What if?

My sister needed a wardrobe for her room and bought one at Casas Bahia. Since she lives in a two-story house with a spiral staircase leading to her room, she was emphatic with the salesperson: “Please, deliver the wardrobe disassembled, otherwise, it will be impossible to get it up the stairs.”

The next morning, a truck arrived at her house carrying the wardrobe—fully assembled. The truck left with the wardrobe, which couldn’t make it up the spiral staircase. She went back to the store, found the salesperson, and argued that she had been very clear about the need to send the wardrobe disassembled. The salesperson replied:

“When I arrived in the morning, the guys had already assembled your wardrobe. So I thought, ‘Maybe it’ll work!’ and told them to deliver it.”

We all make decisions daily, of course, but why do some of us insist on making decisions that go against all odds of success?

I’ve written before about coherence and risk propensity, which are fundamental to understanding decision-making dynamics, but today I want to talk about evaluating uncertainty. I believe this is at the root of this “problem” and ties into a recurring theme in today’s world of the internet and information overload: the quality of the information we have.

Consider, in the past (and I’m talking about 1993, the pre-internet era), there was genuinely little information available. And that information wasn’t always accessible, as the effort to obtain it was often fruitless or simply not worth it.

Today, the amount of information produced in a single year exceeds the amount produced by all of humanity in the last 40,000 years. Of course, not all of this information is good or useful, but thanks to digital means, all of it is at our fingertips. This brings a new problem: how do we distinguish between good and bad information?

Let me give an example. I might want to know if a student who comes to class happy learns more than one who comes in unhappy. How do I assess if my students are happy? The best way would seem to be to ask them. So I place a sheet of paper on each desk with a simple multiple-choice question: “Are you happy? Rate from 1 to 5, where 1 means ‘very unhappy’ and 5 means ‘very happy.’” I quickly analyze the results and decide whether my class should be tougher or lighter. Right? Wrong!

What is the quality, the credibility, of people’s answers to the question, “Are you happy?” Even without a background in psychology, I can think of 100 reasons why someone might respond with a bias toward “very unhappy” or “very happy,” unrelated to their actual emotional state.

If you decide to use the information provided by this questionnaire, no matter how good your decision-making method is (e.g., a Bayesian statistical approach), your decision will be no better than a guess.

So how do we know if information is good? In the absence of a verification mechanism, we have to trust our judgment.

To be certain, the salesperson from Casas Bahia could have gone to my sister’s house with a measuring tape, taken the dimensions of the staircase and doorway, and compared them with the dimensions of the wardrobe—both assembled and disassembled. This would have allowed her to make a decision without any doubt. Without such confirmation, she should have trusted my sister’s word, as my sister knows her own house better than the salesperson does. And unlike my hypothetical students, my sister had no psychological reason (that we know of 😉) to provide unreliable information.

But even so, the salesperson made a decision contrary to logic and reason. Why?

At first, I wrote a long answer to this question (which will become my next text) discussing judgment (or the lack of it), but then I applied Occam’s razor to my ideas and arrived at a simpler conclusion.

The salesperson acted against logic out of laziness! It has nothing to do with a lack of statistical tools or judgment. It was laziness and irresponsibility. She wasn’t the one who assembled the wardrobe unnecessarily, nor was she the one who drove the truck or unloaded the wardrobe unnecessarily, nor was she the one who would go another day without a wardrobe. Her responsibility was to deliver the wardrobe that day (even if assembled, which adds a lack of ethics to her qualities).

The conclusion is that laziness is not a good decision-making criterion.

First published in Portuguese in ‘Você que é Biólogo…’ in January, 2010.