Evolution and cognitive biases: the decoy effect
As discussed in the lecture, it is not the case that evolution always leads to optimal decision making. Humans have many ‘cognitive biases’, that make us consistently deviate from an accurate perception of the world, and therefore from optimal decision making. One example is the ‘decoy effect’: the choice between two options is influenced by a third, relatively unattractive option.
The decoy effect was famously shown in an experiment where subjects had to choose between restaurants . Some of the subjects had to choose between two options: a five-star restaurant that was a 25-minute drive away, and a three-star restaurant at a 5-minute drive away. For other subjects, an inferior third option was added to the choice set; this was either a four-star restaurant at a 35-minute drive, or a two-star restaurant at a 15-minute drive. Although the decoy options were almost never chosen, they did influence the choice for the other two options. When the first decoy option was added, this made subjects more likely to choose the five-star restaurant; when adding the second decoy option, subjects were more likely to choose the three-star restaurant. The reason is that the first decoy is worse than the five-star restaurant in both dimensions: it is farther away, and it has fewer stars. It is only worse than the three-star restaurant in one dimension: it is farther away, but it has more stars. Thus, the decoy option makes the five-star restaurant look particularly attractive, and subjects are more likely to choose it. The reverse is true for the second decoy option.
Interestingly, the decoy effect has also been shown in animals, including ants, honeybees and various bird species. In an experiment with hummingbirds , two nectar foraging options were offered: one with a high sugar concentration, but relatively low in volume, and another with a high volume of a low sugar concentration. Also, two different decoys were added, using the same principle as the study on humans referred to above: one of them was worse than the first option in both concentration and volume, but only worse in volume than the second option. The other was worse than the second option in both concentration and volume, but only worse in concentration compared to the first option. Just like the humans, the hummingbirds tended to choose for the option that was better than the decoy option on both counts.
The decoy effect is definitely not the only cognitive bias that is apparent in humans . Many other cognitive biases are known, such as the confirmation bias (the tendency to focus on information that confirms already held preconceptions), the endowment effect (the fact that people demand more to give up an object that they own, than they would pay to acquire it), and the anchoring effect (discussed in the lecture). Similarly, many other biases are well-documented in the animal behavior literature .
As discussed in the lecture, one of the main reasons that evolution has not produced perfect decision makers, is that organisms simply face a too complex world. They face so many different circumstances in their lives that the constraints on their physical systems (such as their neurological machinery) make it impossible to have a perfect response to every possible situation. As a result, psychological mechanisms have evolved that perform well on average, but may be suboptimal in specific circumstances.
Here is a Wikipedia-article with a list of cognitive biases.
 Huber, J., Payne, J. W., and Puto, C. 1982. Adding asymmetrically dominated alternatives: violations of regularity and the similarity hypothesis. Journal of Consumer Research, 9, 0-98.
 Bateson, M., Healy, S. D., and Hurly, T. A. 2003. Context-dependent foraging decision in rufous hummingbirds. Proceeding of the Royal Society of London B, 270, 1271-1276.
 Kahneman, D. 2011. Thinking, fast and slow. New York, NY: Farrar, Straus, and Giroux.
 Fawcett, T. W., Hamblin, S., and Giraldeau, L. 2013. Exposing the behavioural gambit: the evolution of learning and decision making rules. Behavioral Ecology, doi:10.1093.
© University of Groningen