Bringing out your dead data: Data management challenges in higher education


Do you have “dead data”?  Data that is not relevant or trustworthy? Data that muddles reports and makes data-driven decision making difficult, or even impossible, to manage? If you do, you are in good company as this is an enormous problem throughout higher education. Not to mention, as with all dead things, it can be very hard to fix.

It is often very clear when reports and critical information are being generated from data that has long since lost its life. People treat these dead artifacts in a variety of ways but seldom actually believe in their current value. Very capable, even heroic, people are regularly performing data manipulations and surgery to build and deliver in-time or best-estimate reports for users and executives.  But let’s be honest. Often, even when new and continuing reports are delivered, they are DOA (dead on analysis).

When there is dead and dying data in the mix, the living are threatened.  It would be most beneficial if dead data would just stay dead. When dead data is reanimated and becomes zombie data, the institutional brain is eaten away from the inside until outward signs, such as bad reporting to critical places, starts to ooze out.

Where can you find dead data?

Dead data can be found throughout the processes and databases of a higher education institution. For instance, in the recruitment process, students are tracked from their initial interest to the sending of their acceptance letter to their acceptance of the offer to the down payment on tuition. During this process, markers will be added that describe what stage in the admissions lifecycle a student has reached but as soon as the check arrives the tracking process stops and wherever the student happened to be labeled, that is where they stayed in the data... often times remaining listed as an applicant and not a student.

Some signs that you might have dead –  or walking dead – data include:

  1. When receiving a report, do you immediately ask, "Can I trust all or any part of this for decision-making purposes" or "Can I ask a follow up question or should I just leave this alone?"
  2. When you ask for a report or analysis, does it always require two weeks?
  3. Your current reporting and analytical processes killing your soul.
  4. When asked to fund data initiatives you scream  and run away.

If any of these signs apply to you, then you are probably dealing with dead data. But not to despair. All is not lost.

How to deal with dead data

As in most cases when dealing with the dead, it is best to focus first and foremost on the living. The key to producing high quality, “live” data and information is understanding your institution’s core information requirement and making an active assessment of data health. Just leave the dead where it happens to be and openly, honestly start with all the clearly alive data, then the mostly alive, and work back toward the possibly alive.

In all the years I have been hunting down dead data, I have relied on SAS to be my baseball bat for data integration, data quality, master data management and enterprise data access. These days, the tools I need to beat the dead data back to a manageable place are found in SAS Data Management.

Many of the schools I have worked with have found that by using SAS Data Management, in tandem with good data management policies and techniques, they have been able cut their way through to  a rich and lively data and information-based culture and have ultimately arisen from the graveyard.

Learn more about how SAS Data Management can minimize risk, and how proper planning for data management and usage could have you exclaiming, “It’s Alive!” when you review your reports or need to make timely decisions in the future.


About Author

Eric Donohue

Senior Business Specialist, SAS Education Practice

Eric Donohue is a Senior Business Consultant in the Education Practice at SAS. Before joining SAS in 2006, he was the driving force in the planning and implementation of the University of Washington’s Enterprise Data Warehousing Program. With over 20 years of experience in management, consulting, systems analysis and information technology, Eric firmly believes that reality-based decision-making should be employed at all times.

Comments are closed.

Back to Top