information system case study

Northeastern University (NEU) is now ranked among the top 100 research institutions in the United States.  This was not always the case.  As part of a methodical process and thanks to a significant investment of time, energy and money, the University reengineered its curriculum, improved the quality of its teaching and research faculty, built up its campus facilities, and rebranded itself.   Along the way, the institution could have misdirected its efforts and resources.  Fortunately, Northeastern’s leadership recognized the value of data-driven decision making.  Early on in the remaking of the University, NEU established a decision support system (DSS) to inform planning.    For example, in the early 1990’s, NEU stood firmly in the third tier of ranked U.S. Colleges and Universities, due in no small part to a student retention rate near 35%.  The University was determined to improve on this vital metric (today retention is closer to 85%), but to that end NEU needed to determine what factors/investments might substantial improve student retention rates.  By mining years of student cohort data, the University’s leadership determined those factors that influenced the student’s ability to stay at NEU through graduation.
The enabling information system, (i.e. the University’s first decision support system - DSS) drew on key University transaction systems for its data, including: the Registrar System, the Financial Aid System, the Human Resource System, and the Financial Management System.  Led by a technical team draw from the University’s Information Technology Services (ITS) Unit, a large cross-functional committee of stakeholders, including academic and administrative leaders from across campus, identified the data to be collected by the DSS and devised a common set of definitions, metadata standards and the like to ensure the quality and integrity of the data collection/translation process.  The IT team employed extract–transform-and-load (ETL) tools to draw the data from the aforementioned source systems and to load it into a specially designed data warehouse.  Rather than try to recreate data sets historically, the stakeholder committee agreed to start the DSS process and data collection with the most current body of transaction data, beginning with the Fall term of the 2002-2003 academic year.  The data warehouse now holds nearly nine full academic years of longitudinal data, including 36 data discrete sets as of Summer II term, 2010.
Once the data warehouse was fully deployed and once the automated procedures were in place to extract key data from the school’s various systems of record, the data warehouse was ready for use in decision support.   A team of analysts, working under the general direction of the Senior Vice President for Planning, mined this DSS data source looking for the underlying predictors of student retention and academic performance, tends in course registration and faculty and facilities utilization, demographic shifts in the student body, and many other high-level tactical and strategic issues facing the University’s leadership.  As the process added more academic terms of data to the data warehouse, the opportunities for richer and more focuses analysis became available to the DSS team.
To date the Northeastern University DSS has been used to inform the institutions policies and processes for delivering student services, to direct its building programs choices – especially regarding the construction of new student residents’ halls, and to focus faculty hiring and academic program development.  In turn these decisions have improved overall student satisfaction and retention, and school performance, vaulting NEU into the elite 100.

Question to Address:
What technical components comprised NEU’s decision support system?
Who were the key stakeholders in the process and why were they so important?
To what would you attribute the DSS project’s success and how would you measure and assess outcomes?
Without a DSS, how would NEU inform its decision making?  What are the shortfalls in these alternative approaches to decision making?