Thursday, December 10, 2009

Why Haste Doesn't always make Waste: Mathematical Proof that a "Hunch" Can be Enough


It's logical.

The more time you spend deliberating, researching, and weighing your options, the more educated and informed your judgements will be...


Right?

For example, let's say that you were a museum curator, trying to find out whether or not a statue, supposedly a long lost ancient Greek treasure, was the "real deal" or not.

Whose opinion would you trust more?

(A) A team of researchers who spent 14 months and hundreds of thousands of dollars analyzing, date testing, and researching the statue.

-or-

(B) an expert who looked at the statue for a few seconds, and had a "gut reaction?"

You'd pick (A), right?

According to bestselling author Malcolm Gladwell, wrong.

Gladwell's groundbreaking book Blink introduces a phenomena called "thin slicing."

"What on earth is thin slicing???" you ask?

It has nothing to do with going to a deli...

Thin slicing is something we do everyday. "Thin slicing" means making big inferences based on small pieces of information. In essence, this theory shows that in many cases, following your instincts, can often be more valuable than taking time to do painstaking research...

The issue of the statue was not just hypothetical. This actually happened. In 1983, an art dealer named Gianfranca Becchina came to the Paul Getty Museum in California with a 7 foot tall marble statue called a Kourous.

[this is not the actual Kourous in question..its just to give you an idea...]

There are only 200 kouroi left in the world, most of which had been damaged very badly. The one Becchina brought in seemed to be in good shape, so Becchina started the bidding at $10 million.

Not wanting to fork up the change for nothing, the museum did a great deal of research. They looked up tons of documents relating to its recent history, and found out it had been in the possesion of a Swiss art collector since the 1930's. Geologists examined the surface of the statue with "electron microscopes, electron microprobe, mass spectrometry, x-ray diffraction, and x-ray fluorescence." They determined that the statue must have been at least several thousand years old.

After 14 months, the museum was getting ready to release the statue for display to the public...they showed the statue to several art historians.

All of them, immediately upon seeing it, felt a hunch that something was fishy.

They museum invited a whole slew of art historians to take a look, and all of them had the same reaction..."it doesn't look right."

Eventually researchers dug deeper..and eventually found out..the letters pertaining to its history were forged...and the statue was "aged" using potato mold.

All it took was a "hunch" to turn 14 months of research on its head.


But thin slicing doesn't only mean using "hunches". It can be proved mathematically too.

How about...relationships? Marriage?



Marriage is surely a very complicated thing. Sometimes the most unlikely couples meet, and stay together forever...and other times, couples that seem perfect for each other, end up splitting.

To make an accurate prediction of whether or not a couple will stay married would take a lot of time...weeks, even years..watching the couple's behavior, interactions, how they act when they're tired, hungry, angry, who cooks breakfast, which side of the bed they sleep on, who drives the kids to school...ect...right?

Wrong.

John Gottman, a psychologist and mathematician from MIT, has marriage predication down to a science.

Gottman has brought thousands of couples into his "love lab" to test out his theory. He sits them in a room with two chairs,a table, and two video cameras and asks them to speak about "any topic from their marriage that has become a point of contention" for fifteen minutes. The couples have electrodes clipped to their fingers and ears measuring heart rate, sweat, and temperature, and their chairs are equipped with "jiggle-o-meters"


Gottman created a code called SPAFF, which has twenty separate categories, correlating to each emotion that might come up in a conversation. Disgust=1, Contempt=2, anger=3, whining=11..and so on. He Assigns a number to every second of the fifteen minutes for the male and the female. This results in a very long strand of code. [1,4,6,11,17,1,3,4,5,6,7,....] thousands of digits long. This SPAFF code is coupled with data from the electrodes and jiggle o meters..and..voila! In fifteen minutes, Gottman can predict with 90 percent accuracy, whether or not the couple will still be together in 15 years. Give him an hour with the couple, and this goes up to 95%. Even by looking at a tape 3 minutes long, he can still predict with pretty impressive accuracy.



Another researcher, Wendy Levinson recorded hundreds of doctors interactions with their patients. Half of the doctors had been tried for medical mal practice at some point in their careers, and the other half had not. She put the audio through a filter that eliminated high frequency sounds, making the words unrecognizable, but everything else about the converstation: pitch, intonation, rythym, were all still there. She cut the clips into 10 second slices, and analyzed each slice for its main quality: warmth, hostility, anxiousness, or dominance. She predicted with overwhelming accuracy which of these doctors had been sued and which had not. Levinson did not need to know the doctors qualifications, degrees, specialties, or even what they were saying. All she needed to know was how they interacted with their patients.




So what does this all add up to?

Your "gut" was put there for a reason. Sometimes, your primal instinct is the most trustworthy source of information. This is not to say that you should make all your important life changing decisions on a whim...but perhaps it goes to show that the power of an immediate instinctive reaction should not go underrated.

So...next time you are about to change your answer on a test after checking it over, or about to buy an multi million dollar marble statue...remember...trust your instincts. They will usually lead you in the right direction.


-Josh


P.S. Go read Blink!