Thursday, February 13, 2014

Lies, Damn Lies, and Statistics

I don't have a NGLC grant from the Gates Foundation.  I probably never will.  My poor luck.  But it does afford one advantage.  It is far easier to be critical of what they are doing. 

Here is a case in point.  I have started to read the book, Blended Learning: Research Perspectives, Volume 2, edited by Picciano, Dziuban, and Graham.  My purpose is to see if I can identify interesting points from this book to use for Blended Learning research at the Business School at Illinois.  But I'm readily distracted as I was not part of the data collection effort and still haven't seen the data to which this earlier research might speak.   

I am on chapter 3, written by Patsy Moskal and Thomas Cavanagh.  It is about a large scale project to take the blended learning approach pioneered at University of Central Florida and transfer it to other AACSU member institutions that are participating in the project.  In concept this seems like both a sensible thing to do and an interesting thing to evaluate.

Moskal and Cavanagh do a reasonable job at creating a profile for typical students at such institutions.  Most are non-traditional - they have a family, are working, and are older.  Many are low income.  Many are the first in their families to attend college.  Moskal and Cavanagh also provide some data about the completion rate to degree.  It is low. These students face enormous challenges.

So far so good, but now the problem that raised the hairs on the back of my neck.  Near the beginning of the chapter the authors provide some information about the economic benefits from having a college degree.  They quote directly from a Gates Foundation document (page three top of the middle column):

In 2008, the average wage for adults 25 and older with a four-year degree was $60,954, compared to $33,618 for those with only a high school diploma and $24,686 for those with no high school diploma. 

The problem with this, a big one in my view, is that the quote pertains to the entire U.S. population, not just the population of students profiled by Moskal and Cavanagh. It likely substantially overstates the gains to a degree for the profiled students. 

In today's New York Times there is an editorial Making College Pay.  Whether you agree with it's conclusions or not, it is far better in talking about the data, for example by making a clear distinction between the average sort of statistics, reported above, and the picture for recent grads, where the employment situation is far grimmer. 

So the question emerges while reading Moskal and Cavanagh, are there data about the employment of grads from AACSU member institutions?  If so, why isn't that reported in lieu of the Gates stuff or in addition to it?

And if no good data of this sort exists, why include the quote from the Gates Foundation materials?  I can guess at the answer to that question, but not in a way to come to a benign conclusion.   Someone needs to tell the Gates Foundation and participants alike that including quotes like that tends to discredit the results. 

Is it really necessary to do this to get such projects going? 

No comments: