The Anchoring and Adjustment Heuristic¶
Barbara Hernandez & Charles Hicks
What factors do you take into consideration when making an estimate? Do you rely on your previous experiences, do you think about whether a guess is reasonable, do you compare values? What about the information you were previously exposed to? In this chapter, we will discuss the anchoring and adjustment heuristic, by which individuals’ estimations are biased on an initial value they do not sufficiently account for (Tversky and Kahneman 1974). We will summarize several articles that discuss this heuristic and explain how this bias can be found in the medical field.
Estimating Unknown Events¶
Amos Tversky and Daniel Kahneman published an article in 1974 in which they presented a series of heuristics that we use when making decisions about things we are unfamiliar with (Tversky and Kahneman 1974). One of the heuristics Tversky and Kahneman included was the anchoring and adjustment heuristic. Before explaining the studies they conducted to test this concept, let’s see how you respond to this question!
Instructions: Quick! You have five seconds to calculate the answer to this multiplication. If today’s date is an odd number, answer the first multiplication. If today is an even number, look at the second multiplication.
8 x 7 x 6 x 5 x 4 x 3 x 2 x 1
————————————-
1 x 2 x 3 x 4 x 5 x 6 x 7 x 8
The correct answer to this multiplication was 40,320. How did you do?
This example is a short version of the experiment carried out by Tversky and Kahneman (1974). They found that the sequence of multiplication individuals were assigned influenced their final estimation, as individuals with the high-to-low multiplication scored an average 2,250 while those with the low-to-high list averaged 512. In this example, since you only had five seconds to answer, the information you anchored your decision on was likely the first few numbers in the multiplication, which is why we can expect people who were assigned to the first multiplication to have a higher average estimate while those who were exposed to the second multiplication calculated a lower average response. To explain the results of their experiment, the authors note that individuals may make a few estimates that serve as a baseline and do not change their computations enough. Essentially, what the authors are trying to demonstrate is that individuals will use the first information they observe to calculate a final value, without sufficiently adjusting their estimates.
A second experiment carried out by Tversky and Kahneman (1974) asked participants to spin a wheel and based on the value the wheel fell on, indicate whether it overestimated or underestimated a value such as the number of African countries that are part of the UN and then provide an estimate of that value. Participants with higher numbers on the wheel had higher estimations than those with lower numbers on the spinning wheel. Through Tversky and Kahneman’s experiments, we observe how random information researchers provide can lead to over or underestimations.
In their 1999 article, Thomas Mussweiler and Fritz Strack elaborate on the origins of Tversky and Kahneman’s anchoring heuristic. Mussweiler and Strack propose that the anchoring heuristic occurs by individuals selecting information that convinces them that the anchor provided is the same as the correct response (Mussweiler and Strack 1999). Essentially, this theory suggests that when we are given an anchor, our decision making is more focused on finding information that confirms that value than on finding information that supports a different response.
Like Tversky and Kahneman’s study, Mussweiler and Strack’s research looked at externally-generated anchors. However, what happens when individuals generate their own anchors? Nicholas Epley and Thomas Gilovich (2001) studied individuals’ anchoring and adjustment mechanisms when anchors are created by participants. Epley and Gilovich challenged Mussweiler and Strack’s theory on selective accessibility, arguing that since individuals know whether their anchor is accurate or not, individuals adjust their answers based on their initial estimate rather than using prior experience and selective information to confirm a value. In one experiment, researchers asked participants questions they would have an initial self-generated value to and questions they would not have any self-generated anchors to base their answer on. In their experiment, researchers recorded participants’ reasoning for arriving at an answer and had two independent reviewers assess whether participants used anchors in their reasoning. The researchers found higher rates of anchoring and adjustment for self-generated values than for those that were external, suggesting that the anchoring and adjustment heuristic is present when individuals generate the anchors themselves, as well as when the anchors are provided by the researcher.
These three articles have provided an overview of initial conception of how anchors, regardless of how relevant they are, are used by individuals to generate estimates while also questioning whether anchors generated by experimenters yield the same results on estimation as when they are created by individuals.
Reinventing the Wheel¶
In all of Tversky and Kahnemnan’s findings, some of the most confounding were those focused on randomly selected anchors. When presented with a logical problem regarding approximation, why should a value selected by a spinning wheel influence our opinion? As our minds are presented with the wheel’s selection, our association between the value and the estimation is factored in our approximations. While this speaks to our ability to make quick estimations, it also exposes some faults in our cognitive processes that are explored in the following section.
Real World Implications¶
Although heuristics can help us make quick judgments or make complicated decisions on a daily basis, these mechanisms can lead to biases. The following section is focused on providing examples of the effects the anchoring and adjustment heuristics can have on patients and medics.
One way in which the anchoring and adjustment bias may present itself in the medical field is how patients misconstrue the probability of having a disease (Senay and Kaphingst 2009). Although patients may be told they are statistically unlikely to have a disease, they continue to think they are likely to get sick because they base their judgment on their own perception of self-risk instead of incorporating the probability provided by the physician. In this case, a self-generated anchor impedes patients from adjusting to an externally-calculated probability.
Another way this heuristic is manifested in the medical field is in the diagnoses physicians give to patients. In a case review, Edward Etchells described the case of a man who was incorrectly diagnosed by doctors several times despite showing worsening symptoms (Etchells 2015). In analyzing this case, the author commented on the different biases, including the anchoring and adjustment heuristic, that may have influenced the doctors’ misdiagnosis (Etchells 2015). In the article, Etchells argued that the anchoring and adjustment heuristic, along with other biases, prevents physicians from changing their diagnoses despite patients displaying symptoms that do not match their initial diagnosis.
Conclusion¶
The anchoring and adjustment heuristic emphasizes the important role exposure to information can have in coming to an estimation or judgment. In this final section, we have some questions to think back on the content presented in this chapter and ways the content can be applied to the real world.
Review Questions¶
What is the anchoring and adjustment heuristic?
What are some studies conducted on the topic? What did they show?
What are the two types of anchors people can be exposed to? Do these make a difference in the calculation of final results? Why or why not?
What are the ways in which individuals modify their decisions?
What are some real-world examples of this heuristic that were not mentioned in this chapter?
Why does this heuristic matter?
Additional readings:
References¶
Epley, N., & Gilovich, T. (2001). Putting Adjustment Back in the Anchoring and Adjustment Heuristic: Differential Processing of Self-Generated and Experimenter-Provided Anchors. Psychological Science, 12(5), 391–396. https://doi.org/10.1111/1467-9280.00372
Etchells, E. (2015, June). Anchoring Bias With Critical Implications [PSNet]. Web M&M Cases and Commentaries. https://psnet.ahrq.gov/web-mm/anchoring-bias-critical-implications
Mussweiler, T., & Strack, F. (1999). Comparing Is Believing: A Selective Accessibility Model of Judgmental Anchoring. European Review of Social Psychology, 10(1), 135–167. https://doi.org/10.1080/14792779943000044
Senay, I., & Kaphingst, K. A. (2009). Anchoring-and-Adjustment Bias in Communication of Disease Risk. Medical Decision Making, 29(2), 193–201. https://doi.org/10.1177/0272989X08327395
Tversky, A., & Kahneman, D. (1974). Judgment under Uncertainty: Heuristics and Biases. 185, 10.