Post-Data-Fitting Theorizing

If you see an effect in physical science, and it is surprising, you try to propose an explanation, one that fits in with other data. BUT then you do experiments and theoretical calcualations to be sure that the explanation and its mechanism are actual.

Often, in social science, people will do a fine statistical analysis of their data, often big data, and then try to explain various regression coefficients, etc. All these explanations are likely to be made up, often not linked to any well articulated theory. This need not be a problem. But if you propose an explanation, you then have to go out and test that explanation in a different context or with different data or with a finer theory that incorporates the explanation. You can't just tell a story and expect anyone to believe it, even if they find it cogent. For just telling a story does not exclude other good explanations.

So gather data, develop some theoretical explanation (one that at first might be merely a linear or some sort of regression), do your statistical analysis. If all fits, you might be surprised by some of the coefficients, and then you will need to explain them or plea fluctuation (no sin!). And then you need to theorize and test.

1. Data
2. Theory that leads to testable propositions about the data, or at least testable propositions, which might well be a set of reasonable regressors.
3. Statistics
4. Surprise?-- Yes -->2
                        No-->Develop a theory that justifies your #2 theory.-->Write the article.

T'row Da Bumps Out! --Error Bars, Fluctuation, Don't let your eyes deceive you.

When particle physicits wanted to claim that they had found the Higgs particle at the Large Hadron Collider at CERN in Geneva, Switzerland, they had to find a 5-sigma signal before it would be credible. Moreover, two different experiments had found the same result at the same mass (the mass being the "first name" of the particle).  Moreover the particle needed to have a particular angular momentum (spin) and mirror symmetry (parity), and that could only be ascertained by studying particular modes of decay, and those facts took lots more experimental data to ascertain.

Put differently, just because you see a bump, or just because you see a trend, does not mean it is significant and real. It might just be a fluctuation.

When we make claims in public policy or social science, about society, that are empirically grounded, we'll rarely get 5-sigma quality (too few observations, too little theory, too little precision). But, in general, you want to be assured that the claims make sense. Hence you must always attach error bars to your points or claims, where the bars might be 1-sigma plus or minus. Moreover, if you are claiming a trend or a shape, you need to fit the data to see if constancy and a straight line are reasonable zeroth-order assumptions. And if you are making a claim about when something began or the like, there are subtle tests of such in the statistical literature.

Moreover, Bayesian ideas should be on your mind. Even if you have rough measures and not so ideal statistics, can your measurements be seen in the light of what we take as priors and used to revise them. Often, in the policy arena, poor data may still allow you to improve practice, albeit not with the assurance you would like, but at least now you are doing better than without any data and only your presumptions and priors.

Also, never draw a line connecting points unless it is a "fit" to the data. Surely in the case of railroads you can link stations with lines since you know that trains go from A to B to C to...  And even here they may not follow straight lines between stations. However, in studying time dependent data, your straight lines presume trends when what you may have is random fluctuation.

Finally, if you want to claim changes from one time to another, be sure to normalize those changes by the standard deviations of the data, so that, again, fluctuations are more apparent. And if you plot the data and you have data that begins at say zero, you do not just show say from .5 to .6, but present it as 0 to .7, or if not put a zig-zag on the y-axis to indicate that you are skipping lots of y-axis—that is, the-y-axis begins at zero, you put in a zig-zag at say 0.1 and resume at 0.4 in the above case.

-----------------------------------

What motivated the above: A propos of yesterday’s seminar on economic conditions and social capital, I wrote this post. I enjoyed the talk, and unusually for me, I was not so much concerned with what was the punchline. It seemed clear—to provide some evidence about a common belief. Jenny Schuetz asked incisive questions about causation. The speaker responded that he was trying to find out the facts of the situation, and the connection seemed to be causal given the time frames and some of trends in the disaggregated data. I woke up this morning thinking some more. None of what I say here diminishes my interest in the seminar, but all of these things are needed to calm various objections, none of which are necessarily fatal but all of which need to be dealt with. A seminar may not be the place to make sure all is perfect, but there is no reason to leave out obvious practices even if it is “just” a talk. You want people to concentrate on your substance, not go crazy over your statistics.

Madison's Managers (Bertelli and Lynn): judgement, rationality, balance, and accountability.

I am enthusiastic about Madison's Managers by Bertelli and Lynn, Jr. What I like about it is its grounding in the law and political process, its making clear that policy school managers are very different than those from a business school, and its "transcendent" tone (albeit grounded in separation of actual powers). It gives to our students a distinctive vision, one that many already share but have not well articulated, and it gives to our faculty a purpose that is much more than providing courses and curricula. (Think of a medical school where what you want to produce are physicians, not students who happen to learn a variety of subjects and are good diagnosticians...  Or USC, where we not only produce BA holders, but also Trojans (faithful, scholarly, skillful, courageous, ambitious--with an overarching sense that you contribute to society).)

I don't think most of students would be able to read MM without lots of help. You need to know too much. But the basic ideas of judgment, balance, rationality, and accountability can be inculcated and exemplified. 

Now, perhaps all of this already happens. But in my years I am not sure I have ever heard the kind of arguments Bertelli and Lynn make, or those values, in our seminars or in hallway conversation.  "Governance" is no substitute for what MM says. 

The veterans and active duty students I see know all of this, albeit not the MM version but at least the "transcendent" values are ones they understand, no matter how much they know too much about their service. We are not just a school that trains people technically. We are not just about intersectoral, or leadership, or ... We are a school that has a distinguished mission. Our students surely know how to be USC Trojans, or at least many do. What we need to do is make them members of the Price Club.

When people argue about what was Hayek's or Marx's true position.

Thinkers can be cherry picked for any position. Sun Tzu, Clausewitz, and the Bible come to mind.

What’s interesting are the arguments Hayek provides, not Hayek as a guru. If his arguments change that may be interesting for biography, but both arguments may have value and ought be employed, perhaps each by different people.

Hence, while Bohr’s atom and quantum ideas and their beautiful pictures and sort-of vague philosophy are powerful, so are Sommerfeld’s demands for equations (which led Schroedinger to look for an equation, Debye asked him for one), and his more phenomenological analysis of spectral data (“number mysticism” it was called by some) led to Heisenberg’s matrix mechanics with its focus on only observable quantities and on combinations of allowed/forbidden atomic transitions. [Keep in mind that Heisenberg did not realize he was doing matrix operations, as such. Only soon after did Jordan inform him that he was speaking linear algebra. Heisenberg had a scheme, not so different from Sommerfeld's playing with numbers albeit more systematic, but its beautiful unification in operators and matrices came afterwards.] Each of Sommerfeld’s two viewpoints can  be quoted out of context to provide cannonfodder for some philosophical argument, but Sommerfeld's authority would seem not to be the point. By the way, S. Seth refers to Sommerfeld's physics of problems, vs. Einstein's physics of principles. Also, Sommerfeld found Bohr's models incredible and fudged, and so he separated out the technical details of the models, from their effects and focused on those effects, what he took to be quantum phenomena.

I would imagine that complex thinkers, with long histories, will have diverse viewpoints. What’s interesting is the quality of their arguments. If you want to use them as icons and gurus, you can, but the problem is that their texts are available to all. You can argue what was Hayek’s true position, but for my money I want to know about the arguments and justifications and evidence for each position. But then of course neither Hayek, Einstein, nor … are icons for me. 

They're checking your thesis/dissertation with Turnitin!

1.      Your Graduate School may randomly check theses and dissertations using Turnitin.

2.      “I am not a crook [a cheat]”, “I was taught to do it this way.” Many students do know the rules about plagiarism. But some do not. Others bring practices from professional and governmental work to university, where they are not acceptable.

3.      “I have a 4.0 GPA” This may well mean that none of their previous instructors discovered their violations of academic integrity. Hence, such a claim, when the work in front of you is inadequate, may well be a sign of undetected past violations.

4.      University regulations require that we report violations to the appropriate committee. In part this is educational, in part it is to prevent future violations. Penalties are still in the faculty member’s hands.

5.      Mosaic plagiarism is rife. Namely, a passage from a work is quoted verbatim with a change of one or two words, a reference may be given, but no quotation marks.

6.      Paraphrase demands a reference. Turnitin finds paraphrase since some unacknowledged copying of part of the passage is likely.

7.      Turnitin is effective in finding problems. You can indicate you don’t want to count stuff in quotation marks, or similar passages less than N words. You still need to examine the paper since indented quoted passages are counted as similar although indentation indicates quotation and should not be counted in the similarity score.

8.      Students may well threaten to petition to have a grade change, to get you dismissed, to accuse you of other violations. They are not bad at browbeating. My counsel is to send it to the university committee immediately and let them deal with it.

 

"Please remove my email address from your email distribution list. Thank you for your consideration in this matter."

Don't send email notes like this. Just BLOCK the sender, or send their email to your Trash.

This is from a student to a professor. Some day the student might need a favor from the professor. Why make it hard for the professor to be helpful? This person would never send a note of this sort to their boss. This person knows how to be polite and respectful.

You may be offended. You may despise someone. You may just be pissed at them.

But why leave artifacts of your sentiments when you otherwise would not express them.

We all need the kindness of strangers.

Never act out.

Big Data Needs Big Theory

The world of Big Data is here. Sophisticated statistical techniques are available to work with it. But what we also need are nice theories which allow for subtleties that might be revealed through such statistical precision. In physics, one is never doing a regression, as such. One has a rich theoretical structure, and maybe millions of more events or measurements, and the problem is a matter of decision how to make "cuts" on the data, choosing the relevant events, grounded in a theoretical physical understanding of why you are doing that. You may eventually fit a data set, but in general that is at the end of enormous amounts of theorizing and data analysis. You have a bump or a whatever, and you have a good idea of its shape and all the kinds of confounding physics below it.  If all you had was a regression, it won't be physics.

"Greg could really hold a blood grudge."

In Henry Crumpton's book on working in the CIA's Clandestine Service, he refers to one of his colleagues, saying "Greg could really hold a blood grudge." Hence, Greg was motivated, to say the least. I wonder when we such motivation in the scholarly or academic life. I am not thinking of athletics and sports, or academic politics, or mean competitiveness. Rather in just doing the work we do in our research and scholarship, do we ever have such powerful motivation? To show them, to show Nature, ... To prove to others who denied us tenure that we are ten times better than they are.

When the Used Car Salesman Deceives Herself

In an earlier post, I used George Akerlof's analysis of asymmetric information in economics, The Market for Lemons, to suggest that universities treat deans as used car salesmen, knowing more about their car for sale (the tenure candidate) than could be known by committees or provosts. Hence, you want to be sure (saleman's guarantee, your own mechanic) that what you are being sold is not a lemon.

Let us say, however, that the dean or salesman is possessed of at best imperfect information, is doing their best to  be honest, and even offers a guarantee. Still, we know that some cases prove to be lemons none the less. How do we protect the dean from herself?

You could raise the standards. Or you could have a devil's advocate (DA) on the dean's staff, or on the provost's who brings up problems and makes sure the dean addresses them. The DA is responsible to the provost or at least is not under anyone's command. The DA is the expert mechanic, the legal beagle, the person who asks the embarassing questions. Perhaps the DA has an algorithm that predicts future performance, and uses it to raise red flags.

Now, you might make mistakes of rejecting candidates or cars that will prove to be fine performers. And surely some lemons will make it through. And some apparently fine candidates will for various reasons disappoint.

But, since these are lifetime commitments, and they involve millions of dollars, and opportunity costs of not hiring a stronger performer (or a weaker one!), it may be worth these precautions.

By the way, good performers should have no trouble with this arrangement. You are appointing people for their strengths. Their weaknesses are not crucial, as long as they are not debilitating for schoalrly performance.

Don't call yourself Doctor (of Philosophy, Education,...) until you have the degree in hand.

At my university, students who are almost through with their degree requirements are allowed to march in graduation. Graduation pictures, celebration, etc... but not there yet. This does not insure they fulfil those requirements, or that they receive the degree. At MIT, I gather that those who march are checked to be sure they are actually ready to receive the degree.

Once you actually receive the degree, you can change your license plate to DR SMITH, or SMITH PHD, or... Your email address can say whatever you need to say, doctorsmithleader@gmail.com. But before then it makes no sense. It is a mockery of yourself, for until you have the degree in hand, you are a cartoon.

As for marching at graduation, despite being permitted to so march, you want to march when you are done. You want to be hooded, if this is a doctorate, when you are done. (Otherwise, you are more like the a cartoon of the Ku Klux Klan, albeit in black than in white.)

Of course, if your degree is a scholarly degree, the PhD, what counts is what you do next in terms of scholarly publishing and research. If it is a professional doctorate, EdD for example, what counts is what you do with your career. The degree is just a sign of one step on your road.

If your degree is just a sign that you could do it, and for many that seems to be the case, I have little to say.

Vindictivness and Charity

It took me a long time to achieve tenure, only in part my own fault, in part whatever else goes on in the world. I do my work since it interests me, but I believe there is always the need to prove myself. So I have some sympathy for the following situation:

Smith comes up for tenure and there is enough disagreement in the department so that tenure is not awarded. The dissenters have juggled the letter writers and have been quite verbal in their dissent. Eventually, after a national campaign that reveals the dissenters' schemes, the provost gets involved, and reverses the departmental decision, in part because Smith is at least as good as Jones who was tenured at that time. The dissenters become distraught and vindictive, and for the rest of their careers and even after, they try to help Jones and hinder Smith. Their problem is that Smith proves to be very productive, has a wide reputation for excellence, even if not among the dissenters' sub-field. In time the department becomes more like Smith, in the scholarly work and eminence, the dissenters become more isolated, so (in desperation presumably) they keep up their anti-Smith campaign, what with honors etc. But, the dissenters prove not to provide the scholarly leadership for the next generation, and they become even further bitter. Still, they try to be sure that Jones receives all the goodies before Smith.

Of course, what the dissenters should have done was to realize the game was over once the tenure was awarded by the provost, and actively promote and help Smith. Smith is now their permanent colleague, and Smith's success or failure redounds to them.  Perhaps Smith would not have been so productive if Smith did not need to keep proving that the dissenters were wrong, and so the dissenters would even have gotten their secret wish?

Academics tend to melodramatics and extremes about rigor and quality, and so they cannot admit they have been defeated. More often than not those who do not prevail find that they are further in the minority, as their department actually become stronger.

David Petraeus on Your Doctoral Project

When David Petraeus met with some doctoral students he emphasized again and again how your doctoral project should serve your long term needs. The professional doctorate encourages this. Petraeus's PhD dissertation at Princeton was about the counsel the brass gave after the Vietnam war, and the caution they expressed. He had a chance to interview top people, who would otherwise not be available to him since he was a captain? at the time. In reading his dissertation, I was struck how he had found subject matter experts to guide him and to argue with him, his greatest thank-you was to a Princeton professor who was on the other side of many issues. Be sure your chair is expert in your field. By the way, it is a good read, albeit encumbered with all the stuff that dissertations require in terms of surveys of the field. On the other hand, the footnotes are very interesting.

General Petraeus was, as I said, enthusiastic about the projects he heard about, glad they were of use to people’s actual work. In reading his dissertation, I was struck by how important it was for him to have the right advisors, early and often. They knew the fields, they knew the literature, and they could challenge him. Make sure you choose your advisor(s) with the same criteria. I mentioned this yesterday, but I was struck last night in recalling the dissertation’s acknowledgments how important that was. Of course, someone had to make sure he did all the right things bureaucratically, and I suspect that was not his advisor--  But what mattered was the substantive argument he received from his advisor.

One other thing. About methods courses. Make sure they suit the work you are going to do. People in education or psychology or perhaps economics think statistics as the methods. Or survey methods. But if your work is not quantitative, and involves other skills, maybe a course on how to write up cases, or a course on history writing, or… will serve you well.

A Truck is Coming . . . Whiskey, Tango, Foxtrot . . . Academic Roadkill: Spielberg's DUEL (1971)

From Stephen Spielberg's DUEL (1971)

I often say, in my new book THE SCHOLAR'S SURVIVAL MANUAL and in everyday discourse:

A truck is coming, and you better not say:

What truck?

Much as Alfred E. Newman, of Mad Magazine fame would say:

What, me worry?

For trucks are real and if they are coming you had better get out of the way,

Otherwise you will become Academic Roadkill.

Publish that book or get that grant or have N papers out and seen.

You had better do it.

You don't want to say,

Warshel did a half a book and it was in draft

And he got tenure.

For others will think to themselves,

Whiskey, Tango, Foxtrot,

And they will get the broom and dustpan

To clean up the road.

Maybe Warshel had other virtues, maybe he had slept with the president, maybe they made a mistake. But you are likely to be academic roadkill, be splat on the ground, unless you do what you must. You are lucky to have seen the oncoming truck. [Arieh Warshel of USC just won the Nobel Prize in Chemistry. I use his name since he is so far from being the example here.]

Stephen Spielberg's Duel (1971), one of his first films, is about a truck that keeps coming. You really don't want to be the character played by Dennis Weaver, who is being pursued by an actual Etruck.

Tough Requirements and Discipline...Ricks on Professional Military Education

Posted By Thomas E. RicksFriday, October 11, 2013 - 10:09 AM

Recently, in a note to a reader, I summarized my views of how to reform professional military education. It occurred to me that I might share them here:

• Make admission competitive.

• Make the education as academically rigorous as it was during the interwar period. (And don't train them. Educate them to think critically.)

• Make everybody write a lot and get graded on their work. As Orwell said, if you aren't writing clearly, you probably aren't thinking clearly.

• Post class rankings weekly.

• Fail at least 5 percent of the class, and dismiss from the service anyone caught plagiarizing or otherwise cheating.

• Upon graduation, publicly list graduates in order, and give the top 10 or 25 percent preference in subsequent posts.

Shameful and Disgraceful

If you are ever in a position to write a formal letter, criticizing someone for their behavior, and perhaps penalizing them, and the formal letter is likely vetted by your legal people, don't then indulge yourself with gratuitous expressions. Someone may well have proved to be unstable and unreliable, quitting on you in the middle of a semester or project, leaving you with the task of cleaning up after them, and they might well have been manipulative and demanding, but you need not say more than, "We regret your resigning mid-course. And your claims about your colleagues turn out to be not probative. We do hope your find a position more suited to your talents." Get the letter in the mail immediately--for you really do not want them to change their mind.

Never add in something like, "Your behavior is shameful and disgraceful.", even if you believe that as a colleague they have been unprofessional and an embarrassment. There's no need to say more. Wish them well, if need be throw them a big good-bye party. If someone calls for a reference, there are ways of providing something that is bulletproof yet conveys your problem: "While X was assiduous in their paperwork, we do not believe we would employ them again." Say no more.

Keep in mind that their ability to top their past idiocies is very great, so you will find they escalate their outrageous behavior, their accusations against others, their obsessive focus on innocent bystanders, and their assurances that they are right. Don't talk to them if you can possibly avoid it, see them only when others are present, and don't confuse their kindly manner with their being safe and reliable.

The Scholar's Survival Manual is now available!

The official publication date is 22 October. But I believe the book is now available more widely. It is handsome and of course as an author I am not the best judge. But I do feel good!

Martin Krieger

http://www.amazon.com/The-Scholars-Survival-Manual-Administrators/dp/0253010632/ref=sr_1_1?ie=UTF8&qid=1381409838&sr=8-1&keywords=krieger+scholars

Goldbricks and F-Offs: Do you trust your colleagues (professors) to "watch your back" and contribute to the effort.

In talking with military NCO's I am struck by one thing:

A good fraction (3/4?) of the recruits in this all-volunteer armed services are really only partly capable of being relied upon. That is, would you want them with you on a mission, expecting that they would watch your back and contribute rather than become more a burden than help. They might well be useful for driving or for less urgent tasks and missions, but when things get tough they won't deliver. (By the way, this issue strikes me as much more urgent than women in combat. I also suspect that most officers would not do well in on-the-ground missions.)

My point here is that in any bureaucracy, some fraction of your colleagues are OK, but are not to be relied on when the going gets tough: academics whose scholarly work is not really terrific, whose ability to deliver in class or in committees is not great. The net effect of this is to undermine our excellence and any claims to legitimacy of our appointment and promotion procedures. Again, these underperformers might well do a decent job, but what you need is superior reliable performance.

Too Many Theorists?

It would seem that a bunch of elementary particle physicists who are theorists are disappointed that with the opening of the new big machine in Geneva, their immediate expectations have not proven out. It may well be that further experiments in the next few years, at higher energies and intensity, will find something sufficiently peculiar to justify their theoretical work.

On the other hand, perhaps too many researchers have devoted their lives to speculation, sophisticated and illuminating and interesting, but in the end it would seem a not so fruitful path. Now, it may be that all this work will turn out to be useful for purposes not otherwise anticipated. But it has struck me that in the twentieth century at least theorists tended to be closer to available data. The justification for the very large particle accelerators is that they enable us to look closer, more intensively, at the deepest features of particles and the forces of nature. It has always proved fruitful to do this, whether we end up finding new particles and phenomena, finding stuff we know must be there but have yet to see, or finding big surprises--probably the last ones were CP-violation by Christianen, Cronin, Fitch, and Turlay in the 1960s and the J/psi in the 1970s.

What seems to have happened is that "Beyond the Standard Model" and the wonders of string theory and topological quantum field theory could not be guided by the data, but the mathematics and the beauty of some of the theories were attractive. There may well be payoffs in condensed matter physics or in mathematics. But for the moment I hear of a great letdown. There is a famous study, When Prophesy Fails. It suggests that believers may well double down on their commitments.

When the memoirs don't take responsibility...

When I read some of the memoirs of the GWBush folks, what was striking is that they seemed not to be present when he big decisions are made. That is, no one told them something important, or something just happened. (By the way, the first 2/3 of Doug Feith's War and Decision is very different.) Then I found this on Tom Ricks' blog today.



"Gold Star Father," a Marine vet who lost a Marine son in our recent wars, mentioned this in a comment the other day:

I was tootling around a Big Box Store one day with my wife. I left the book section and moved on. Wife disappeared. I back tracked to find her at the book section taking all of Bush's memoir copies off the shelf and dropping them on the floor. I panicked for a second as I figured every camera in the store probably just targeted us. But, I kinda shrugged, watched her do it, and we moved on. She told me she did the same when Rummy's memoir hit the shelves.

WSJ Article on Making Students Stronger

WSJ: 28 Sept 13: I am not sure just which of the list is right, but the general idea strikes me as correct: 

make sure people learn stuff, failure is possible, be demanding, creativity is a 

matter of accumulated experience and risk-taking innovation, grit or perseverance is

crucial, praise should be rational, based on performance, real work is always stressful.

Put differently: 

A truck is coming and if you ask,What truck?, 

                                                you will become roadkill. 

Performance matters. 

No Excuses.

 

Why Tough Teachers Get Good Results

• Article
• Comments (413)

MORE IN LIFE & CULTURE »

• smaller
• Larger

• facebook
• twitter
• google plus
• linked in

• Email
• Print
• Save↓ More

By 
• JOANNE LIPMAN

I had a teacher once who called his students "idiots" when they screwed up. He was our orchestra conductor, a fierce Ukrainian immigrant named Jerry Kupchynsky, and when someone played out of tune, he would stop the entire group to yell, "Who eez deaf in first violins!?" He made us rehearse until our fingers almost bled. He corrected our wayward hands and arms by poking at us with a pencil.

Today, he'd be fired. But when he died a few years ago, he was celebrated: Forty years' worth of former students and colleagues flew back to my New Jersey hometown from every corner of the country, old instruments in tow, to play a concert in his memory. I was among them, toting my long-neglected viola. When the curtain rose on our concert that day, we had formed a symphony orchestra the size of the New York Philharmonic.

Kupchynsky Family

Mr. K began teaching at East Brunswick High School when it opened in 1958.

The Saturday Essay

• A Nation Built for Immigrants (9/21/13)
• America, Syria and the World (9/14/13)
• The Ultimate End-of-Life Plan (9/7/13)
• Hard-Wired for Giving (8/31/13)
• The Failed Grand Strategy in the Middle East(8/24/13)
• In Defense of Football (8/17/13)
• Take Back Your Pregnancy (8/10/13)

I was stunned by the outpouring for the gruff old teacher we knew as Mr. K. But I was equally struck by the success of his former students. Some were musicians, but most had distinguished themselves in other fields, like law, academia and medicine. Research tells us that there is a positive correlation between music education and academic achievement. But that alone didn't explain the belated surge of gratitude for a teacher who basically tortured us through adolescence.

We're in the midst of a national wave of self-recrimination over the U.S. education system. Every day there is hand-wringing over our students falling behind the rest of the world. Fifteen-year-olds in the U.S. trail students in 12 other nations in science and 17 in math, bested by their counterparts not just in Asia but in Finland, Estonia and the Netherlands, too. An entire industry of books and consultants has grown up that capitalizes on our collective fear that American education is inadequate and asks what American educators are doing wrong.

I would ask a different question. What did Mr. K do right? What can we learn from a teacher whose methods fly in the face of everything we think we know about education today, but who was undeniably effective?

Luci Gutiérrez

As it turns out, quite a lot. Comparing Mr. K's methods with the latest findings in fields from music to math to medicine leads to a single, startling conclusion: It's time to revive old-fashioned education. Not just traditional but old-fashioned in the sense that so many of us knew as kids, with strict discipline and unyielding demands. Because here's the thing: It works.

Now I'm not calling for abuse; I'd be the first to complain if a teacher called my kids names. But the latest evidence backs up my modest proposal. Studies have now shown, among other things, the benefits of moderate childhood stress; how praise kills kids' self-esteem; and why grit is a better predictor of success than SAT scores.

All of which flies in the face of the kinder, gentler philosophy that has dominated American education over the past few decades. The conventional wisdom holds that teachers are supposed to tease knowledge out of students, rather than pound it into their heads. Projects and collaborative learning are applauded; traditional methods like lecturing and memorization—derided as "drill and kill"—are frowned upon, dismissed as a surefire way to suck young minds dry of creativity and motivation.

But the conventional wisdom is wrong. And the following eight principles—a manifesto if you will, a battle cry inspired by my old teacher and buttressed by new research—explain why.

1. A little pain is good for you.

Psychologist K. Anders Ericsson gained fame for his research showing that true expertise requires about 10,000 hours of practice, a notion popularized by Malcolm Gladwell in his book "Outliers." But an often-overlooked finding from the same study is equally important: True expertise requires teachers who give "constructive, even painful, feedback," as Dr. Ericsson put it in a 2007 Harvard Business Review article. He assessed research on top performers in fields ranging from violin performance to surgery to computer programming to chess. And he found that all of them "deliberately picked unsentimental coaches who would challenge them and drive them to higher levels of performance."

Arthur Montzka

Mr. Kupchynsky helps his daughter with her bow stroke in 1966.

2. Drill, baby, drill.

Rote learning, long discredited, is now recognized as one reason that children whose families come from India (where memorization is still prized) are creaming their peers in the National Spelling Bee Championship. This cultural difference also helps to explain why students in China (and Chinese families in the U.S.) are better at math. Meanwhile, American students struggle with complex math problems because, as research makes abundantly clear, they lack fluency in basic addition and subtraction—and few of them were made to memorize their times tables.

William Klemm of Texas A&M University argues that the U.S. needs to reverse the bias against memorization. Even the U.S. Department of Education raised alarm bells, chastising American schools in a 2008 report that bemoaned the lack of math fluency (a notion it mentioned no fewer than 17 times). It concluded that schools need to embrace the dreaded "drill and practice."

3. Failure is an option.

Kids who understand that failure is a necessary aspect of learning actually perform better. In a 2012 study, 111 French sixth-graders were given anagram problems that were too difficult for them to solve. One group was then told that failure and trying again are part of the learning process. On subsequent tests, those children consistently outperformed their peers.

The fear, of course is that failure will traumatize our kids, sapping them of self-esteem. Wrong again. In a 2006 study, a Bowling Green State University graduate student followed 31 Ohio band students who were required to audition for placement and found that even students who placed lowest "did not decrease in their motivation and self-esteem in the long term." The study concluded that educators need "not be as concerned about the negative effects" of picking winners and losers.

4. Strict is better than nice.

What makes a teacher successful? To find out, starting in 2005 a team of researchers led by Claremont Graduate University education professor Mary Poplin spent five years observing 31 of the most highly effective teachers (measured by student test scores) in the worst schools of Los Angeles, in neighborhoods like South Central and Watts. Their No. 1 finding: "They were strict," she says. "None of us expected that."

The researchers had assumed that the most effective teachers would lead students to knowledge through collaborative learning and discussion. Instead, they found disciplinarians who relied on traditional methods of explicit instruction, like lectures. "The core belief of these teachers was, 'Every student in my room is underperforming based on their potential, and it's my job to do something about it—and I can do something about it,'" says Prof. Poplin.

She reported her findings in a lengthy academic paper. But she says that a fourth-grader summarized her conclusions much more succinctly this way: "When I was in first grade and second grade and third grade, when I cried my teachers coddled me. When I got to Mrs. T's room, she told me to suck it up and get to work. I think she's right. I need to work harder."

5. Creativity can be learned.

The rap on traditional education is that it kills children's' creativity. But Temple University psychology professor Robert W. Weisberg's research suggests just the opposite. Prof. Weisberg has studied creative geniuses including Thomas Edison, Frank Lloyd Wright and Picasso—and has concluded that there is no such thing as a born genius. Most creative giants work ferociously hard and, through a series of incremental steps, achieve things that appear (to the outside world) like epiphanies and breakthroughs.

Prof. Weisberg analyzed Picasso's 1937 masterpiece Guernica, for instance, which was painted after the Spanish city was bombed by the Germans. The painting is considered a fresh and original concept, but Prof. Weisberg found instead that it was closely related to several of Picasso's earlier works and drew upon his study of paintings by Goya and then-prevalent Communist Party imagery. The bottom line, Prof. Weisberg told me, is that creativity goes back in many ways to the basics. "You have to immerse yourself in a discipline before you create in that discipline. It is built on a foundation of learning the discipline, which is what your music teacher was requiring of you."

6. Grit trumps talent.

In recent years, University of Pennsylvania psychology professor Angela Duckworth has studied spelling bee champs, Ivy League undergrads and cadets at the U.S. Military Academy in West Point, N.Y.—all together, over 2,800 subjects. In all of them, she found that grit—defined as passion and perseverance for long-term goals—is the best predictor of success. In fact, grit is usually unrelated or even negatively correlated with talent.

Enlarge Image

Arthur Montzka

Tough on the podium, Mr. K was always appreciative when he sat in the audience. Above, applauding his students in the mid-1970s.

Prof. Duckworth, who started her career as a public school math teacher and just won a 2013 MacArthur "genius grant," developed a "Grit Scale" that asks people to rate themselves on a dozen statements, like "I finish whatever I begin" and "I become interested in new pursuits every few months." When she applied the scale to incoming West Point cadets, she found that those who scored higher were less likely to drop out of the school's notoriously brutal summer boot camp known as "Beast Barracks." West Point's own measure—an index that includes SAT scores, class rank, leadership and physical aptitude—wasn't able to predict retention.

Prof. Duckworth believes that grit can be taught. One surprisingly simple factor, she says, is optimism—the belief among both teachers and students that they have the ability to change and thus to improve. In a 2009 study of newly minted teachers, she rated each for optimism (as measured by a questionnaire) before the school year began. At the end of the year, the students whose teachers were optimists had made greater academic gains.

7. Praise makes you weak…

My old teacher Mr. K seldom praised us. His highest compliment was "not bad." It turns out he was onto something. Stanford psychology professor Carol Dweck has found that 10-year-olds praised for being "smart" became less confident. But kids told that they were "hard workers" became more confident and better performers.

"The whole point of intelligence praise is to boost confidence and motivation, but both were gone in a flash," wrote Prof. Dweck in a 2007 article in the journal Educational Leadership. "If success meant they were smart, then struggling meant they were not."

8.…while stress makes you strong.

A 2011 University at Buffalo study found that a moderate amount of stress in childhood promotes resilience. Psychology professor Mark D. Seery gave healthy undergraduates a stress assessment based on their exposure to 37 different kinds of significant negative events, such as death or illness of a family member. Then he plunged their hands into ice water. The students who had experienced a moderate number of stressful events actually felt less pain than those who had experienced no stress at all.

"Having this history of dealing with these negative things leads people to be more likely to have a propensity for general resilience," Prof. Seery told me. "They are better equipped to deal with even mundane, everyday stressors."

Prof. Seery's findings build on research by University of Nebraska psychologist Richard Dienstbier, who pioneered the concept of "toughness"—the idea that dealing with even routine stresses makes you stronger. How would you define routine stresses? "Mundane things, like having a hardass kind of teacher," Prof. Seery says.

My tough old teacher Mr. K could have written the book on any one of these principles. Admittedly, individually, these are forbidding precepts: cold, unyielding, and kind of scary.

But collectively, they convey something very different: confidence. At their core is the belief, the faith really, in students' ability to do better. There is something to be said about a teacher who is demanding and tough not because he thinks students will never learn but because he is so absolutely certain that they will.

Decades later, Mr. K's former students finally figured it out, too. "He taught us discipline," explained a violinist who went on to become an Ivy League-trained doctor. "Self-motivation," added a tech executive who once played the cello. "Resilience," said a professional cellist. "He taught us how to fail—and how to pick ourselves up again."

Clearly, Mr. K's methods aren't for everyone. But you can't argue with his results. And that's a lesson we can all learn from.

Ms. Lipman is co-author, with Melanie Kupchynsky, of "Strings Attached: One Tough Teacher and the Gift of Great Expectations," to be published by Hyperion on Oct. 1. She is a former deputy managing editor of The Wall Street Journal and former editor-in-chief of Condé Nast Portfolio.

A version of this article appeared September 28, 201