quality assignments were given—which was seldom—the quality of student work was also higher. The researchers note that this is not to say that students who receive low-quality assignments couldn't do better, but rather that the opportunity to show what they know is limited by the assignments.

A subsequent report ("Authentic Intellectual Work and Standardized Tests: Conflict or Coexistence?") found that students in the 3rd, 6th, and 8th grades who received such assignments scored higher on the Iowa Test of Basic Skills than those who didn't. For example, in classrooms with high-quality assignments, scores topped the national average by about 1.2 percent. In contrast, students in classes with low-quality assignments fell below the national average by 0.6 to 0.8 percent. Similar results were reported on the Illinois Goals Assessment Program (IGAP).

The high achievers include those living in the most disadvantaged socioeconomic conditions—a rebuttal, say the authors, to those who say that a back-to-basics approach is the best way to get students in such circumstances to achieve at higher levels. In fact, say the authors, embedding basic skills in challenging, "authentic" assignments can accomplish a number of instructional goals at once.

3. Improving schools keep pace in instruction.
Consortium research has shown that many Chicago schools do not offer grade-level instruction to their students (see the report "Setting the Pace: Opportunities to Learn in Chicago's Elementary Schools"). For example, researchers discovered that in different schools introductory lessons on the parallelogram were being taught in the 2nd, 5th, 8th, and 10th grades—essentially offering 2nd-grade lessons to students in all the classes.

In slow-paced classes on literature, students repeated the same kinds of exercises year after year. Students in both the 2nd and 10th grades were asked to identify a book's setting, events, and main characters. In contrast, in classrooms where study was paced to grade level, those same 2nd graders would be asked in the 5th grade to identify idioms and guess their meanings; in the 8th grade they might have to cite examples of hyperbole and explain its usefulness as a writing device; by the 10th grade, they would be expected to engage in de tailed analysis of plot, including the uses of foreshadowing, flashback, and irony. In each of these classes, assignments built on prior learning.

A number of factors may slow the introduction of new material, according to the report. Teachers may rely too heavily on review and repetition, particularly in the weeks leading up to preparation for state-mandated achievement tests. The tests themselves and the stakes they carry may undermine teachers' belief that it is crucial to go beyond "teaching to the test." Weak homework assignments (or none at all), poor classroom management, and low expectations can also slow the pace.

4. Improving schools bolster instruction with social support.
While students in such schools are challenged to achieve at high levels, they also benefit from such supports as tutoring and good relationships with teachers. Social support and challenging instruction must go hand in hand, says the Consortium report "Social Support, Academic Press, and Student Achievement: A View from the Middle Grades in Chicago." The phrase "know every child" has become a familiar mantra in discussions of school improvement and, indeed, personal relationships have proven to be a key part of better learning. But it's not enough. Those relationships must also be geared toward instructional improvement. "Teachers who are friendly toward their students but do not demand serious academic effort are not helping students reach their full potential," the researchers write. Likewise, assigning challenging work without giving students the necessary support will be counter productive.

5. Improving schools emphasize "interactive" instruction.
In a high-stakes testing environment, should teachers use so-called didactic methods—that is, lectures, drill and practice, and worksheets that encourage students to memorize facts and procedures—or an "interactive" approach that emphasizes inquiry-based, hands-on activities; knowledge-building discussions; and projects that connect students to their larger world?

Of course, it's not an either/or question. Nearly all teachers use a mix of styles. But the Consortium report "Instruction and Achievement in Chicago Elementary Schools" shows that in a single school year, Chicago elementary school students in classes with high levels of interactive instruction scored higher on year-end tests than the city average—5.1 percent higher in math, 5.2 percent in reading. Students in mostly didactic classrooms scored below the city average in both—3.9 percent lower in math, 3.4 percent in reading. The researchers suggest that students who learn in interactive classrooms through the eight-year course of elementary school may end up a year ahead academically of those who receive didactic instruction.

Whether teachers use didactic or interactive means, all of them face the issue of how—and how much—to review previous lessons before moving forward in the curriculum. The Consortium study found that students scored better on year-end tests when instructional review was limited—4.2 percent better than the city average in math, and 4.1 percent better in reading. "Although reviewing familiar content may help build a solid knowledge base for new learning, this could also diminish learning by taking away from teaching new material," the authors write.

Didactic instruction and review get used most after 5th grade; where behavioral problems and irregular attendance are usual; where students are low achievers; in large schools; and in schools with a predominately African American and/or low-income student body—all of which may suggest that those who might benefit most from interactive instruction aren't getting it, according to the report.

6. Improving schools use effective professional development to upgrade instruction.
To teach their children well, schools must teach their teachers well. It is what ties together these other characteristics of improving schools. Instructional focus, appropriate pacing, effective teaching practice, challenging assignments, and supportive relationships are greatly enhanced by high-quality professional development, according to the Consortium report "Teacher Professional Development in Chicago: Supporting Effective Practice."

What is effective professional development? According to a growing consensus among education professionals cited by the Consortium reports, such development gets teachers to reflect in an organized way on their practice, assess student work together, share resources and strategies, and build a sense of collective responsibility for improvement of the whole school. Such development emphasizes ongoing learning in terms of both subject matter and teaching practices. It is frequent, intensive, and includes follow-up exercises; centers around a school's instructional goals; and includes perspectives from beyond the school's walls—work with outside coaches, perhaps, or with education researchers—that can freshen the pool of ideas.

To achieve this requires certain organizational supports: strong instructional leadership from principals, sufficient time, and a school culture that encourages innovation and open discussion about what's working and what's not.

Much of the Consortium's research suggests that professional development will make or break reform. For example, the study on pacing cited above demonstrated that teachers were more likely to teach at grade level in schools with strong professional communities where they had had common goals and frequent communication about instruction. Similar findings were reported in the study on interactive and didactic instruction.

A New Era in Chicago

Of course, these are not the only factors leading to improved schools, nor are they the only ones Chicago reformers have tried.

Chicago's first decade of reform focused largely on issues of school organization and governance. In the first phase of reform from 1990 to 1995, decentralization was the focus. Each school was given control of its curriculum and budget through an elected Local School Council made up of the principal, teachers, parents, and community leaders. In phase two, beginning in 1995, the central administration under Mayor Richard M. Daley and the chief executive officer of schools, Paul Vallas, imposed some necessary fiscal and administrative discipline on the system and oversaw the introduction of a high-stakes accountability system centered around standardized tests. Test scores rose and reached a plateau in 1999. Yet the city still had no systematic, coherent instructional strategy.

That may be changing in phase three under CEO of schools Arne Duncan, who took over in 2001. Duncan's first major effort has been to improve reading in K-8 schools. (A high school program is expected soon.) Schools are required to set aside two hours of instructional time each day to teach reading, divvying that time up into the framework's four areas: word knowledge, comprehension, writing, and fluency. They are also mandated to use strategies supported by research. Reading specialists have been assigned to 114 schools—the city's bottom 25 percent in reading. All schools have received money to create classroom libraries and funds for teacher and principal training in promoting literacy. After years of wrangling over governance and organization, is classroom instruction finally moving front and center in Chicago's reform efforts?