Blog

Deming in Higher Education, Part II

An Opposing View

A few years ago, I published an article in the Industrial Physicist Magazine titled, "Physicist Transformed the Quality of Management." It was an introduction to the late Dr. W. Edwards Deming's "system of profound knowledge." The article generated some letters to the editor, and I was asked to respond to them in subsequent issues of the journal. (For a couple of those responses, see blog titled "Deming in Higher Education," posted on this web site in December 2013.) After several very positive responses, a letter arrived that expressed strong opposition to my article, responses and Dr. Deming's principles. The letter read, in part,

"I was surprised by your grading advice in letters to the editor in December's Industrial Physicist. Shouldn't this have been saved for the April Fool's Day issue?

"I work in a high tech profession (Space Suits) where I see increasing numbers of poorly qualified, poorly motivated new graduates who cannot think through problems by themselves. Industries' response to this pool of mediocre talent is to lower its expectations, and treat these technical people as an expendable resource, certainly not to be rewarded with the compensation and career opportunities previous generations of engineers and scientists enjoyed...

"Universities seem to try hard to indoctrinate their students in the popular concepts of team work (considered important in these days of lazy, ill prepared employees) and promulgate the politically correct notion that criticism of personal performance is bad. This is the message I get from your comments regarding Deming's philosophy.

"I contrast this with the scenario when I went to college some 30 years ago.

"One quarter of my freshman engineering class didn't come back for the second semester. Half weren't back for the sophomore year. Most of the absent had come to the conclusion that study was too hard and they either could not or were not willing to do the solitary study. Of the over 200 who started with me in Aerospace Engineering, only 16 finished with a diploma in that field...

"One of my calculus professors was nicknamed cube root Kent because he was said to pass only the cube root of his class. Of course, this was not literally true; he only failed around 50 percent. Most of my professors graded on a bell curve, with a middle grade of 'C.'

"This system produced a generation of scientists and engineers who had self confidence born from conquering failures. We came out of school with personal initiative and determination, something today's graduates lack... I doubt Deming was correct when he maintained that learning and making a grade were very different things...

"Mr. Smith [not his real name] P.E.
Boeing Aerospace Operations
Houston, Texas"

Mr. Smith's narrative reminded me of a Boston-area CEO who disagreed with Dr. Deming at a roundtable held at MIT in the early 1980's. During the question-and-answer period that followed Deming remarks, the executive insisted that Dr. Deming was wrong to propose the elimination of traditional grading, appraisal and merit rating systems. "I can see how the systems may create fear," he said. "But to advocate the elimination of these practices -- wouldn't that be throwing out the baby with the bath water? It's the front end of our appraisal system when people are told what's expected of them; people are given their goals."

Deming's terse response will be forever burned in my memory. "That question's so lacking in logic I don't know how to begin to answer it. Next question!" Some of Mr. Smith's points were equally lacking in logic, but I did give him the courtesy of a response that included the following points (among others).

For example, he claims to belong to "a generation of engineers and scientists who had self confidence born from conquering failures." But Mr. Smith was among the 16 Aerospace Engineering graduates in his school that didn't fail. His other 184 classmates failed; and they certainly didn't conquer that failure. I can't help but wonder what contributions and technological advances we've been denied by the professors who drove scores of aspiring engineers out of the field.

Mr. Smith referred to his experience in engineering school "some 30 years ago." It was not unlike my own engineering education experience more than 40 years ago. The motto among some of my classmates was, "I must walk the straight and narrow because my major is Aero." Fear of failure was the primary motivator. Even at that young age, however, I wondered why my professors assigned grades at all, let alone assigned grades arbitrarily or "on a bell curve, with a middle grade of 'C'."

Any physics professor who issues grades violates everything he's teaching. Physics is the science of systems and the behavior of components therein. Why does the professor grade only the student? Why not grade the syllabus? Text? Lesson plan? Computer center? Why not -- God forbid! -- grade the professor? To grade the student alone confounds the student with all the other components of the teaching and learning system that influence test scores. It is contrary to the science of physics.

Mr. Smith suggests that "the popular concepts of team work" are "considered important in these days of lazy, ill prepared employees." Like his support of the draconian grading practices in his university, Smith's denigration of team work further illustrates his lack of understanding of the science of physics. When a consumer buys a car, he or she does not want an engine in that car that's made up of perfect individual components. Rather, the consumer wants an engine with components that work well together.

By the same token, shareholders and executives don't need corporations that are made up of perfect individual divisions; and a plant manager doesn't want a plant made up of perfect individual departments or employees. Rather, we need divisions, plants, departments and people that work well together. Team work is necessary for cooperation, synergy and optimization between and among departments and work groups. Dr. Deming exposed grading, appraisal and merit rating systems for what they truly are: barriers to cooperation, synergy and optimization.

The science of physics is also at the foundation of some of Deming's other radical proposals. For example, Point 11 of his 14 Points for Management reads, in part, "Eliminate numerical goals for management." He wasn't coming solely from some team-based, touchy-feely, let's-all-feel-good-about-each-other, behavioral psychological perspective. He was coming from the science of physics.

Sometimes the best thing an Operations Manager can do for the good of every customer, employee and stakeholder in an organization is to not meet this month's goals for Operations. Sometimes the best thing he can do for the good of everyone is to send some of his limited value-adding resources -- his manufacturing engineers -- to go work with the Development Engineering group. Working together, the manufacturing and development engineers could improve the process of transfer and scale-up of new designs and products. That way, products will achieve and sustain high first-pass yields a lot sooner in their design life and provide high returns in the short, medium and long term. Everyone in the organization benefits.

But in releasing his manufacturing engineers to go work with the development engineers, the Operations Manager won't have the resources to meet this month's goals for the Operations group. And sometimes that's the best thing he can do for everyone in the system.

To move to another academic discipline of importance to the engineer, any statistics professor who issues grades violates everything he's teaching. The American Statistical Association defines statistics as "the science of learning from data, and of measuring, controlling, and communicating uncertainty..." In the study of uncertainty, professors teach the Normal, Binomial, Poisson and other distributions as models of random variation.

Are the differences in the students' test scores random, or are they significantly different? Grading on the bell curve violates the principles of the Normal distribution. Half the scores will be equal to or above average -- that's normal. Half the scores will be equal to or below average -- that's normal. Test scores are either different or they're significantly different. If they're significantly different, the students should be graded differently. If the scores are not significantly different, they should not be graded differently.

This is neither brain surgery nor rocket science. One would expect an aerospace engineer to be able to figure it out. Next question!