Big Data in Higher Education

Big Data in Higher EducationBig data in higher education is one topic I will finally admit that I just can’t get excited about. Mostly because I question the existence of big data in higher education and the collective willingness to take any action on it.

Like so many overly hyped technology topics, big data in higher education seems more like a solution looking for a problem. Or should I say a tech industry looking for next quarter’s numbers.

UPDATE (9/12/12):

IDC just released its most recent software sales survey results which predicts a 9.8% decline for business analytics software that will still be a $50.7 billion market that is up from $37.1 billion. IDC attributes this growth to the media hype surrounding Big Data.

Definition of Big Data

The definition of big data generally refers to data sets that are so large or complex that they become difficult or nearly impossible to handle. Others describe a definition of big data as the condition when the data has grown so large that the existing software and systems can no longer effectively process it.

In more practical terms. Big data can be a large number of small data files, a massive number of single data elements, a massive database, or even piles of data in disparate systems. Big data isn’t simply a matter of size since big data can also be the result of a high volume or high velocity data stream.

Top 5 Myths About Bid Data

I thought I would share the list of the Top 5 Myths About Big Data from a recent Mashable article as part of helping bring some clarity to the whole big data in higher education quandary I find myself in.

  • Big Data is Only About Massive Data Volume
  • Big Data Means Hadoop
  • Big Data Means Unstructured Data
  • Big Data is for Social Media Feeds and Sentiment Analysis
  • NoSQL means No SQL

You should read the explanations for each of these myths about big data if they are not apparent.

Big Data in Higher Education

Now that I have acknowledged the common big data myths, allow me to ask if most colleges or universities even have big data. I am not talking about large data sets being crunched by researchers on systems they support.

I am asking if big data in higher education academic and administrative systems exist. I am asking if there is a significant percentage of the ~4500 institutions in the US have a big data problem.

In many ways some of the stuff I read about on big data in higher education is like an episode of the X-Files except Scully and Mulder are played by your favorite storage and BI sales reps.

In terms of size, how many institutions have an SIS/ERP with more than 100GB covering 25 years, or even 500GB. And, just because you can’t get the data you want out of your SIS doesn’t mean you have big data.

How many are carrying more than a semester or two in their LMS even approaching with data approaching 100GB (not counting static content).

What? So What? Now What?

The real issue for me is born out by a simple decision aid I have used for many years:

What? So What? Now What?

When considering the possibility of big data in higher education I attempt to answer those three questions.

  • What is the big data in higher education that we might care about in terms of sources, types and relevance.
  • So What if there is Big Data. What problem does it solve or maybe more to the point is their a collective willingness to solve the problem?
  • Now What do we do if we can crunch the big data and feed it into a decision support process.

Don’t Ask If You Don’t Want to Know

This is really a bit more of the So What question. Let’s say we solved the big data challenge and the complexities of multivariate analysis of all SIS/ERP and LMS data and more.

Now we know with certainty the likelihood of success for an incoming freshman with certain risk factors. And we know with certainty the failure rate of students in the advanced course work based on which faculty taught certain foundation courses.

  • Don’t the central state system databases already contain these answers for the publics?
  • Don’t we already know with some degree of confidence some techniques that are good enough for boosting STEM success?
  • Don’t we already know if pre-program strategies result in program enrollments and graduations.
  • Don’t we already know if a dollar invested first year success programs has more success than just adding more instructional support?

Couldn’t you figure that out today with the tools you already have? I know most people don’t touch the data in their LMS and still struggle with the SIS/ERP reporting.

Of course the real question is does anyone really want to know? Why hasn’t anyone ever asked?

In any other industry you strive to understand the performance of the value chain and each component or process. Even in health care we track infection and mortality rates to individual clinicians, staff and clinics.

So, if you could get past the (insert air quotes here) hurdle of big data in higher education what would you do with it. Is it actionable? Would it offer any new insights or understanding on the top issues affecting higher education today?

Well for me I need to ruminate on this a bit more because I am just not convinced there is a problem to be solved in terms of the data itself or the need for information to support decision making. Until I see a convincing argument to the contrary I will hold the belief this is typical vendor hype bolstered by CIO’s who have taken the bait and can’t spit the hook out.

This entry was posted in IT Strategy and tagged , , , . Bookmark the permalink.

3 Responses to Big Data in Higher Education

  1. Jimmy says:

    Some schools are doing great things with big data. In some ways, its refreshing to see schools think more like businesses. As long as its being collected, schools might as well make use of it (as long as it doesn’t violate students’ privacy).

    ASU in particular is doing some very cool things with it:

  2. The Higher Ed CIO says:

    I read the link (some links on that page are broken) which I appreciate but still challenge the premise of their data being ‘big’. Software for assessing students at risk or degree auditing applications are not new and don’t handle big data. Similarly course scheduling systems can handle numerous variables but most often do not reach their full potential for minimizations/maximizations because faculty only work certain days/hours and rooms are restricted for departments and programs.

    So the article sounds good but what is really going on.

    But the real point is still if the big data analysis would even be actionable. Admissions selectively will likely never be fully automated so the big data benefits may never be realized there especially when demand outstrips supply for most of higher education today and limited accountability.

    As to privacy, you may want to stay tuned this next week for some real interesting stuff there….

  3. Pingback: Big Data or Big ProblemsThe Higher Ed CIO

Comments are closed.