Blogs + Podcasts

Home > Blogs + Podcasts > The Teagle Liblog

The Teagle Liblog

Return to List
March 14, 2007

Kids Die
By Bob Connor - The Teagle Foundation

podcast

I was flying to New York for a meeting of the Teagle board of directors, wondering what I should say to them that wasn’t just the same old stuff, totally boring. In frustration I reached into my brief case and pulled out an old The New Yorker article that Dick Hersh had sent to me. Why he did so was a mystery. The article had nothing to do with our mutual interest in liberal education. It was about kids dying from cystic fibrosis. A terrible subject but it got my mind off the impending report to the board.

About half way through the article, the light came on. The article is called “The Bell Curve” because centers treating cystic fibrosis are distributed on a bell curve when you measure their outcomes. In 1997 “patients at an average center lived to be just over 30 years old; patients at the top center typically lived to be 46.” If you bring your kid to a nice, well run, well regarded center at the middle of the bell curve, rather than one at the far right end of the curve, you know what that means. She dies earlier. Or maybe you don’t know, because the center has no incentive to tell you. No incentive, that is, except concern for the patient. Kids die. Some die sooner than others. Some places care. Some places don’t.

But if your institution cared, what would it do? The far right lip of the bell doesn’t coincide with reputation or measures such as where the doctors got their degrees, or whether they are using the most expensive equipment. Maybe it has something to do with fanaticism, uncompromising determination to do whatever it takes to get that patient to survive as long and as well as possible. We may not know exactly what accounts for the difference between “average” and “top”, but it appears that in medicine we do know two prerequisites for moving a medical center towards the right end of the bell curve. Let me quote the hero of the The New Yorker story, a physician named Don Berwick, who runs a non profit in Boston called the Institute for Health Care Improvement.

“To fix medicine,” Berwick maintained, “we need to do two things…measure outcomes, and be more open about what we are doing.”

“Go naked” as one doctor put it. And it turns out that some centers have done just that, made the data about their outcomes totally public. That left them with no alternative but to do everything they could imagine, as fast as they possibly could, to help those kids survive.

So is there any analogy to higher education? Centers treating cystic fibrosis have crystal clear outcomes. Bodies live; bodies die. Count them. If we could get some comparable measure for what happens in higher education, would we find another bell curve? Would it coincide with the pecking order reported each year in US News? More important: could an institution that really cared, one that thought the mind was as important as the body, move from the middle of the bell curve to the far right lip? Could it change from doing a perfectly good, “average” job to “top”? And to do so, what would it take? “Going naked?”

When I put the article down, I realized how much I wanted to know the answers to these questions, and how we could best help those places that really cared. There was my “report” to the Board. That’s what we have to figure out.

You have to read the The New Yorker article. It’s by Atul Gawande and it’s called “The Bell Curve” in the December 4, 2004 issue.

Email This Article | Subscribe to E-Updates

User Comments		Post Your Comments

It’s refreshing to see the question being posed in concrete terms in these days when (as Bill Readings has trenchantly observed) trumpetings of “excellence,” whatever that is supposed to mean, are ubiquitous in college catalogues and websites. In the medical context Gawande describes so eloquently, everyone agrees on the outcome. The person with CF is sick; the aim of the center is to keep them alive. But of course a major part of the challenge of measuring outcomes in the context of higher education is that all interested parties may not agree on what the desired outcome actually is. This is a question of enormous scope, so let me cut it down to the problem of aligning “our” criteria—what we faculty and administrators care about—with those of the students and their families. So long as students enter college with ill-formed or misguided conceptions of why they are there, how can we hope to measure the effect of “what happens” to them? I’m talking about the whole range of students, from those who simply don’t know what a liberal education is supposed to mean to those who dismiss the study of literature and culture as a boring obstacle to their acquisition of their pre-professional scout badge (i.e. the diploma). Is there a coherent and persuasive way to frame the “outcome” of higher education without stumbling into bland formulations like “critical thinking” (which, as bright economics and science students are always pointing out to me, isn’t usually the highlighted priority in a standard English or art history class)? And how may that framing enter into public thinking about higher education?
March 14, 2007 – Joy Connolly - New York University

I enjoyed reading your essay "KIDS DIE" and I re-read Gawande's article in the New Yorker titled: "The Bell Curve." I actually read it with interest several months ago. I also know and have worked some with Don Berwick, the "hero" in your essay, who is one of the major advocates of continuous quality improvement in health care. That effort is but one of many contemporary extensions of Deming's work on Total Quality Management (TQM) - See html # 1 below.

You are in a better position than I am to judge, but I assume that measures of outcomes of a college education, which we all agree are important, either are or could be demonstrated to distribute along similar "bell shaped curves." After all, what doesn't?

In my world of health care there are several uses for such bell-shaped curves that describe outcomes. First, the width of the curve illustrates variation and one goal of improvement can be to reduce variation. A second goal is to track your own curve over time (so-called control charts) to measure change. And third, the mean or average of a distribution curve can be used to compare the outcomes of one organization to another, and by then looking at what makes the best better you can identify not only best practices but also targets for process improvement.

The starting point in all such efforts is a philosophy at the top of any organization devoted to improvement. The models for implementing that philosophy are generically straight-forward, published, and have been applied to some extent in a host of different organizations from automobiles, to Motorola, to health care, and to education. So there is a history worth learning about. The focus is on improvement of processes that contribute to the desired outcomes.

Below are 2-3 helpful web sites I found. Berwick is a great speaker and broadly knowledgeable inside and outside health care. I also learned that Donna Shalala at U. of Miami has thought about this in higher education.

http://en.wikipedia.org/wiki/W._Edwards_Deming
http://www.ihi.org/IHI/Topics/Improvement/
http://www.greenwood.com/catalog/C8389.aspx

March 14, 2007 – Andrew G. Wallace - Dartmouth College

I am pleased that the article I sent you had the same impact on you that it had for me—indeed it impelled me to write my November 2005 Atlantic Monthly article “What Do Colleges Teach?” In that piece I explain the rationale behind the Collegiate Learning Assessment (CLA) project, to wit: to stimulate “going naked”—measuring outcomes, being transparent about what we are doing in higher education, and by doing so leaving ourselves no alternative but to do everything imaginable to help students learn at the highest levels.

You ask if there is an analogy between medicine and higher education. Absolutely there is and both are matters of life and death. To emulate what medicine is doing all one has to do is gather some comparable learning data across colleges and universities. That is exactly what the CLA measures of critical thinking, analytical reasoning, problem-solving, and writing are intended to do. In the results to date, with a nationally representative sample of over 200 colleges and universities, we find some answers to your questions:

1. Would we find another bell curve? Yes, that is exactly what we have found to date.
2. Would it coincide with the pecking order reported each year in US News? No. We find no evidence that there is any relationship to the usual pecking order.
3. Can an institution move from the middle to the far right of the bell curve? Only the results of our longitudinal study can tell us that but the presumption is YES. What CLA does is provide one useful metric of important learning outcomes that allows for comparability BUT also allows an institution to measure itself against itself and thus see change over time since a longitudinal study allows for baseline and change data.
4. What would it take to move along the curve? We have found that 20% of campuses provide significantly greater value-added than comparable campuses, e.g. similar size, admissions, faculty, etc. By constructing careful case studies we plan to ask what is taking place on those leading value-added campuses in terms of cultural ethos, pedagogy and curricula that might account for their greater efficacy. And with a useful learning assessment metric, each institution can make changes in pedagogy and curriculum, for example, and see if it causes greater outcomes.


So, whereas in earlier years a few people “streaking” across campus was all the rage but probably did not influence learning, we now may have to face the prospect of entire campuses going naked.
March 19, 2007 – Richard H. Hersh - Co-Director, Collegiate Learning Assessment project

I am struck chiefly by the sentence in which you identify what is to me the real point of the analogy. Do academics work to improve souls as passionately as the culture at large works to maintain bodies? The answer is no. But the reason is not that academics don’t care about souls. The reason is that they don’t envision their academic writing and teaching as “soul-craft.” As against the moderns, the ancient political philosophers agreed, one and all, that politics was or consisted largely in providing for citizens an education in virtue. Statecraft was a form of soul-craft dedicated to enabling citizens to make healthy choices, psychologically and otherwise, and thus to live well. As all ancient political thinkers recognized, providing for the health of bodies was a necessary precondition for such educational projects. But maintaining physical health was subordinate to fostering the development of virtues of intellect and character. It would require a much longer conversation to say whether (and how and to what extent) higher education ought to concern itself with the virtues of character. It should be uncontroversial, however, to say that higher education both is and should be concerned with developing the intellectual virtues. To my knowledge, we in higher education do not currently have a broadly acceptable language to describe such virtues across all the disciplines: the intellectual virtues that we would admire in a human being as such, not the human being as historian or scientist or litterateur. Yet the ancients – particularly Aristotle – provide an excellent starting-point for constructing such a language. Consider that contemporary political philosophers have attempted to resuscitate the Aristotelian virtues of character – in a suitably updated form – in support of cosmopolitan and universal political values. Why should we not attempt the same thing with respect to the Aristotelian intellectual virtues – if not in such a broadly cosmopolitan way, then at least in a sufficiently general way to enable us to articulate what we value intellectually about human beings as such? What virtues of the intellect should we be working to foster? Publicity, transparency, and so forth sound like excellent ideas down the line, but I believe that, before making our outcomes public, we need to know first and foremost what we should be “going naked” about.
March 22, 2007 – Ryan Balot - Associate Professor, Department of Political Science, University of Toronto
Post Your Comments