STEM Stupidity

Seven Day. Week Out. And frost once more upon the ground. Oh!, the joys of Climate Change and Global Warming. One of which is the ignorant shoutings of Repulsian stalwarts that such incidents definitively prove the absence of their effector. Never mind that these people are math blind and have no capacity to comprehend that the standard deviation of the temperature distribution is proportional to the mean temperature, they have to be right because they “feel” it to be right. Deity bless Amerika where Ignorance is king and knowledge is sinful.

On which note I feel the need to let loose some blot on a spate of STEM articles I have accumulated this week. First, an article [Link] entitled “Are Men Smarter at Science Than Women? Men Certainly Think So, Study Shows.” This article from the New Yawk Times proclaims that 

“Men were more than three times as likely as women to say they were smarter than the classmate with whom they worked most closely.”

The problem with the article, for once, isn’t the journalists. It’s the researchers.

Simply put, in the last twenty years I have read twenty studies that report the same results as here. Admittedly these results didn’t come out of a physiology department and that’s the damnation. They all came out of Management “Science” departments (or similar, including psychology) and they all informed us that men have bigger egos and more arrogance than women. That part was consistent, the primary variable was the nature of the workplace ranging from service jobs through executive suites.

So the problem here is one of lack of integrity and honesty on the part of the researchers. What isn’t clear is whether they failed to tell the journalists that their study was derivative and confirmatory or they failed to do an adequate literature search. Either way, they’re damned. 

Sadly, it’s also a symptom of our failure as a society. Because some components of society have decided they are underrepresented in an activity, the cure is to mandate leveling. The problem with this is it destroys creativity and internal self-respect. Yes, non-pink and non-male representation in many “professions” is different from the population as a whole. But no validated evidence has been presented to indicate that is detrimental. And organizational mandate to level has divided the workplace socially to the point of severe decrepitude. 

As a manager I recognize that diversity is good and increases the strength of the organization, but when I am making a personnel choice and I am told that I have to select X characteristics specifically, I am unable to acquire that strength. 

Are there managers who don’t do things honestly? Yes. Most of them. But they also cheat on the mandates as well. WHich makes the problem worse. 

This is a rotten tooth. The more it is prodded, the worse it gets.

Next, an article [Link] entitled “Lesson learned? Massive study finds lectures still dominate STEM ed.” This article claims that lectures are the dominant vector of STEM coursework and that they don’t engage students.

Again, not a new thing. I have read about as many papers on this in the last twenty years as I have the previous. And they all decry the ineffectiveness of lectures.

The question however, is what is the measure of effectiveness? Over the years I have spent some time in the classroom and I have discovered several things that aren’t talked about in these studies.

Most of the data is collected from “service courses”. Those are courses like the widely dis-loved “freshman physics” courses. They are taught in large auditoriums with hundreds and thousands of students and one lecturer/teacher and a few TAs. Maybe. These courses are a love-hate albatross for their departments. First of all, they are the primary source of existence and educational funding for the department because the (vast) majority of the students in the courses are there to satisfy a general degree requirements. Students who will major or minor in the discipline are a vanishing minority.

But these latter students are the ones who are really important to the department. For what should be obvious reasons. They are the students who, in the main, will be the STEMs of that discipline. 

The other students are transients. Many of them – more than in my day – are there only for ticket punch. They cannot be engaged. No matter what. 

We also have to consider the demographics of STEMs. STEMs are more likely to be INTRO than EXTRO. INTROs and EXTROs engage radically differently. So which way do we structure the course to engage? I won;t even belabor the other problems, like lecturers who are overworked and under-timed. The point remains that studies like these are almost always EXTRO Supremist, which is recognized by the actual practitioners as extinctionist. 

The lecture format isn’t perfect but it represents a compromise, perhaps the best one, between resources and student needs. 

Third, an article [Link] entitled “The Scientific Paper Is Obsolete.” The author argues that the scientific paper is obsolete because of computer software. And, no, this isn’t one of those let’s publish results on social media things. The argument is that the software is compromising the transmission of the information.

Not surprising.

In Galileo’s day, the same problem existed except that the maths of the day – infinitesimals – were contrary to the delusions of the religion organization. In Newton’s day the Royal Society attacked the problem by making people read their papers aloud to the meetings of the society. 

IOW, not a new problem. 

When I was a grad student, computers were not the desktop fixture of today. In those days we had to go to temples and present offerings to the computer-deity. And only the folks who knew coding AND computer maths could do so well. 

At one time I was writing code for most of the faculty in the physics department. And having to write the computer section of their papers. 

Research is supposed to be new. It isn’t scholarship. And yes, there were problems with repeatibility in those days. That’s why we published the code, so people could run two pieces of code in parallel and see how they crunched differently.

One of the problems is that of Babel. Too many languages. I write code in FORTRAN or BASIC, depending on the level of detail to the calculations. If I can do it in a spreadsheet, BASIC, housemother FORTRAN. Other folk write in other languages. And usually we can’t talk to each other.

What gets missed here is that as bad as NERD papers are, they do work. And no other method has.

When societies fail it is usually because “make it better” overwhelms “get it done.”