Replicating Good Practice: Schools and Starbucks

One of the most troubling aspects of school reform over the past 30 years has been the inability to replicate good schools at scale. Since the mid 1970s, researchers have been able to identify the characteristics of effective schools—high expectations, clear standards, strong leadership—but no one has been able to identify a similar strategy for replicating such schools. The systemic standards movement has been the primary response to this problem within the public system, while entrepreneurs seeking to spread charter schools or model curricula face similar challenges from a different angle.

Why is this so difficult? After all, I'd bet the Starbucks on my block looks pretty similar to the Starbucks on yours. Over the past week, I've had the chance to hear two presentation on the challenges of creating effective schools at scale: one from within the system and one from without. Despite their differences as insiders and outsiders, the two had startlingly similar stories to tell about the challenges of creating good practice at scale.

Former Boston superintendent Tom Payzant talked about the challenges of bringing high quality new practice to scale in a recent HUGSE accountability workshop. Speaking from the perspective of the superintendent he used to be, he emphasized the role of trained intermediaries--coaches, district superintendents, and, of course, principals--as the key components to creating the capacity for system-wide improvement. In one sense, the vision here is unavoidably top-down, as the direction comes from the center and moves outward into the districts and schools. But at the same time, Payzant stressed creating opportunities, within this larger vision, for schools to shape the programs to their needs, to choose curricula that fit their missions, and to avoid overly prescriptive programs that denied talented teachers the room to work. If the flaw of “tight on ends/loose on means” was that schools, given the freedom to innovate, didn’t know how to improve the quality of their instruction, the coaching-heavy strategy seeks to remedy the problem by providing external support and expertise for how to improve their practice.

At the same time, Payzant acknowledged that there were some significant challenges to this model. Finding knowledgeable and skilled coaches can be just as difficult as finding high quality principals and teachers. Some schools are much quicker than others to take advantage of new opportunities (such as new curricula for reading), and the lowest performing schools are, not surprisingly, the least likely to do so (more on this in the next post). And a fair number of teachers, particularly those in middle schools and high schools (not to mention universities) are highly resistant to the idea that an external coach should work with them on their teaching, which they see as a core part of their professional expertise.

Approaching similar questions from a different angle, Eastern Michigan researcher Donald Peurach gave an AERA presentation on the efforts of Success for All (SFA), a Baltimore-based reading model, to replicate their efforts at scale. Drawing on his 10 year study of SFA and a literature on organizational replication, he challenged the idea that replication was achieved through developing “best practices” in a hub and then transferring this knowledge to field sites. Rather, he argued, field sites needed to find ways to take what was offered from the hub and make it their own, and the hub needed to learn how to revise its core set of knowledge on the basis of feedback from the field. In this notion of interactive learning, as opposed to transferring knowledge, Peurach’s findings mirror earlier research on the limitations of top-down implementation, and are consistent with a broader constructivist approach to organizational learning and change.

Peurach argues that the challenges to replication are found in four central places: the model, the outlets, the environment, and the hub. Among the barriers he cites: the difficulty of knowing exactly what it is about a successful model that makes it work (the model); field sites that lacked the leadership or previous knowledge needed to make sense of the reform (the outlet); cultural resistance among teachers to the introduction of external routines into their work (also the outlet); a policy and funding environment that was too uncertain to ensure long-term continuity (the environment); and the many managerial challenges facing the hub as it seeks to address the problems in the model, in the outlets, and with the broader environment (the hub). While this analysis is of a program sponsored by an external academic entrepreneur, he argues that similar problems would face state or district efforts to boost capacity. Judging by Payzant’s analysis, he is not wrong.

So where does this leave us? Peurach places his faith in “routines,” which he sees as the key building blocks of organizational replication, tools that encompass both knowledge and tacit skill and can be standardized for widespread usage. The analogy here is a familiar medical one: that doctors are not mavericks who risk patients’ lives in the name of professional discretion, but rather a profession governed by standards of care that serve to ensure a certain consistency in the quality of practice. There is something appealing here, in that it points to the longstanding weakness of teachers’ individualistic notion of professionalism, a vision that is more about being free from oversight and less about working together to achieve quality practice. At the same time, there is something unappealing about applying “routines” to teaching, because teaching, as a form of collective thinking, is most alive when it is open to new or unforeseen possibilities. A middle ground (often proposed but infrequently implemented) might combine the best elements of both: teachers would be given pedagogical strategies but not routines, and other teachers, coaches and principals would consistently observe classroom practice and offer ongoing feedback about how to improve practice. Such a strategy would not solve all the problems of hub, model, outlets or environment, but it would be a start towards spreading quality practice in professional work. Starbucks might learn a thing or two.

Crossed posted at Usable Knowledge.