Wednesday, September 02, 2015

Psychology of Intelligent Problem Solving

Everyone knows good leaders must be intelligent, but how do we know who is intelligent? One possibility is to ask the psychologists, but I have my doubts about that procedure. Let me explain.

Two Sample Problems
Some time ago, I read a magazine column prepared by Mensa, "an organization of people who score in the top two per cent on standardized IQ tests." Among the test questions they offered were the following two questions, which bent my mind in a direction perhaps not intended by Mensa.

1. All the secretaries in my office are under 21. All the young ladies in my office are very beautiful. My secretary has long blond hair and blue eyes. Which statements below can be justified by the information given?
a) My secretary is under 21.
b) My secretary is a beautiful young lady.
c) Neither of the above.
d) Both of the first two.

4. In a certain field there are both horses and men. There are 26 heads and 82 feet (or hooves) in the field. How many men? How may horses?

Answers
1. The "correct" answer is given by Mensa as (a). What's supposed to trip you up is the assumption that all the secretaries are female, which isn't stated. But what about the assumption that my secretary is in my office? It doesn't say that, but then it doesn't say my secretary is female. Depending on which set of unstated assumptions you choose, all four answers are possible. Although I've never personally had a male secretary, I have worked in many situations where my secretary was in another office—even in another city. Does this make me less intelligent than the psychologist who posed the question?

4. By simple algebra, you get the official answer: 15 horses and 11 men—if you assume that each horse has four feet and each man has two. (You must also assume there are no other beings with heads or feet in that field, and that heads of cabbage, for instance, don't count.) I don't know much about horses, but plenty of men have only one foot, or none. So, for example, one possible solution is 16 horses and 11 men, one of whom is a wounded veteran with no legs. Then, also, a quick Google search of wikipedia yields this quote: "Two-headed people and animals, though rare, have long been known to exist and documented." What does a two-headed horse do to your solution? 

Multiple Solutions
In other words, there are oodles of other solutions—if you're not a psychologist.

Anyone with any intelligence who has been exposed to this genre of testing has experienced the same kind of frustration. You can think of several possible answers, but you know the psychologists want only one—and you're not allowed to ask questions. It's not so bad when the questions are in an entertaining magazine article, or even in the Mensa admission test. (Who wants to be in Mensa anyway?)

But what if you want to be admitted to college? Or win a scholarship? Or put your child on a favored track in second grade? Or get a job? (I sometimes use questions like these when I'm hiring software testers, but I'm looking for people who can think of the most different answers to each question, not the one "right" answer.) The psychologists hold the power to keep you from getting what you want, and their power is rarely questioned.

The Central Dogma
Quite possibly, modern psychology is the most arrogant profession of all time (unless it's programming). In effect, the central dogma of modern psychology says this:

"There is exactly one right answer,
and the psychologist knows it."

This dogma applies equally to test for people at work or rats in mazes. Any rat that displays a modicum of suspicion for the psychologist's setup runs the maze a bit slower than the others—and is labeled "less intelligent." To me, anyone who fails to be suspicious of psychologists should receive quite the opposite label.

This central dogma is damaging enough to the individual person or rat trapped by the psychologist, but its long-range effects on society may be even worse. Schools and employers reward people whose thinking happens to match the narrow thinking of psychologists who produce these tests, so people either learn to think that way or find themselves out in the cold.

After a while, we find  people in problem-solving situations who literally believe that every problem has one and only one solution—a solution so inevitable that they will recognize it when they discover it.

Managers infected by the central dogma act like psychologists. They assign work to their subordinates and expect to have it accomplished in the one right way—their way. Infected designers rarely consider an adequate number of alternative designs—and never consider testing other than by their own intuition. Infected programmers and testers are powerless in the face of a bug that deviates in any way from the "obvious" answer.

Another Problem
In our Problem-Solving Leadership (PSL) workshop, we naturally devote much of the time to problem solving. No matter what problems we use, no matter what we think the "right" answer is, the technical leaders who participate always come up with something better. Try this one, for example, which some of our participants concocted for their classmates:

A man hires a worker to do seven days of work on the condition that the worker will be paid at the end of each day. The man has a 7-inch bar of gold, and the worker must be paid exactly one inch of that gold bar each day. In paying the worker, the man makes only two straight cuts in the bar. How does he do it?

The "right" answer to this problem was supposed to be that the man cuts the bar into lengths of 1, 2, and 4 inches. By "making change," he can pay the man exactly an inch a day—certainly a rather clever solution. Being an old binary machine rat, I found that solution immediately. I felt awfully intelligent as I watched some of the participants struggle to find it.

I no longer felt so intelligent when several participants came up with another solution. The problem says nothing about bending the bar, so they had the man forming the bar into a figure S. Then with two cuts, you can get exactly seven pieces—and make a dollar sign, to boot. I wonder if they did this just to revenge themselves on the psychologists.

My own experience with technical leaders tells me that the best of them operate on a different central dogma:

Any real problem has one more solution
which nobody has found—yet.


They might not be able to find that next solution, or find it right now. It may not be worthwhile finding under the present circumstances. But it's there, waiting to be found. And someone will find it, someday—if they have not been brainwashed by those darn psychologists. That's ultimately the principal reason we want our software developers and testers to work in teams.

4 comments:

John Flach said...

Interesting observations. While it is true that much of the laboratory research on decision making and problem solving is framed around closed problems that are designed to have one "absolutely right answer" so that performance can be scored right or wrong by the experimenter, there are significant numbers of psychologists (among others) who recognize that most problems in the world are "open" with many satisfying solutions. This was particularly true of the early Gestalt psychologists (Seltz, Duncker, deGroot, Wertheimer). You might look at two papers by Lindblom on "Muddling Through)" (1959, 1979) that are more representative of how people manage complexity in everyday contexts. Many of us psychologists have become skeptical and critical of a decision literature designed around puzzles that trivialize thinking.

I agree that "intelligence" testing is one of the areas where psychologists have tended to be blinded by hubris and by the financial rewards associated with the testing industry. But this is not true of all.

If I were to blame anyone for trivializing complex problems and decision making it would be the Harvard Business School's approach to management by Spreadsheet that is analyzed quite nicely in Saul's Voltaire's Bastards.

Gerald M. Weinberg said...

John,
thank you for your comments, which are totally correct, relevant, and useful. As you point out, I was only commenting on a subset of all psychologists, trying to emphasize how dangerouis and widespread their influence can be. I hope all my readers will read the references you cite, and take your message to heart—along with my message to stop trivializing thinking.

Danny Faught said...

"Who wants to be in Mensa anyway?" - I've thought about taking a test to qualify for membership, but what if I fail? I'd rather not join than possibly find out I'm only in the top 3%. :)

Lars Outzen said...

If you can bend it into an S figure you can also 2d-ZigZag it to do the split using one cut, or go into 3d and spiral the gold bar and also only use one straight cut. If you have a 4d cutter, you also only need one cut. Don't think there are any other 1d-binary-2-cut solution easily available. No mention of a straight cut though... also not mention of the no-pay-spend-2-exchange assumption. Yes I always have issues in-side the box :-)

Currently I work for/with a solution architect - 10 years younger than me - who tells us to write down our assumptions when we do the estimates / sw-design. Of course there is not enough "room" for all of the assumptions - some are just so implicit and tacit that everybody knows then . . I do wander what they are.