A recent discussion between David K. Cohen of the University of Michigan and the Fordham Institute’s Chris Tessone used the term infrastructure. Cohen, in an earlier post on ShankarBlog (from the American Federation of Teachers’ Shankar Institute), argued that individual reforms such as the DCPS IMPACT teacher performance review system were insufficient to fix the system overall. Cohen argues that what is needed is an instructional infrastructure that provides teachers with tools for their job. Cohen lists four things that infrastructures should have:
- curricula or curriculum frameworks,
- exams to assess students’ learning of the curricula,
- instruction that centers on teaching that curriculum,
- and teacher education that aims to help prospective teachers learn how to teach the curricula.
Cohen further argues that school systems around the world commonly have infrastructures with these components. Tessone’s response focuses on the need to get rid of bad teachers and points to IMPACT’s benefits in weeding out those teachers that are the cause of an underperforming system. These two arguments have a small intersection around the IMPACT system. Otherwise, they are addressing different issues: Cohen focusing on the need for a broad infrastructure supportive of teachers and Tessone focusing on the quality of staff. In this post I present another perspective. I argue that the IMPACT system is infrastructural. It may not be the infrastructural in the sense that Cohen and many educators would wish it to be. But, it is a component in a developing sociotechnical infrastructure that has many dimensions and can figure into the future of education. Like any infrastructure, we may find it has both good and bad aspects.
What is an Infrastructure?
As Cohen points out, electrical systems and railroad networks are infrastructures. We can think of these infrastructures as technical things. These are national projects that operate to integrate different regions and localities. The Internet and the cellular phone system, like broadcast television systems before them, are also infrastructures that operate across distances and enable markets and industries. An important dimension of these infrastructures, and all infrastructures, are standards. In most cases these standardizations remove variation that adds little value. Rather than several independent communication systems, the technical communication standards allow us to make calls between different carriers easily. This standardization, most would agree, is a good thing.
Infrastructures can also be thought of in another way. From the perspective of social activity, an infrastructure can be viewed in terms of how it enables and makes different activities convenient. The infrastructure that is the Internet supports a range of social practices not possible to envision without it. Many advocating for new tools for teachers, including the cyberinfrastructures for learning and the Digital Promise, no doubt envision infrastructures that can transform teaching in the way that social media have transformed parts of daily life. These infrastructures, however, may contain embedded standardizations that can impact teaching practice in ways not evident now.
IMPACT is Infrastructural
The IMPACT system is a form of infrastructure. While it is a measure to both reward and remove teachers based on performance, it includes many of those things we would find in an infrastructure. It has technological components. These include the student test scores and the value-added model that is used to provide a measure of the teacher’s contribution on student learning. These scores and models have been criticized as being limited in what they measure, limited to only some subjects, and technically problematic as Cohen points out. However, the scores and models IMPACT uses are similar to those being used in states across the land. While we have no unified system of testing in this country – for political, legal, and technical reasons – in the way we have a common Internet, the DCPS test system and IMPACT’s value-added models are not unique. They would scatter across a map of the US today and next year and the year after that map would contain more of them. IMPACT happens to be one of the more fully developed at the present time and one used for visible changes to DCPS teachers. these niches then over time may come to form a broad network of systems that affect the social profession of teaching across the land.
IMPACT also has other components besides the value-added measures to consider. It involves a rubric-based assessment of teaching practice where teachers in both tested subject and non-tested subjects are evaluated by master teachers who come and observe their teaching and rate them according to “nine commandments” of good teaching. Few would argue that the aspects of teaching in these commandments are good. However, their use within IMPACT causes teachers to focus on them specifically. Over time, we may see a standardization of some teaching practices, especially with teachers who are evaluated in the middle of the pack so will look to the evaluation system for how to improve. In fact, as Headden pointed out that many teachers find the evaluations and feedback useful as a way to reflect on their practice, although the reviews were mixed and the evaluations far from a perfect measure.
In short, DCPS’ IMPACT system is similar to FaceBook and LinkedIn. All are sociotechnical developments that are tied to infrastructures and form mini0infrastructures themselves for certain types of activity. A major component of it and other similar systems emerging throughout the country is the testing and data distribution infrastructure put in place by No Child Left Behind.
Teachers, Infrastructures, Management Science, and Systemic Reform
What do these issues have to do with educational systemic improvement?
Both Cohen and Tessone focus on the role of teachers. Clearly this is necessary when considering the performance overall of educational systems. Is it sufficient? Can we imagine that just cleaning out the teacher corps and rewarding the high-performers as Tessone argues for or providing all teachers with the coherent tools for their classroom jobs as Cohen argues for is enough? What about the principals, curriculum specialists, special educators, and all of the other individuals that work in a modern school system. School systems, most of them, are complex organizations with structure and differentiation of roles. What tools do we provide them? How do we support those who support the classroom teachers? How do we measure them? This is new territory for educational researchers. While performance management approaches and calls for management science to be applied to education have been occurring for several years – often from the same quarters that are bringing greater use of data systems and dashboards for decision making – the theoretical connections between these fields and education have been slow to emerge . The focus on education as a collection of teachers rather than as a complex organizational problem has been too common. However, one of the legacies of education legislation in recent decades, including NCLB and the Individuals with Disabilities Education Act (IDEA) has been to broaden instructional responsibility beyond teachers. This broadening of instructional responsibility – often mediated with test data that flow over district data infrastructures – calls us to look beyond the classroom teacher. The technical challenges being raised by the efforts to link teacher and student data records are providing indications of the complexity of educational practice and the need for a wide organizational and systemic lens in considering how new tools, including those that Cohen and Tessone argue for, can really help transform education.