Methodology

In 2020, James Madison University contracted with Equifax to develop an analysis of JMU alumni graduate outcomes. JMU provided Equifax with information on graduates from 2005 to 2019 (approximately 70,000 records). Data were provided on a number of graduate attributes, such as the type and level of degree earned, as well as demographic variables such gender and race. Equifax matched the submitted records against its data repository of over 300 million payroll files and, when eligible data were located, returned aggregated information to JMU on the following fields:

Income: Income was provided to JMU for “Pay eligible employees” only. A pay eligible employee was one that was employed full-time (employed for at least 30 days, worked an average of at least 28 hours/week, and made more than $8,000 annually). Income was adjusted using the Chained Consumer Price Index for All Urban Consumers (C-CPI-U), normalized to current US dollars.

Industry: Industry results were returned at the Supersector level by NAICS Code - https://www.bls.gov/iag/tgs/iag_index_naics.htm.

Geography: Location of employment was returned by State, Region, and Division following the Census Regions and Divisions of the United States.

Data Security

Although the source data was based on individuals, the reports provided to JMU by Equifax were only in the aggregate, based on the attributes defined by JMU. Further, the information was completely anonymous. Equifax only provided aggregated “slices” of data (groups of individuals by reporting attributes) with 10 or more matched records. The identity of JMU alumni was protected at all times throughout this project. Data were sent securely via Secure File Transfer Protocol using PGP Encryption. At no time did JMU have access to individual payroll or earnings data on its alumni.

Because of the high level of security required for this project, certain graduate attributes yielded more results than others. As you navigate the tools developed with the Equifax data, certain majors or degree levels (e.g., Doctoral degree, Master’s degree, Bachelor’s degree) may not appear. In these instances, Equifax did not match enough records to provide that “slice” of data. Further, in many instances, JMU required a higher degree of matched records to ensure that the sample of data obtained was reasonably generalizable to the greater population of alumni. Therefore, in many instances, at least thirty records from JMU had to be matched by Equifax in order for JMU to include the data point in the tool. For the mapping tool, at least 10 records were required to share locations where JMU graduates worked. The overall match rate of records for this project was 70 percent, meaning that at least partial data was returned for 70% of the records submitted.

For questions about this tool, please contact Chris Orem, Director of Institutional Research (oremcd@jmu.edu).