This is a guest post by Jim Farmer.
As you know, the Statewide Longitudinal Data Systems program is funding states’ work to improve their data systems. Over the past four years, 41 states and the District of Columbia have received more than half a billion dollars from this program. It has supported states as they link data from preschool, K-12, and postsecondary education. In some states, it supports their work to track students into the workforce. We’re committed to helping all states.
Keynote presentation – Secretary of Education, Arne Duncan
For 40 years institutional and policy researchers and information technologist have met with staff from the U.S. Department of Education’s National Center for Education Statistics to learn and advise on NCES data collection efforts. NCES is a source of useful statistics on K-12 and higher education. As described by Secretary Duncan, this year NCES staff and consultants provided detail data on the anticipated changes in higher education being implemented through the state agencies and conditions for any federal funding.
At last year’s STAS-DC conference the primary focus of the 600 participants was qualifying to receive federal grants to develop statewide longitudinal data systems (SLDS). The $500 million was for in grants to state education agencies. Higher education did not receive any of those funds directly or via the state higher education executive officers.
Now NCES has become the leader of perhaps the largest information technology (IT) implementation ever attempted in the U.S. It spans 4.339 colleges and universities and 98,916 public schools with annual revenue of $1.1 trillion. The Stats-DC 2010 conference, held 26-30 July in Bethesda, Maryland, focused on this implementation.
This year’s conference provided evidence of the Department’s perspective of higher education and the requirements that would be placed on public colleges and universities and incentives for private non-profit and for-profit colleges to participate as well. Statewide Longitudinal Data Systems (SLDS) have become an unfunded mandate for higher education through new state-required data reporting requirements and data exchanges among colleges, universities, and schools. States that received funding agreed to build state-level data warehouses of these data; the model was centralization within each of the states or the use of servicers.
In the long term higher education will be restructured by the metrics used to represent the success of college and university management and teaching faculty. The Stats-DC Conference offered an opportunity to learn the underlying assumptions about education and perception of the value and use of data. A caution: The U.S. Department of Education has a number of knowledgeable, dedicated, and hard-working educators. This is also true of the state education agencies. But the design and beginning implementation of a national education data system within the available time does not permit the thoughtful planning that would have reduced the risks of unintended consequences.
National Model of Education
All of the discussion focused on the pipeline model: Pre-kindergarten, kindergarten through 12th grade, higher education, and the workforce. The students who have limited objectives, take courses outside the U.S. or study at two institutions simultaneously, have credit for military training, or transferred to a non-reporting private college or university would not be considered in the model as it is being implemented.
Two roles for higher education were discussed. One is to provide information about student performance back to the student’s high school (via the statewide longitudinal data systems) Graduation rate becomes the metric for institutional effectiveness and comparisons.
Common Data Standards
The most discussed topic was Common Data Standards. Although discussions and presentations the past two years wrote about 10 or 15 data elements, the “Consortium” said a list would be available the following Monday—that list had 199 data elements. Only a few initial data elements for student progress had been published on the CDS web site earlier. None were included in the August list.
One speaker observed “98% of the [CDS] data elements were from SIFA [specification].”
The Common Data Standards data elements support the use of high school graduation rates, college and university bachelor’s degrees with a nod to the associates degree, and employment and salary for the workforce. These are the specific metrics public higher education will be judged or managed.
The incentives for colleges and universities may be an unintended consequence: Accept only students that can be expected to graduate, focus on full-time students since they graduate more quickly, and reduce the rigor of degree requirements.1 Recent research by Georgetown University’s Center for Education and the Work show that 63% of the jobs in 2018 will require some higher education or better; only 34% require a bachelor or graduate degree, about half.2 The Department’s metric for progress and learning is the classic “Carnegie credit hour.”3
The CDS Consortium includes the U.S. Department of Education, Council of Chief States School Officers (CCSSO), and the State Higher Education Executive Officers (SHEEO). The work of the consortium has, in large part, been financed by NCES. AEM Corporation, an NCES contractor, had been responsible for the technical work. The comment period for the standards to be released 30 September closed 4 June. Recognizing the September 2010 specifications will not be completed and could be improved through changes, PESC President Michael Sessa (2010) has organized a task force to develop materials for the consortium. The CDS initiative is expected to continue for two more years even though the two years exceeds the implementation of the data systems.
The Complexity of Higher Education
The conference had 605 pre-registered participants. This included 48 from colleges and universities—up from 6 last year. Only two of the 48 had information technology titles or organizations.
Perhaps because of the proportion most of the presentations and discussions were on K-12.
Hans L’Orange, Vice President for Research and Information Resources at SHEEO (State Higher Education Executive Officers) devoted his brief presentation to the complexity of higher education. He identified the many different forms of higher education institutions—some public, some private; some non-profit, some for-profit, some focusing on research, some focusing on teaching, some large, some small. He commented the complexity of higher education—likely resulting from the increased number of specialized areas of study—makes it difficult to implement student longitudinal data systems. He could have continued to show effect of the specialization of studies on higher education. As an example from a LETSI conference call, an engineer from MIT, from San Jose State University, and from DeVry are all perceived by business as graduating top engineers, but for sharply different types of work.
L’Orange’s thoughtful perspective and clear articulation should have been a caution to the audience But the K-12 representation, appearing to be uninterested in higher education, likely did not follow his argument.
Privacy, but not FERPA (Federal Education Rights and Privacy Act), remains an issue. And the architecture of the infrastructure for exchanging data depends upon the level and type of privacy protection that is needed. Panelists considered FERPA irrelevant since it never has been and likely could not be enforced (See Krebs v. Rutgers for the limits). But a panelist cautioned that state privacy law and regulations are stricter than FERPA and can differ from one state to another. Distance learning has risks unless implemented to meet the standards of all of the states and countries where there are students.
Privacy is also an issue in using learning services, learning materials, and assessment services not on the local learning management systems.
The infrastructure to support student authentication and authorization—separately when remote services are used—will require investment. Metcalf’s Law theorizes first adopters will have little benefit from the implementation of a data transport network; hence little incentive for anyone to implement. Since implementation will be required for K-12, the SIFA specifications will likely be a component of the more general solution. The specification will be available in January 2011 and several software vendors expect to be “feature complete” at that time as well. If colleges and universities are required to exchange data with K-12 then the SIFA specification would apply unless format translations with appropriate security are used.
NCES’ next conference will be the MIS (Management Information Systems), 23-25 February 2011 in Austin, Texas. At the MIS conference earlier this year state education agencies described their current or plans for the state longitudinal data systems and their use. In 2011 many SLDS will be new systems nearing completion; the focus likely will be the areas that are not well defined such as the Web Services data transport, draft specifications for messages to be exchanged–content and choreography, and privacy.
- Babcock, Philip and Mindy Marks. (2010: 9 August). Leisure College, USA: The Decline in Student Study Time. Washington DC USA: American Enterprise Institute. [↩]
- Carnevale, Anthony P., Nicole Smith and Jeff Strohl. (2010: 11 June) Help Wanted; Projections of Jobs and Education Requirements Through 2018. Washington DC USA: Georgetown University. [↩]
- NARA. (2010: Jun 18) Federal Register Part II Department of Education: 34 CFR parts 600, 602, et al. Program Integrity issues; proposed Rule. Washington DC USA: National Archives and Records Administration Pages 34805-34890. [↩]