Title: Predictably Irrational
Author: Dan Ariely
Publisher: Harper Collins Publishers
Having found that I am rather shabby with light theoretical books (in some strange way, if a book does not make me feel a bit uncomfortable about myself and my past ventures within the subject matter, I decide it’s a picnic and mentally throw it away), I developed a method. It’s called “Give the book at least two chances”. I’m still working on the nomination … It runs like this: If first I spot what then seems to be a blatant, even if unreservedly advocated for error, I’m intrigued: “This might be of the revolutionary sort!”, I say to myself. Secondly, if I spot another one within the arguments put forward for justifying that the former was not an error, I’m deflated. The third stage is the mental propulsion. I may not always right, of course, but this is how I function.
A friend of mine acquainted me with this much-publicized book called Predictably Irrational, by Dan Ariley. A hit, admittedly, since it had already been translated into 15 languages and it promised to show that “... we are not only irrational, but predictably irrational—that our irrationality happens the same way, again and again”. Wouldn’t you be intrigued? I most certainly was. Even though, before moving on from the introduction, I thought rationality is a peculiar word to use in behavior research, I was prepared to be dazzled.
Here’s the story. In the ad below, three subscription offers were available:
As one might notice, the ‘print subscription’ is lousy. Assuming one would want only the printed versions of The Economist, he or she would definitely choose the third option for the same amount of money without regard to his or her real need for an ‘online access to all articles from The Economist since 1997’. So far so good; no one would think this is irrational (and, in this sense, neither the author), for it could always be subsequently motivated by claims such as “Well, I knew I didn’t need it then, but I thought that since it was free, maybe I would need it later so it felt good to have it secured.”
But Dan Ariley made an experiment to prove that, in some other sense, it is irrational. That the ‘lousy variant’ – which he calls the ‘decoy’ -, by its mere presence, makes us think in a relative manner and shapes our decision. The experiment was this: 100 students were faced with the wicked ad and made their choice as follows: 84 for the Internet & Print, 16 for the Internet Only and none for the Print Only. These were students from the MIT’s Sloan School of Management and, observably, none of them was drunk or in any other way not lucid.
But how can you prove that they’ve done it because of the ‘mere presence of the decoy’? Well, remove the decoy, see what happens. Which the author did, and the results came astonishingly different. Let me quote here, so you can feel the enthusiasm:
Would the students respond as before (16 for the Internet only and 84 for the combination)? Certainly they would react the same way, wouldn't they? After all, the option I took out was one that no one selected, so it should make no difference. Right? Au contraire! This time, 68 of the students chose the Internet-only option for $59, up from 16 before. And only 32 chose the combination subscription for $125, down from 84 before.
So, 68 (up from 16) for the Internet Only offer, and 32 (down from 84) for the Internet & Print. This is admittedly a big difference. In fact, at the end of the passage we have a footnote assuring us that it is “statistically significant”, which would have been nevertheless obvious, but it’s always good to know.
OK, now the conclusion: What changed the decision? “Nothing rational, I assure you. It was the mere presence of the decoy that sent 84 of them to the print-and-Internet option (and 16 to the Internet-only option). And the absence of the decoy had them choosing differently, with 32 for print-and-Internet and 68 for Internet-only”. When an experimentalist ‘assures you’ of something, you are almost forced to regard obliquely the issue presented. Which he did, which I did.
Let us review what the two experiments consisted of. The first 100 decisions (I will thenceforth assume they are “statistically significant”) were reached with consistent advantage for the Internet & Print offer (84%). Then, with the same (maybe not the same, but equally statistically significant, I presume) consistent advantage for the Internet Only offer (68%). What changed? The ‘decoy’.
Oblique question nr. 1: WHO made the second decisions? Were the subjects the same subjects from the first experiment? If they were, and we agreed that they were neither drunk nor stupid, they could not have forgotten the fact that the ‘decoy’ existed, that the ‘decoy’ is at least still possible if not, for the sake of the experiment, available. Did he men-in-blacked their memory in order to ascertain the fact that all conditions remained the same? And if they weren’t the same subjects,
Oblique question nr. 2: HOW were they exactly the same? Of course, no one’s asking for genetic precision, but when studying decision making, in order to consider 100x = 100y, you must at least have some indication that their cognitive (cerebral?) similarities are reliable. If the subject were not the same, could it not be possible that the 100y’s simply made other decisions?
Oblique question nr. 3: HOW were the second set of decisions made? Did all 16 of the Internet-Only subjects stayed at the Internet Only option? Did 32 of the Internet & Print stayed at the Internet & Print? Because, if not, than it means that it is possible that only 52 of the students changed their view. I really do not know anything about statistics, but I assume 52% is not that big a significance.
Oblique question nr. 4: Were the subjects even INTERESTED in buying such an offer? One could think MIT students are supposedly and genuinely interested in such magazines. And one could be right. But if measuring their price selection (in other words not the I would/I wouldn’t decision, but the this much/not that much one), wouldn’t be nice to know that all 100 (or 200) students would be willing to pay more or less the same amount for the chosen object? Is it even possible to know that?
Oblique question nr. 5 (and the most important): HOW is that irrational? What ideal of rationality are they abusing? If we accept that decision is based mainly on the relation between the ‘decided-upon’ and its relations to other ‘not-decided-upons’, then we must suppose that regardless of the relativity, one option is rational, while others are not (if not, how are we to judge something as irrational?). Which one could be this rational decision? The Internet Only offer? Why? Just because more, at the second round, more students chose it? If the Internet & Print option is the rational one, isn’t the ‘decoy’ actually rationalizing, rather than … irrationalizing?
It took me some time to unwrap my head from the first question. In order to justify the claim that different decision have been made after only one modification in the context, everything else being the same, you must at least have the same students. But ‘having the same students’ means ‘having the same students’ brains’, moreover since we are about to evaluate their rationality.
 Of course, there is always the ‘take the wide sense’ getaway, but rationality – at least since Kant’s unreadable metaphysical journeys – is in fact only suited for mathematical, logical or geometrical thought. We are, strictly speaking, being rational only when operating in abstract systems where – again, strictly speaking – decision is not taken, but necessarily deduced.