What to do about the undecided?

There has been a fabulous discussion on LinkedIn (which of course now I can't locate) about the proportion of students at our various institutions who are undeclared or undecided about their education goals when they join us.

One dilemma is what to do about their data for our reporting and analysis of educational intent.  In the current higher education environment at the federal and state level, much of the discussion is about success in the form of degree completion.  This predicates the collection of a stated educational goal, or intent, from the student at their entry so that we can then measure whether or not we supported them in achieving it.

Another issue is the integrity of our data.  We wonder if some students are just being entered as undecided if they don’t immediately self-identify with a specific program or major goal.

Our third challenge is with blanks.  As we say around here, blank is not data.  Was the field skipped?  Was the student undecided?  Is it an error in data retrieval?  Even worse are fields with only a space.  They look like blanks but read like data to our system.

Those are the nerd issues.  Then there are the academic ones.  Should we require the students to make a decision/commitment?  Some research indicates this might help retention.   

What if they are genuinely undecided?  Can we leave a space for that exploration in an environment in which our ability to serve them rests on how many we can get to degree or certificate completion?  What is best for the student in the long and short term?

We are still mulling these questions…comments are welcome.

Decision Agility and the Dream of Big Data

Don’t get me wrong…I love data like only geeks can.  But even I get cross-eyed at some of the proposals going around now about accountability and benchmarks and comparison tools.

Imagine with me, please, what our enterprise data systems would look like if we were able to gather, store, organize, and retrieve all the data that our many analysts inside and outside the academy have suggested.  Never mind the significant state and federal data collection efforts. 

I do dream of huge multidimensional cubes of data waiting to be mined for nuggets.  I salivate like Pavlov’s anticipatory canines at the possibility of predictive modeling using all possible variables.   

Yet if we actually had all that data, would we really be able to use it to make sensible recommendations on a reasonable timeline?

Alas, probably not.  In the current web-mediated world, information is plentiful.  How much of it can we absorb, utilize, or make sense out of?  

Collecting the data is only the first step of a full scale process.  Data have to be cleaned, organized, and presented in a format and fashion understandable to the audience. All this is complicated by an overabundance of data.  It requires humans with the training and talent to choose and deploy data to maximum effectiveness.

The result is that we become slower to process our information, slower to make it into usable data, and slower to interpret the streams of data now at our disposal.  This is no service to the academy.   

Wise decisions rest upon data gold, but if we spend our limited resources on gathering every straw, spinning it into actionable form may suffer from a need for magical rather than procedural solutions.