These days, education stakeholders are constantly clamoring for data – data on student achievement, data on spending, data on teachers – you name it, they want it.
The Federal Education Budget Project (FEBP), Ed Money Watch’s parent initiative, houses a database on its website that provides much of these data for both K-12 and higher education. In addition to relying on state sources, FEBP relies heavily on the U.S. Department of Education for data, particularly for higher education data at the state and institutional level. The primary source for these data is the Integrated Postsecondary Education Data System (IPEDS), a publicly available, but somewhat confusing database on postsecondary achievement, costs, and demographics.
While IPEDS has provided institution-level data for years, they recently launched a State Data Center to provide aggregated data at the state level. And though that data center is publicly available, IPEDS has not yet figured out how, exactly, to make that dataset as useful as possible. Our recent experience with the State Data Center’s graduation rate data is a case in point.
When FEBP launched its higher education database in mid-2011, we gathered state-level data on a host of indicators from the IPEDS State Data Center. This included average tuition and fees, four-year and two-year graduation rates, and student demographics. At the time, data were only available from the 2007-08 school year. Everything seemed on the up-and-up.
Fast forward to a couple weeks ago when we discovered that the State Data Center had made data for the subsequent year (2008-09) available. We downloaded the new state-level data and merged it with the 2007-08 data.
While most of the data looked fine, we discovered that the 2009 data for four-year graduation rates were dramatically different than the 2008 data. In some cases, graduation rates in 2009 were half of what they were in 2008. Though graduation rates do fluctuate from year to year, this sort of variation seemed highly unlikely. Unfortunately, the 2008 data are no longer available on the IPEDS State Data Center, making it impossible for us to do a fresh comparison.
Suspicious of some kind of calculation error and curious about the lack of availability of the 2008 data, we called the IPEDS helpline. We were informed that IPEDS is still figuring out the capacity of the State Data Center and what role they want it to play in the greater database. As a result, the 2008 state-level data are no longer available for download and are not stored anywhere offline at IPEDS. Further, IPEDs could not tell us exactly how the graduation rate data were calculated for 2008 or 2009, so we don’t know what might have caused the discrepancies.
Given this, it was clear to us that we could not include the state graduation rate data in our database as they were collected from IPEDS. Instead, we used the IPEDS institution-level data system to calculate group averages for each state for both total graduation rate and four-year graduation rate for 2008 and 2009, an iterative and long process. Though these data are not automatically available from IPEDS, we can ensure that they were calculated in the same manner and are therefore comparable across years. These are the data now displayed for both 2008 and 2009 in the FEBP database.
The IPEDS State Data Center could be a powerful tool for stakeholders attempting to understand how higher education systems and outcomes vary across states. But as long as IPEDS treats the center as a second class project, it will not reach its full potential. Key to this process will be making multiple years of data available and ensuring that data for each year are calculated in the same manner. Hopefully IPEDS will have the motivation and capacity to head in this direction. If not, we will continue to be concerned about the reliability and comparability of the data available at the State Data Center.