Making Informed Decisions in Education
Download this article
I’ll start this conversation by making an apparently audacious (but nevertheless true) statement: After the implementation of the Common Core State Standards , the most important change in K-12 education in the next decade and beyond will be the implementation of the Common Education Data Standard .
I know that members of the assessment consortia  are likely to take exception to that statement – their work is especially important in the development of meaningful assessments that go beyond the types of assessments that have been used mostly for accountability in the past. There is no doubt that new assessments are needed and that those assessments must be used in different ways.
But I’ll contend that those assessments cannot change education unless individual learner performance on those assessments can be linked to the Common Core State Standards, to the effectiveness of resources used to create learning related to those standards, to the preparation of educators who will facilitate learning related to those standards, to the effectiveness of professional development related to facilitation of learning based on those standards, to the measures of accountability for learning, and to other important factors that will have significant impact on the systemic transformation of education for each learner.
As Sir Arthur Conan Doyle points out, through his character Sherlock Holmes (who, by the way, never heard of a statewide longitudinal data system): “It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.” 
This is a fairly good accounting of K-12 education over the last several decades. We’ve stated that data can’t tell us everything there is to know about learning, teaching, learners, and teachers; so, we’ve made educational data into a bad thing – simply because the data don’t always align perfectly with our theories of education. The facts, as Sherlock Holmes points out, indicate that we need our educational data in order to develop theories of education that might actually work for learners and teachers. 
We actually have a lot of data in K-12 education. In the vast majority of cases, those data are inaccessible, uninterpretable, misused, abused, and (worst of all) unused. Most educational data are collected for purposes of reporting what has already happened – gradebooks, report cards, and accountability include indicators of things past. Educational data are rarely, if ever, used to guide what should happen next for the learner or the teacher – to identify a learner’s needs, to suggest effective resources to meet those needs, to inform the learner of what is needed, or to provide other meaningful information that illuminates a path forward.
Educational data are some of the most segregated data in the world. The student information system can’t talk to the lunchroom management system or the library management system. The online gradebook is not connected with other student data, including learning standards and high-stakes assessment data. Learning resources are not linked to individual student needs . Figure 1 illustrates the strategy developed by many state education agencies to integrate the types of data that exist in K-12 education today. Figure 1 also shows the lack of meaningful links between and among those data within those existing strategies .
So – with all the data we already have – and with all the new data that will be available from implementation of the Common Core State Standards and the new meaningful assessments, it is essential that we prepare ourselves to access, analyze, and use educational data in new and different ways.
But we only have to do that if we want to make a difference for teachers and learners. And that’s why I can contend that the implementation of the Common Education Data Standard is the second most important thing that will happen in education during the first quarter of the 21st century.
Implementation of the Common Education Data Standard will create the capacity to integrate the disparate sets of data within each state education agency, to inform teaching and learning as a result of that integration, and to transform K-12 education through the use of meaningful information in the teaching and learning processes. The Common Education Data Standard will enable us to bring together our data in new and important ways and will empower us to use those data to the benefit of each teacher and each learner – and, thus, to the benefit of the nation.
The following conversation will describe the Common Education Data Standard, what it means to education, why it’s essential, how to implement it, and who must be involved (and why they want to be involved) in making it happen. We will also talk about how the Common Education Data Standard can bring information (that is, data turned into meaning) to each teacher and each learner.
Implementation of the Common Education Data Standard is the second most important thing you can do as a state education agency, as a school district, and as a school. The Common Education Data Standard can – and will – change the way education does its work.
So – what’s in it for you? And what are you in for? The rest of this conversation is about helping you answer those two questions, among others.
Structure of the Conversation
The following conversation will be about implementing the Common Education Data Standard. The conversation will address the following questions, among others:
- What is the Common Education Data Standard and where did it come from?
- What does the Common Education Data Standard mean for the State Chief School Officer and her or his state education agency?
- What does the Common Education Data Standard mean for curriculum, content, and assessment?
- What does the Common Education Data Standard mean for teachers and learners?
- What does the Common Education Data Standard mean for data and information technology staff?
- What does the Common Education Data Standard mean for education application developers and content providers?
- What is the work of implementing the Common Education Data Standard at a state education agency?
- What are the benefits for implementing the Common Education Data Standard?
- What are the long-term plans for sustaining the Common Education Data Standard?
This introduction will point out some of the general issues related to the implementation of the Common Education Data Standard. Following this introduction, you will find more detailed conversation related to the questions listed above and written for specific audiences. And, of course, my hope is that you will read all the sections, regardless of your membership in one specific audience, because all the audiences must be tied together through a comprehensive data system if we are really to change the way we do the business of education.
As we talk about the answers to the questions above (and a few more that will come up along the way), I hope to help you see the value of the Common Education Data Standard so you will be able to develop your own strategy for implementing the Common Education Data Standard efficiently, effectively, and economically.
The Common Education Data Standard and your Statewide Longitudinal Data System
While Figure 1 illustrates, generally, the needed relationships (most of which are non-existent) between and among important education datasets and data systems, Figure 2 amplifies the complex nature those relationships. Both figures seem to imply that the statewide longitudinal data system (your SLDS) would be the solution for which all of us are looking.
That implication is incorrect.
State education agencies (SEAs) and other state agencies (in some states) began building their SLDS strategies as early as 2004, which was well before the development of the Common Education Data Standard (CEDS). As a result, the following K-12 data issues are not addressed (and, generally, are not addressable) in existing SLDS implementations:
- Your SEA’s statewide longitudinal data system was designed and built before the Common Education Data Standard was developed;
- Your SLDS was built to address federal reporting requirements, to meet accountability needs, and may include some capacity to do research in de-identified data subsets;
- Most likely, your SLDS is not designed to include elements that can inform teaching and learning – that is, your SLDS cannot help teachers and learners;
- Your SLDS is not compatible with nor comparable to another SEA’s statewide longitudinal data system;
- Your current SLDS does not support (and, in fact, discourages) inter-state collaboration because of the high costs associated with converting data to be compatible and comparable;
- Conversion of your current SLDS to another database format based on the Common Education Data Standard is prohibitively expensive and time-consuming; and
- If you did have the resources to convert your current SLDS to a CEDS format, you would have to re-write all the applications you’ve already developed to work with data from your SLDS – which also would be very expensive and time-consuming.
So – what is the relationship between your existing SLDS and the Common Education Data Standard?
The answer to that question is at the heart of transforming the use of educational data to transform K-12 education practices and results.
Your SEA, along with other SEA’s across the country, wants to become more service oriented. That is, state education agencies wish to change their focus from enforcement and compliance with federal and state regulations to provision of services to teachers and learners.
Your current SLDS is not designed to support that change in mission. The Common Education Data Standard provides the mechanism by which your data can support that new mission. CEDS provides the foundational elements and the tools by which your SEA can create – and use – a comprehensive real-time dataset that can inform teaching and learning. CEDS provides the mechanism through which your SEA can address its new mission with a cost-effective strategy that creates compatible and comparable datasets that are accessible and usable.
Figure 3 illustrates the basic manner in which your state education agency can
- use its existing datasets,
- map its existing elements to the elements in the Common Education Data Standard,
- convert its existing element values to the values defined by the Common Education Data Standard, and
- store the converted elements and values in a physical operational data store.
That operational data store will contain all the elements defined by the Common Education Data Standard, based on the logical data model that is part of that standard; however, some of the elements in your state’s operational data store may be empty (or, in geek-speak, some of those elements may have “null” values). Your SEA can determine which elements it wants to populate and which elements it wants to leave empty. That decision can change as your agency’s data needs change.
Before we continue this conversation about the Common Education Data Standard and its physical implementation as an operational data store, let me be perfectly (or as perfectly as possible) clear about what the Common Education Data Standard and the operational data store are – and, as importantly, what they are not. This will be of more interest to those of you who will have the responsibility for implementing than to those of you who will consider a policy decision to implement or those of you who will be using the data after implementation; however, it is important to understand the basic context in which the Common Education Data Standard exists and the purpose of the operational data store (the ODS).
So, with that in mind, please consider the following as you begin your consideration for implementing the Common Education Data Standard:
- Implementation of the Common Education Data Standard is entirely voluntary . There is no requirement to adopt or implement the standard. Period.
- The Common Education Data Standard is NOT data; the standard defines a set of elements and formats for the values to be stored for those elements.
- The Common Education Data Standard does NOT require you to change or re-build any of your existing databases or datasets – in fact, this conversation is about how you can implement the Common Education Data Standard without changing what you’re already doing or what you already have.
- Neither the Common Education Data Standard nor the operational data store is related to the Ed-Fi  data model developed in Texas by the Michael and Susan Dell Foundation .
- The implementation strategies that are described in the following pages are NOT related to the Shared Learning Initiative (SLI)  that is being piloted in several states by the Bill and Melinda Gates Foundation . Neither the Common Education Data Standard nor the operational data store is required for SLI. Implementation of the Common Education Data Standard and the creation of the operational data store are NOT part of the SLI project. Period.
- The operational data store is NOT a “big black box in the cloud.” The operational data store is a standardized physical copy of your data stored on one of your servers or in another environment that you have chosen.
The implementation of the Common Education Data Standard through the creation of an operational data store has been described in earlier papers and discussions as the “Blue Box.” This conversation will refer to the “Blue Box” concept through the more general term of an “operational data store.” If you are familiar with the Blue Box, this conversation may help broaden your understanding. If you are not familiar with the Blue Box, don’t worry; the following conversation does NOT rely on your becoming familiar with it.
- The operational data store is a dataset; it is NOT a data system. The operational data store is a set of data, owned and managed by your state education agency, which can be used as the foundation for software applications that your state may need in order to inform teaching and learning and to provide other educational services to educators.
- The ultimate purpose of the Common Education Data Standard and its physical implementation in an operational data store is to create the capacity in your state to build and use comprehensive datasets within your state in a format that is compatible and comparable with the datasets in other states.
Now that you know, basically, the relationship between CEDS, the operational data store, and your existing data systems, we can expand the conversation about the Common Education Data Standard and about strategies for implementing your own operational data store to support teaching and learning and to change the way teaching and learning are done.
Figure 4 illustrates the relationship between your existing data, the Common Education Data Standard, and the operational data store that represents the physical implementation of CEDS. Later in this conversation, I will discuss the “how” part of this strategy. But, there is still the need to talk in more depth about the “what” and the “why” parts. I hope you’ll continue the conversation in the following pages.
Gary West 
April 24, 2012
Permission is granted to share this document with anyone, in print or electronic format, as long as the content is not changed in any way.
Endnotes for General Introduction
 Common Core State Standards: http://www.corestandards.org/
 Common Education Data Standard: http://ceds.ed.gov/
 USED Announcement: http://www.ed.gov/news/press-releases/us-secretary-education-duncan-announces-winners-competition-improve-student-asse; Smarter Balanced Assessment Consortium (SBAC): http://www.smarterbalanced.org/; Partnership for Assessment of Readiness for College and Careers (PARCC): http://www.parcconline.org/
 From “A Scandal in Bohemia”; Sir Arthur Conan Doyle; 1891; You can download and read this “adventure” of Sherlock Holmes at http://www.gutenberg.org/files/1661/1661-h/1661-h.htm. You can find more thinking about fact and theory at http://en.wikiquote.org/wiki/Fact_and_theory. You can find more information about Project Gutenberg at http://www.gutenberg.org/.
 Historically, educational data have been used to support the politics of folks who want to show what education has done (or, more probably, not done) in the past, to support the theory that traditional education strategies (that is, my ways of teaching) are better than all the new-fangled strategies, to support the myth that learning must be fun instead of work, and to support other theories of teaching rather than theories of learning. As we will discuss in other parts of this conversation, our data have been twisted to conform to our theories – and we’ve been using selective data in the twisting. If we are to transform K-12 education, we must have more, new, and comprehensive data – and we must use those data to develop real solutions rather than to support our existing solutions, which can be demonstrated to be ineffective and downright unsuccessful. This is one of the reasons that the Common Education Data Standard is so important; that standard defines the data we need and builds our capacity to use those data effectively to inform teaching and learning.
 Much of the K-12 education discussion over the last several years has centered on the Common Core State Standards, which define what students should know (and, thus, learn). Publishers of educational resources (textbooks, digital content, video resources, online materials, etc.) have worked diligently to align their resources with those standards. As a result, we have students “aligned” to the standards and we have educational resources “aligned” to the standards. But we are missing one important link – and our data must be able to fill in the missing information. We are unable to determine which of the resources are most likely to align with a single student’s learning needs. Our data systems must be able to help teachers and learners predict which resources will effective with Johnny in this afternoon’s learning activities. This is another reason that the Common Education Data Standard is so important in building our capacity to inform teaching and learning.
 Figure 1 illustrates the failure of your current statewide longitudinal data system to meet the needs related to informing teaching and learning. Your current SLDS is based largely on convenience data – that is, the data that have historically been collected. Those data have been collected and used for reporting and accountability. In Figure 1, the solid arrows connecting disparate datasets to the SLDS indicate most of the existing data being accumulated in your SLDS. The dashed arrows indicate that other data are needed if your datasets are to inform teaching and learning. Those data are defined or can be defined in the Common Education Data Standard and can create the capacity for your data systems to inform teaching and learning.
 Like the Common Core State Standards, implementation of the Common Education Data Standard is voluntary. Your state education agency is not required to implement the standard. However, implementation of CEDS – like implementation of the Common Core – builds specific capacity for your agency to transform K-12 education through the use of comprehensive, compatible, and comparable data. Implementation of CEDS also builds capacity for your agency to work with other state education agencies in collective and collaborative ways, whenever appropriate. Implementation of CEDS integrates your efforts at implementing college- and career-ready standards with meaningful assessments to create capacity to inform teaching and learning in real ways and in real-time.
 Ed-Fi data model: http://www.ed-fi.org/
 The Michael and Susan Dell Foundation: http://msdf.org/
 The Shared Learning Initiative (SLI) is funded by the Bill and Melinda Gates Foundation through a non-profit LLC named the Shared Learning Collaborative (SLC): http://www.gatesfoundation.org/Grants-2011/Pages/Shared-Learning-Collaborative-LLC-OPP1041367.aspx; you can find more at http://www.slcedu.org/
 The Bill and Melinda Gates Foundation: http://www.gatesfoundation.org/Pages/home.aspx
 Gary West was high school math teacher for ten years, a federal projects coordinator for a small rural school district for nine years, a technology and data systems director for a moderately sized school district for more than nineteen years, the Chief Information Officer for the South Carolina Department of Education for more than three years, and the Strategic Initiatives Director for Information Systems and Research at the Council of Chief State School Officers in Washington DC for one year. He has been emphasizing, since the beginning of time, the use of data to guide teaching and learning – not just the use of data to report what has already been taught and learned.