Cognitive Biases

Cognitive Biases are instinctive ways we think and see things, and thus behave, which are sometimes inappropriate in particular circumstances. They are a major source of poor thinking in both ourselves and others and are exploited by scammers and people looking to take advantage of us. By being better aware of cognitive biases we can significantly improve the quality of our thinking and be less susceptible to being fooled or taken advantage of.

 

Scrollbar, Popup, and Listing of different types of Cognitive Biases | Base Rate Bias

More about Cognitive Biases? | Test Yourself to Improve your understanding | Useful Sites for better understanding Cognitive Biases

 

Scroll Bar Listing of different types of Cognitive Biases

Scroll through here for a brief description of common cognitive biases, with a particular bias, the Base Rate Bias being separately described because of the need for a bit of arithmetic being required to appreciate it.

 

[Note that you can navigate through with left and right arrows or hold the screen by hovering over it.]

 

PopUp Listing of different types of Cognitive Biases

Or click on any given Bias name to see a pop up of its definition. Use this is a learning tool by seeing if you can remember the gist of the definition before popping it up.

 

| Using Popups

| Actor-Observer Bias | Ambiguity Bias | Anchoring or Focusing Bias | Attention Bias | Attribution Bias | Availability Bias | Bandwagon Effect Bias | The Barnum Effect BiasBelief BiasBenefit BiasBlind Spot BiasEven-Chance BiasChoice Supportive BiasClustering BiasConservativism BiasConfirmation or Preconception BiasCongruence BiasConsistency BiasContrast BiasControl BiasDistinction BiasEarly Evidence BiasEarly Reward BiasEgocentric BiasEndowment BiasExtreme Aversion BiasObserver Expectancy BiasExpectation BiasFalse consensus BiasFamiliarity Bias Framing BiasThe Frequency illusionThe Halo Effect Bias Hindsight BiasAn Illusion of Control BiasImplicit Assumption BiasInformation BiasIngroup BiasA Just World BiasA Negativity Bias The Normalcy BiasOmission BiasOptimism BiasOutcome BiasOverconfidence BiasPartsum BiasPattern Bias Personal Validation BiasPessimism BiasRecall BiasRecency BiasRestraint BiasRisk Framing BiasSecrecy BiasSelective Outcome BiasSocial Comparison BiasUnconscious BiasThe Zeigarnik Effect |

 

 

Base Rate Bias

The base rate bias, which is also known as a Fallacy, under the heading The Prosecutor’s Fallacy, relates to misinterpreting the meaning of a test or detection rate accuracy.

Consider a detection system with a 1% failure rate, whereby it fails to go off when it should 1% of the time, and it goes off when it shouldn’t 1% of the time. Note that most ‘detection’ systems are not 100% accurate, and it is common for them to occasionally ‘detect’ something that is not there, or fail to ‘detect’ something that is there.

Most people in finding the detection system has gone off will assume that the stated failure rate of 99% means that that it is correct 99% of the time. This however is likely to be wrong, and could be badly wrong.

Take the case of whatever it is that is of interest actually only being present 500 times in a million. Note that this is nothing to do with the detection system itself or its accuracy. We might for example be seeking to detect the presence of some rare form of cancer, or seeking to detect the presence of some mineral in a sample we have taken. The likelihood that the cancer is actually present, or the mineral is actually present, is what we term ‘the base rate’. It is a value that exists independent of the means by which we are trying to detect its presence.

Suppose, for illustration purposes, a million tests were undertaken.

Then with there being 500 ‘true’ occurrences in this million, the detection system, with an accuracy of 99%, will detect (on average) 495 of these ‘true’ occurrences (ie. 99% of 500).

It will fail to detect 5 of these ‘true’ occurrences, ie. 1% of the 500.

In the other 999,500 cases, where there is no ‘true’ occurrence, the detection system will believe there is an occurrence in 1% of the cases, ie. it will be believe there is an occurrence in 9,995 cases.

For 99% of the 999,500 cases it will correctly identify there is no occurrence.

Thus, of the 1 million cases, there will be 495 ‘true’ occurrences, but also 9,995 cases where the detection system believes there is an occurrence. Thus the true accuracy of the detection system in detecting true occurrences is 495/(9995+495), ie. a little under 5%. This is clearly a long way from the 99% that many people will assume based on the supposed ‘99%’accuracy of the detection system.

Appreciating and being able to calculate the implications of base rates is often aided through creating a table of the form of the following:

 

Has occurred Has not occurred
Test positive 495 9,995 Thus there is only approximately a 5% chance of a detection actually being correct.
Test negative 5 989,505 The chance that a negative test is correct is approximately 99.9995%.
500 999,500

 

Note:

495 is 99% of 500 [500 being the (on average) number of ‘true’ occurrences in a million]

5 is 1% of 500 [ie. 1% chance that a ‘true’ occurrence is not detected]

9,995 is 1% of 999,500 [ie. 1% chance that we incorrectly believe we have an occurrence when in fact we don’t, with 999,500 being the number of ‘non-occurrences’ that we have in a million.]

989,505 is 99% of 999,500 [ie. the 99% chance that we correctly recognise we don’t have an occurrence.]

 

This base rate error is most prominent when the accuracy of the detection system is much worse than the ‘base rate’ of whatever is being measured. However:

  • Even if the detection system has the same accuracy as the base rate of what is being measured, there is still a 50% chance that a detection is in fact false.
  • If the detection system is 10 times more accurate than the base rate of what is being measured, there is still a 10% chance that a detection is false.
  • Only if the detection system is 100 times more accurate than the base rate of what is being measured does the chance of a false detection fall to about 1%.

 

More about Cognitive Biases?

Cognitive biases happen unconsciously. Unless we learn about them and become sensitive to them, we are not aware of them when they happen.

Our brains have become susceptible to cognitive biases because in our distant pass they helped us survive. Our ability to react quickly and instinctively would have in certain circumstances saved our lives and those with brains which reacted in such a way would have been more likely to survive and pass on these brain characteristics to their offspring. However our environment has changed and these biases are no longer required for survival and can often lead us astray.

Cognitive biases don’t necessary lead us astray. We may, based on ‘instinct’, jump to a conclusion about a person or a circumstance, and we might be right. But sometimes we won’t be. Moreover it will be largely a matter of chance whether or not in a given circumstance we are right based solely upon our instinct. Our instinct can’t differentiate between what is really right or what is really wrong, and sometimes it will be wrong when it really matters.

Cognitive biases arise largely as a result of our brain seeking to maintain a positive self-image so we feel good about ourselves, and our tendency to be lazy and want to do things, including thinking, with the minimum of effort. Most people are unaware of their own cognitive biases, though readily see them in others.

By allowing these Cognitive Biases to dominate our thinking we are allowing ourselves to make incorrect judgments, and poorer decisions than we might otherwise have done.

By recognizing our own cognitive biases, and being able to correct for them, we will be much better thinkers, much more able to draw correct conclusions and take appropriate action. By understanding and being able to label given cognitive biases, we will be much more able to recognize shortcomings in other people’s thinking and make allowances accordingly.

Note that a poor understanding of cognitive biases, or a naïve view that we are somehow magically immune to them, leaves us open to the influencing and persuading of others, whether by salesmen, general advertising or through being targeted by scammers.

In practice even when we recognize our cognitive biases it can be different to overcome or correct for them. However we usually can to some extent, and can usually come to much better conclusions than we would if we hadn’t tried.

The greatest barrier to countering our own cognitive biases is our belief that they don’t apply to us. And thus the most important step towards overcoming them, or at least significantly lessening their hold and impact on us, is to accept without question that we too are susceptible to them, in the same way as we readily see that others are.

Once we accept that we are as readily susceptible as others, itself a vital first step if we are to mitigate the impact of cognitive biases upon us, we can lessen their impact by:

  • being clear about what is clearly facts and what is opinion or assumption, with far more being the later than we may otherwise have realized;
  • always recognise that you might be wrong or mistaken, that you don’t magically know anything for certain;
  • don’t rely upon your memory if you want accurately remember something, write it down as objectively as you can before anyone can influence your memory;
  • actively listening to or seeking out alternative viewpoints and opinions;
  • being aware that others are also susceptible to biases, including experts and friends;
  • being aware that circumstances play a major part in what people do and decide, and that often you may not be aware of those circumstances;
  • avoiding making important decisions when tired or after you’ve been working hard for an extended period of time; and
  • ensuring you are not rushed into making up your mind or rushed into accepting what others are forcing upon you.

The main types of cognitive biases are well known, and indeed have been known about since the times of the Greeks and Romans and probably before. It is very useful to know the names of the biases, which are for the most part sufficiently descriptive that we will then readily be able to remember the basis of the bias. The training we include here is thus largely about teaching you to recognize the names of the common biases and be able to associate them with their meaning. We also give tips on how you can try to compensate for these biases should you recognize you may potentially be subject to it.

By regularly running through the training below, and regularly returning to the related exercises, you will become more aware of these Cognitive Biases, and better able to stop and think in a given circumstance to ask whether or not it is inappropriately affecting your thoughts and judgements.

 

Useful Sites for better understanding Cognitive Biases

http://lesswrong.com/

http://www.howtogetyourownway.com/

https://www.schoolofthought.org/

https://en.wikipedia.org/wiki/Cognitive_bias

http://rationalwiki.org/wiki/List_of_cognitive_biases

If you know of other sites you believe are particularly good at covering cognitive biases, or have good free cognitive bias related tests or quizzes, please let us know:

 

Quiz/Tests

Test Yourself to improve your understanding:

Reminder on taking tests: It’s not about trying to prove you already know it, it’s about learning.

 

Cognitive Biases -> Just to get you Thinking

Question Think.1

Susceptibility to Cognitive Biases. Which of the following do you consider to be closest to the truth:

a. It is clear many other people are susceptible, but we ourselves are much less so.
b. Susceptibility to Cognitive Biases is largely a matter of intelligence. The cleverer someone is the less susceptible they are.
c. We are all equally susceptible to Cognitive Biases.
d. Having read about susceptibility to Cognitive Biases we become largely immune to them.
e. We can learn to lessen our susceptibility to Cognitive Biases by learning about them.
f. We can learn to eliminate our susceptibility to Cognitive Biases by learning about them.

 

Question Think.2

Cognitive ease refers to our brain being in an automatic mode, accepting things as they are and not questioning them. In such a state we are more likely to fall for cognitive biases. Which of the following are conducive to putting us in a state of cognitive ease:

a. Being in a good mood.
b. Reading something written very clearly.
c. Something being repeated.
d. Familiarity.

 

Question Think.3

People do evil things because they are evil people. Is this a reasonable thing to believe?

 

 

Cognitive Biases -> Name that Bias

Question Name.1

The belief that we are less prone to biases than most other people.

a. Belief Bias
b. Blind Spot Bias
c. Egocentric Bias

 

Question Name.2

Over 90% of drivers rate themselves as better than average.

a. Control Bias
b. The Halo Effect Bias
c. Overconfidence Bias

 

Question Name.3

People often see their role in things as more prominent than they actually were.

a. Personal Validation Bias
b. Egocentric Bias
c. Recall Bias

 

Question Name.4

Significantly underplaying the seriousness of extreme situations

a. Normalcy Bias
b. Optimism Bias
c. Overconfidence Bias

 

Question Name.5

The tendency to see two items as more dissimilar when evaluating them together than when evaluating them separately

a. Recall Bias
b. Familiarity Bias
c. Distinction Bias

 

Question Name.6

When we look at a situation from a particular viewpoint and fail to recognize there are other viewpoints

a. Omission Bias
b. Framing Bias
c. Familiarity Bias

 

Question Name.7

The tendency to overestimate how much other people agree with us.

a. Base Rate Bias
b. Conservativism Bias
c. False consensus Bias

 

Question Name.8

We see ourselves as flexible and responsive to circumstances but see others as fixed in their ways.

a. Actor-Observer Bias
b. Attribution Bias
c. Blind Spot Bias

 

Question Name.9

People tend to notice risks or threats more readily than they notice opportunities.

a. Negativity Bias
b. Pessimism Bias
c. Choice Supportive Bias

 

Question Name.10

A tendency to view a person or thing favorably, or unfavorably, based on a single incident or trait.

a. Omission Bias
b. Recall Bias
c. Halo Effect Bias

 

Cognitive Biases -> Choose a Definition

Question Defn.1

Attribution Bias.

a. Judging the behaviour of others differently to how we judge our own.
b. Our tendency to be overconfident in our judgments.
c. When we see what we expect to see even when it is not there

 

Question Defn.2

Just World Bias

a. Our tendency to see harmful actions as being worse than equally harmful inactions
b. Chance events even themselves out.
c. Victims, have, in some way, only themselves to blame.

 

Question Defn.3

Confirmation Bias

a. The tendency to see two items as more dissimilar when evaluating them together than when evaluating them separately
b. The ignoring or undervaluing of evidence that is contrary to what we already believe.
c. Our tendency to place more value on something we own than on something we don’t own

 

Question Defn.4

The Barnum Effect Bias

a. A readiness to accept statements supposedly tailored to us as highly accurate even when they are statements are vague and general enough to apply to a wide range of people.
b. A bias towards wanting to complete unfinished tasks.
c. A parent’s claim they are right because they are the parent and the other is the child.

 

Question Defn.5

The Zeigarnik Effect

a. A readiness to accept statements supposedly tailored to us as highly accurate even when they are statements are vague and general enough to apply to a wide range of people.
b. A bias towards wanting to complete unfinished tasks.
c. A parent’s claim they are right because they are the parent and the other is the child.

 

Question Defn.6

Hindsight Bias

a. When something turns out well, we praise ourselves even though the outcome may have been almost entirely due to luck.
b. Given any event, after it has occurred most people seemed to have ‘known it all along,’ even though few of them seem to have been so definitive prior to the event occurring.
c. Our tendency to see patterns even where they don’t exist.

 

Cognitive Biases -> The Clue is in the Name

Many biases take their name fairly obviously from their definition. Identify the name of the biases related to the following definitions:

Question Clue.1

Which bias is being described:
When only the results that favour a give outcome are used, and those that don’t are suppressed. [If you want to ‘prove’ a coin is biases towards heads, undertake hundred samples of coin tosses each with a hundred tosses. Then present only the sample which included the most head tosses and apply a statistical significance test to ‘prove’ that the result was highly unlikely to occur unless the coin was biased.]

 

Question Clue.2

Which bias is being described:
When our attitudes or beliefs change we tend to alter our memories to be consistent with our changed attitudes or beliefs.

 

Question Clue.3

Which bias is being described:
When people overestimate probabilities based on the fact that they can bring particular examples to mind.

 

Question Clue.4

Which bias is being described:
Given two rewards, people show a preference for the one expected to arrive sooner rather than the one expected to arrive later, even if the one arriving later is worth more. We ‘discount’ the value of the later reward; the more so the longer the delay. Of course this assumes the later reward is not significantly disproportionate to the earlier reward.

 

Question Clue.5

Which bias is being described:
When something that has recently come to our attention then appears to be everywhere.

 

Question Clue.6

Which bias is being described:
Our tendency to be overconfident in our judgments. For example over 90% of drivers rate themselves as better than average. People who were 100% confident in their spelling of certain words were in fact only right 80% of the time. Most people are overconfident in their ability to accurately assess risks.

 

Question Clue.7

Which bias is being described:
Where more recent information tends to push out older information, and thus we are more likely to be influenced by it.

 

Question Clue.8

Which bias is being described:
To someone who appreciates wine, a more expensive wine will actually taste better to them because they expect more expensive wines to taste better.

 

Cognitive Biases -> Matching Names to Definitions

Test your ability to recognise different biases with the following Quiz, with each bias being from the following list:
Attribution Bias / Pattern Bias / Availability Bias / Belief Bias / Consistency Bias / Distinction Bias / Endowment Bias / False Consensus Bias / The Frequency illusion / Hindsight Bias / Ingroup Bias / Omission Bias / Optimism Bias / Overconfidence Bias / Restraint Bias / Unconscious Bias.

Cognitive Biases Quiz

Testing knowledge of what different types of Cognitive Bias are well known.

 

Cognitive Biases -> Attribution Bias

Question Attrib.1

Part 1: You are walking along and you momentarily trip and knock into someone. What is the likely cause?
a. Something was on the ground.
b. There was some unevenness on the ground that was just enough to upset your balance.
c. Your mind was elsewhere.
d. It’s the result of your general clumsiness.
[Choose one of a. to b.]

Part 2: You are walking along and someone knocks into you. What is the likely cause?
a. There was something on the ground.
b. There was some unevenness on the ground which caused them to lose their balance.
c. Their mind was elsewhere.
d. It’s the result of their general clumsiness.
[Choose one of a. to b.]

 

Question Attrib.2

Part 1: Think about any mistakes you might have made or been accused of making over the past year or so. Assign a general probability to each of the following, adding up to a total of 100%.

a. I don’t make mistakes. Whilst it may have appeared to others to be a mistake in fact things turned out as I intended.
b. I made a balanced judgement based upon the best information at the time, which turned out to be wrong.
c. I was distracted by other things and didn’t pay enough attention.
d. It would have been ok, but other people made it worst or prevented things turning out the way I intended.
e. I am prone to making mistakes.
f. Some other option.
[Assign probabilities to each of a. to f., such that the total of the probabilities is 100%. This is not intended to be precise, and try to avoid use of f.]

Part 2: Think about mistakes others in general might have made over the past year or so. Again assign a general probability to each of the following, adding up to a total of 100%.

a. They probably weren’t mistakes. It was the way they probably wanted things to turn out.
b. They made a balanced judgement based upon the best information at the time, which turned out to wrong.
c. They were distracted by other things and didn’t pay enough attention.
d. It would have been ok, but other people made it worst or prevented things turning out the way they intended.
e. They are prone to making mistakes.
f. Some other option.
[Assign probabilities to each of a. to f., such that the total of the probabilities is 100%. This is not intended to be precise, and try to avoid use of f.]

 

Cognitive Biases > Base Rate Bias

Question BaseR.1

A test for a disease has a 99.9% accuracy rate. You are one of a population of 100 million who each have an equal chance of having the disease, and some 10,000 people in this population have the disease. You are tested for the disease, and the test shows you have the disease. What is the likelihood that you do in fact have the disease? (You don’t need to be precise, just approximate.)

 

Question BaseR.2

A test for a disease has a 99% accuracy rate. You are one of a population of 50 million who each have an equal chance of having the disease, and some 10,000 people in this population have the disease. You are tested for the disease, and the test shows you have the disease. What is the likelihood that you do in fact have the disease? (You don’t need to be precise, just approximate.)

 

Question BaseR.3

A test for a disease has a 99.9% accuracy rate. You are one of a population of 50 million who each have an equal chance of having the disease, and some 50,000 people in this population have the disease. You are tested for the disease, and the test shows you have the disease. What is the likelihood that you do in fact have the disease? (You don’t need to be precise, just approximate.)

 

Cognitive Biases -> Quiz and Test Yourself Sites

The following sites have also got quizzes and tests you can use to improve and test your understanding of cognitive biases.

http://www.sporcle.com/games/popestcyril/12-cognitive-biases-that-keep-you-from-being-rational

http://www.clearerthinking.org/

 

[Back to Top of Cognitive Biases]
[Do Quiz/Test Again]
[Back to Quiz and Test List]

Leave a Reply

Your email address will not be published. Required fields are marked *