Heuristics & Biases

As you now know, we utilize two very different kinds of thinking in making decisions - quick, effortless System 1 thinking, and slow, deliberate System 2 thinking.  While System 1's main advantage is speed, it comes at the cost of reduced accuracy.  A long list of flaws and biases in System 1 thinking have been well documented, and two of them include the anchoring effect and availability heuristic, ​which you'll learn about on this page.
Try It Out

1.  Write down the last two digits of your phone number on a piece of paper

2.  Answer this question: "Was Barack Obama older or younger than the number you have written down when he became President of the United States?"

3.  Answer this question:  "How old do you think Barack Obama was when he became President of the United States?"
Anchoring Effect

As you completed the task above, you might have been wondering what the last two digits of your phone number have to do with Barack Obama's age when he became President.  Despite the obvious irrelevance of the two numbers, in similar experiments, people whose phone number ends in a high value (say 87) tend to estimate a higher age for Obama becoming President than people whose phone number ends in a low value (say 17).  This makes no logical sense, but is a well replicated research finding.  How can this be explained?

In order for System 1 to come up with a fast, effortless answer, it has to make use of mental shortcuts, otherwise known as heuristics.  These heuristics help System 1 come up with a decent guess to a question with little mental effort.  However, their use leads to biases in decision making.

One example of a System 1 bias is the anchoring effect.  The anchoring effect involves making use of a reference point, or anchor, in coming up with an estimate.  If you have a particular number in mind, no matter where this number came from, it will influence any estimate you subsequently make.  In the Obama example, people whose phone number ended in a high value had a high anchor, and subsequently tended to over-estimate the age that Obama became President, while the opposite was true for people whose phone number ended in a low value.

Let's imagine how this might work.  If your phone number ends in a high value, say 87, you obviously know that Obama was younger than 87 when he became President.  So, you begin adjusting the number downwards, until you arrive at a number that "seems right" - perhaps around 55.  On the other hand, if your phone number ends in a low value, say 17, you would adjust this value upwards until you reach a reasonable figure - perhaps around 40.  By starting from a different reference point, you end up with a very different estimate for Obama's age. (In case you were wondering, Obama actually became President at the age of 47).

Although these sorts of studies may seem like silly little experiments that Psychologists do to show people how stupid they are, they actually reveal something important about how we think, and how our use of mental shortcuts leads to plausible (but often incorrect) answers.




Research: Kahneman and Tversky

Aim: To investigate how anchors influence thinking and decision making

Procedure

  • Participants spun a wheel with numbers ranging from 1 to 100.  However, the wheel was fixed so that the wheel would always land on either the number 10 or 60

  • Afterwards, participants were asked to estimate what percentage of U.N. member countries were African countries

Findings

  • Participants who spun the number 10 tended to give a significantly lower estimate for African membership in the U.N. than participants who spun the number 60

  • The mean estimate for the "low anchor" group was 25%, compared to 45% for the "high anchor" group

Conclusion

  • The random number had an anchoring effect on participant's estimates for African membership in the U.N., even though it clearly had no relation to the topic

Evaluation

  • This is a well-controlled experiment, demonstrating a clear causal relationship between the independent variable (high or low number spun on the wheel) and dependent variable (estimate for African membership in the U.N.)

  • This is a simple experiment which is easy to replicate.  In fact, many similar experiments have been carried out, and the results are highly reliable

  • Participants in this study were all American university students, so it remains to be seen if these results will also be found in people from different cultures or age groups



Availability Heuristic

Imagine you're on holiday in Australia, and are thinking about learning how to surf.  But you've also heard that Australia is notorious for shark attacks.  In trying to weigh up the risk of surfing in the ocean, what will affect your decision?  According to research on the availability heuristic, people judge the likelihood of an event based on how easily an example or instance of the event comes to mind.  If you can easily think of cases of shark attacks in Australia (perhaps because of recent media coverage of such events) then you'll probably think there's a much higher risk of shark attacks than is actually the case.  In fact, the risk of a shark attack is very remote - thirty times more people die from falling airplane parts each year than from shark attacks!

The availability heuristic explains why people are often particularly fearful of shocking and sensational events, such as terrorist attacks, plane crashes, serial killers, and so forth.  These events tend to be widely reported in the media, leading people to overestimate their occurrence because of the ease of recalling instances of the event.  If someone dies in a shark attack, there are often days of media coverage devoted to the story, whereas there is usually no media coverage when someone dies from a routine calamity such as a heart attack, cancer, or even falling airplane parts.

The availability heuristic causes people to make distorted judgements regarding risk, and this can lead to tragic consequences.  After the terrorist attacks of 9/11, in which airplanes were hijacked and flown into New York's Twin Towers, many people became particularly alarmed about the risks of flying.  In the years after 9/11, there was considerable media attention to the dangers of terrorism on airplanes, and so many people decided to travel by car instead of by plane.  However, the risk of dying in a car accident is actually much higher than the risks of dying in a plane wreck.  In fact, there was an increase of 1,595 car crash fatalities in the U.S. in the year after 9/11, most likely due to the distorted fear of flying.


Research: Kahneman & Tversky

Aim: Investigate how the availability heuristic affects judgement

Procedure

  • Participants were asked, "If a random word is taken from the English language, is it more likely that the word starts with the letter K, or that K is the third letter?"

Results

  • ​Over two thirds (105 out of 152 participants) thought it was more likely that words in English would begin with the letter K

  • In fact, there are about twice as many words in the English language that have K as the third letter than there are words that begin with K

Conclusion

  • This results of this study are likely due to the availability heuristic.  It is much easier to think of words that begin with the letter K (such as kangaroo, kitchen, kidnap) than words that have K as the third letter (such as acknowledge, ask).  Because participants find it easier to recall words that begin with K, they incorrectly assume that there are more such words

Evaluation

  • This is a simple study that is easy to replicate, and the results are highly reliable

  • It could be argued that this study has low ecological validity, as estimating the prevalence of words that begin with a particular letter is an artificial task that would not happen in real life

  • The participants in this study were all American college students, so the results might not generalize to other cultures or age groups

IB Psych Matters

Although some of the research into heuristics may seem rather artificial, many cognitive biases have tremendous impact on real world decision making.  For instance, the anchoring effect plays a very important role in consumer behavior and product pricing.  After all, when you go shopping in the mall, do you know how much things are really worth?

Say you come across a T-shirt that looks cool.  The price of the T-shirt is $12.  Is that a good price or a bad price?  You might find it difficult to decide.  But imagine how your judgement would change if the same T-shirt was advertised with a sign saying, "T-shirts on sale for 75% off!  Limited time only.  Regular price $48, now only $12".  You would probably think you've gotten yourself a great deal, although in fact there is nothing different about either the T-shirt or the price in either situation.

Take a look at the video below, which goes into more detail on the anchoring effect can influence consumer decision making.




Checklist

  • I can explain how System 1 heuristics can lead to cognitive biases

  • ​​I can describe the anchoring effect, giving examples of the effect

  • I can summarize the Aim, Procedure, Findings and Conclusion of a Kahneman and Tversky's research on the anchoring effect, and can Evaluate the study

  • I can describe the availability heurstic, giving examples of the effect

  • I can summarize the Aim, Procedure, Findings and Conclusion of Kahneman and Tversky's research on the availability heuristic, and can Evaluate the study
Quiz Yourself

1.  What are heuristics?

(a) System 1 biases

(b) System 2 biases

(c) System 1 mental shortcuts

(d) System 2 mental shortcuts


2.  If someone asks you, "Are there more or less than 10 countries in the world?" and then asks you to estimate the number of world countries, you will probably...

(a) Underestimate the total number of countries, because you have started from a low reference point

(b) Underestimate the total number of countries, because you have started from a high reference point

(c) Overestimate the total number of countries, because you have started from a low reference point

(d) Overestimate the total number of countries, because you have started from a high reference point


3.  In the United States, many parents are anxious about the possibility their child may be kidnapped by a stranger.  However, there are only around 100 annual cases of child abduction by strangers, whereas over 3,000 children drown in swimming pools each year. What can explain the disproportionate fear of child kidnapping?

(a) The anchoring effect.  Participants start off with a high reference point, and fail to adequately adjust this downwards.

(b) The anchoring effect.  Child kidnappings tend to receive more media coverage than child drownings

(c) The availability heuristic.  Participants start off with a high reference point, and fail to adequately adjust this downwards.

(d) The availability heuristic. Child kidnappings tend to receive more media coverage than child drownings.


4.  In the research by Kahneman and Tversky, how would the availability heuristic explain why two-thirds of participants believe that more English words begin with the letter K than have K as the third letter?

(a) Words that begin with K tend to be shorter, and thus more easily remembered

(b) It is easier to think of examples of words that begin with a particular letter

(c) Participants failed to adjust from an initial reference point

(d) The education system leads to biased decision making


5.  What is NOT a valid criticism of the research carried out by Kahneman and Tversky on heuristics and biases?

(a) Samples are often culturally biased

(b) Results are difficult to replicate

(c) Experimental tasks are often artificial in nature

(d) There are multiple ways to interpret the findings 
 
Answers

1 - C, 2 - A, 3 - D, 4 - B, 5 - B