Decision Making: Reason-Based Choice

Reason-Based Choice

How should we think about the data we’ve reviewed? On the one side, our everyday experience suggests that we often make good decisions—and so people are generally quite pleased with the cell-phone model they selected, content with the car they bought, and happy with the friends they spend time with. At the same time, we’ve considered several obstacles to high-quality decision making: People can easily be tugged one way or another by framing; they’re often inaccurate in predicting their own likes and dislikes; and they seek out lots of choice and flexibility but end up, as a result, less content with their own selections. How should we reconcile all of these observations?

The answer may hinge on how we evaluate our decisions. If we’re generally happy with our choices, this may in many cases reflect our contentment with how we made thechoice rather than what option we ended up with. In fact, people seem to care a lot abouthaving a good decision-making process and seem in many cases to insist on making a choice only if they can easily and persuasively justify that choice.

Why should this be? Let’s start with the fact that, for many of our decisions, our environment offers us a huge number of options—different things we could eat, dif-ferent activities we could engage in. If we wanted to find the best possible food, or the best possible activity, we’d need to sift through all of these options and evaluate the merits of each one. This would require an enormous amount of time—so we’d spend too much of our lives making decisions, and we’d have no time left to enjoy the fruits of those decisions. Perhaps, therefore, we shouldn’t aim our decisions at the best possible outcome. Instead, we can abbreviate the decision-making process if we simply aim at an option that’s good enough for us, even if it’s not the ideal. Said differently, it would take us too much time to opti-mize (seeking the optimum choice); doing so would involve exam-ining too many choices, and we’ve seen the downside of that. Perhaps, therefore, we’re better advised to satisfice—by seeking a satisfactory choice, even if it’s not the ideal, and ending our quest for a good option as soon as we locate that satisfactory choice (Simon, 1983).

But how should we seek outcomes that are “satisfactory”? One possibility is to make sure our decisions are always justified by good reasons, so that we could defend each decision if we ever needed to. A process like this will often fail to bring us the best possible out-come, but at least it will lead us to decisions that are reasonable— likely to be adequate to our needs.

To see how this plays out, consider a study in which half of the participants were asked to consider Scenario A in Figure 9.18 (after Shafir, Simonson, & Tversky, 1993). In this scenario, 66% of the participants said they would buy the Sony CD player; only 34% said

they would instead wait until they had learned about other models. Other participants, though, were presented with Scenario B. In this situation, 27% chose the Aiwa, 27% chose the Sony, and a much larger number—46%—chose to wait until they’d learned about other models.

In some ways, this pattern is peculiar. The results from Scenario A tell us that the participants perceived buying the Sony to be a better choice than continuing to shop; this is clear in the fact that, by a margin of 2 to 1, people choose to buy. But in Scenario B, participants show the reverse preference—choosing more shopping over buying the Sony, again by almost 2 to 1. It seems, then, that participants are flip-flopping; they pre-ferred the Sony to more shopping in one case and indicated the reverse preference in the other.

This result makes sense, though, when we consider that people usually seek a jus-tification for their decisions and make a choice only when they find persuasive rea-sons. When just the Sony is available, there are good arguments for buying it. (It’s a popular model, available at a good price for just one day.) But when both the Sony and the Aiwa are available, it’s harder to find compelling arguments for buying one rather than the other. Both options are attractive and, as a result, it’s hard to justify why you would choose one of these models and reject the alternative. Thus, with no easy justification for a choice, people end up buying neither. (Also see Redelmeier & Shafir, 1995, for a parallel result involving medical doctors choosing treatments for their patients.)