Correlation and Causality

Today’s example of lazy thinking uses the religious debate of the day: global warming. If you’re a person who would rather not let the facts get in the way of a good opinion, it’s time to hit the eject button. But be warned: ignorance is not bliss.

There is a wonderful scene in Al Gore’s Inconvenient Truth where he shows this graph and he sniggers that clearly, CO2 causes global warming:

Yet there are at least another 3 possible explanations:

Correlation does not automatically equal causality. While a rise in variable A (CO2) could cause a rise in variable B (temp), it could also be the other way around…

... which is exactly what some scientists have been arguing: based on ice core samples, CO2 follows temperature, therefore temperature, they argue, is the cause for rising CO2 levels, not the other way around.

Further, variables A and B could be caused by an external variable, C. In this third case, neither A causes B, nor B causes A. Global warming has such a possibility : the variance in the tilt in the earth’s axis. There is compelling data suggesting that the most likely cause of the ice ages is due to fluctuations in the intensity and the distribution of solar radiation caused by changes in the tilt in the Earth's axis, as first theorised by Milutin Milankovitch, in 1938. (or it could also be an external variable that we have not yet identified – that’s also a possibility)

The fourth (though in this case probably unlikely) reason for the correlation of CO2 and temperature, could be pure coincidence.

In summary, we’ve already witnessed the judgment errors and risk of “slam dunk” conclusions drawn from convenient data subsets. Let's learn from history.

That is not to say that we should not worry about burning fossil fuels or doing more with less, striving for a zero carbon footprint, baking in sustainable practices or pricing in the true cost of production to incentivise change – far from it. Just that we need to avoid “Ready, Fire, Aim” conclusions and apply appropriate thinking to causality.

Update (15 Dec '08): Data points from Mars supporting the Milankovitch theory: http://www.space.com/scienceastronomy/081204-mars-climate-cycles.html

Famous last words...

"They couldn't hit an elephant from this dist_____"

General John B Sedgwick
Union Army Civil War officer's last words,
uttered during the battle of Spotsylvania, 1864

While one can argue that the above event is evidence of Black Swan theory, it seems overconfidence is an almost genetic trait in us.

Seeing entrepreneurs pitch "conservative" revenue projections to venture capital companies is an almost universal example of overconfidence. Yet, on the bright side, at least they don't get to pay the ultimate price for being wrong.

History of an Obsession

I'm currently reading Bernstein's "The Power of Gold". The irrationality of gold's value is a recent epiphany for me. In Africa at one stage, an ounce of gold was being traded for an ounce of salt. [Ed: about that time machine...]

Here are some notable quotes from the book:
"Is that not strange? Out of steel, we can build office towers, ships, automobiles, containers and machinery of all types; out of gold we can build nothing. And yet it is gold we call the precious metal."

"In a global economy managed by central bankers and international institutions, does gold matter at all?" [Ed: after the current perfect storm credit bubble, will the revised edition make the case FOR gold, based on the MISmanagement by central bankers?]

Yet, my favourite is the Prologue's opening paragraph:
"...John Ruskin told the story of a man who boarded a ship carrying his entire wealth in a large bag of gold coins. A terrible storm came up a few days into the voyage and the alarm went off to abandon ship. Strapping the bag around his waist, the man went up on deck, jumped overboard, and promptly sank to the bottom of the sea. Asks Ruskin: 'Now, as he was sinking, had he the gold? Or the gold had him?'"

[Ed: Recent volatility has caused huge spikes in gold call option purchases. I wouldn't bet on a flight to reason any time soon...]

Easy Money

In the days before ultrasound, an enterprising chap put out an advert:
Money back guarantee : Proven method tells you the gender of your unborn baby. Send $10, your date of birth, date of conception etc, to....

Thousands of suckers sent in their money. In each case the response came back: "It's a boy!"
He was right - in about 50% of the cases. The other half got their money back.
Ka-ching!

Apparently, this sort of scam is now illegal [Ed: proof that we don't have an adequate education system?]

Fine, perhaps we need protection from our own lazy thinking. But how is the above example any different to taking money for providing horoscope predictions?

Bridge Design Experiment

Test yourself:
What is the best response to the following question: Which of the three following factors - economics, safety, or aesthetics - is the most important in choosing between two designs for a bridge? (answers)

Napoleonic Wisdom

"Nothing is more difficult, and therefore more precious, than to be able to decide"

How much time do we spend on developing that ability?
Quod erat demonstrandum

Point to ponder:

Surveys consistenly show that over 90% of us believe we are "above averge".

Manifesto

Regardless of our endevours, we prefer better outcomes. Outcomes are the result of our actions, which in turn are the result of our decisions and our methodology for making them. Common sense suggests we'd be constantly refining our decision making methodology. But alas, sense is far from common...

We like to think we are reasonable, rational people. Yet the more I look at us, is appears as if the opposite is true. [Ed: perhaps you need your eyes checked?]

The aim is to expose some of these consistent or instructive errors, in the hope that we can refine our thinking, decision making and actions, resulting in better outcomes, and thus a better world.

Ironically, I'll make some mistakes while attempting to show others' errors. [Ed: in the land of the blind...] That comes with the territory. Feedback highlighting my errors will allow me to learn and grow [Ed: while providing fodder for later blog entries] Where you have alternate or better solutions, please share.

Questions

This section lists all the questions. While each question apears within the blog, like the answers, all the questions will be grouped here too.

Bridge Design Experiment
What is the best response to the following question: Which of the three following factors - economics, safety, or aesthetics - is the most important in choosing between two designs for a bridge?

Answers

Million dollar box choice:
Whether your chose box A, B or C, my prediction is that after you were told that the money was not in one of either of the two remaining boxes, you did NOT change your original choice. Yet, not changing is the irrational decision. Changing improves your odds from 33% to 50%. This clever experiment [apologies to designer(s) as I don't know who to credit] apparently has an amazing 99% decision error incidence - hardly anybody changes their original choice. [Ed: I don't feel so bad now...] Also consider that the way this example was set up, you were primed to look for errors in judgement. The real world is not so generous.

Overbidding the 'buy it now' price on Ebay and the $1m box example are errors of the same fundamental issue: perceived ownership. That is, because for a moment we have the highest bid (or we chose box X), we have an emotional commitment to that choice, clouding further rational judgement. [Ed: Fine, but I STILL think overbidding on Ebay is stupid...]

Bridge Design Experiment:

This experiment was designed by Jim Shur. The methodology can be found in his book "Choosing by Advantages" and is used extensively in Value Engineering.

The answer will be delayed till October.