A Framework for Deception Draft Report

Thomas Gilovich

Thomas Gilovich [14] provides in-depth analysis of human reasoning fallibility by presenting evidence from psychological studies that demonstrate a number of human reasoning mechanisms resulting in erroneous conclusions. Them mechanisms identified include;

Mechanism Example
(0) Effects should resemble their causes,
(0a) Instances should resemble their categories Similar looking animals must be more closely related genetically than different looking ones.
(0b) Like resembles like Measles come from germs with spotted coatings.
(0.5) Tendency toward oversimplification,
(0.5a) Occum's Razor When a simple explanation will do, choose it over the more complicated one.
(0.5b) Black and White Tends to be preferred over shades of gray.
(0.5b) Rule of 3s Lists of three things are better accepted in some cultures.
(1) the misperception of random events,
(1a) the clustering illusion, Events appear to be correlated even when they are not correlated
(1b) overapplication of representativeness, The 'law of small numbers' - a few examples are taken as more signficant than they really are.
(1c) misperceptions of random dispersions, Various random events are seen as 'shooting streaks' because randomness is not well understood by most observers.
(1d) the creation of casual theories, People have a tendency to create thearies to explain what they see and adopt them regardless of evidence.
(1e) the regression fallacy, People underestimate the effect of regresson. For example, if you usually average two sales a day and make five sales for each of three days in a row, people will think you are in a slump when you only make one or two sales a day for the next week.
(2) misinterpretation of incomplete or unrepresentative data,
(2a) the excessive impact of confirmatory information, A small number of confirmations are treated as proof, while an occasional refutation may be dismissed as invalid for some posteriori reason (perhaps generated 1d above).
(2b) the tendency to seek confirmatory data, If you are looking for red in fires you will tend to count orange as red, and not discount the presence of blue along with red.
(2c) the problem of hidden or absent data, If you justify the quality of your hiring process by tracking only the success rates of people you hire, you are ignoring the missing data on how successful the people you didn't hire might have been.
(2d) self-fulfilling prophecies, If people believe the markets are crashing, they will pull their money out, and thus the markets will crash.
(3) the biased evaluation of ambiguous and inconsistent data,
(3a) ambiguous information is interpreted in context, We tend to interpret ambiguous data in the context of what we are looking for.
(3b) unambiguous data is shaded, An explanation for the invalidity of data that is inconsistent with theories is often found.
(3c) multiple endpoints, If the data is ambiguous we will tend to associate it with our expectations for outcomes, thus biasing the result. For example, some element of a baby's face looks like anyone and will be associated with the parents face even if the child is adopted.
(3d) confirmations and non-confirmations, Non confirmations are often ignored rather than treated as refutations. Selective memory is an example where people will tend to remember predictions that come true over time and forget those that do not come true.
(3e) focused and unfocused expectations, If we believe that bad things come together in threes but don't set a time limit on what it is to come together, we will wait till the count hits three and declare that we were right. If we are trying to associate a dream of a sunny day with events of the day, we will find the moment that the sun broke through the clouds as a confirmation.
(3f) outcome asymmetries and one-sided events,
(3f-i) hedonic asymmetries, There is a tendency to overemphasize things that are more striking to us. For example, it may seem like you almost always get splashed by a passing car on wet days, when in fact you just remember being splashed more than not being splashed.
(3f-ii) pattern asymmetries, You remember when you wake up and see 1:11 or 2:22 on the clock better than when you see 1:52 or 2:17
(3f-iii) definitional asymmetries, Things won't get better till you have hit rock bottom - but since 'rock bottom' is not pre-defined, it is always able to be true since we can call wherever you turned around, 'rock bootom'.
(3f-iv) base rate departures, "Thinking about being healthy will help you cure cancer" is supported by people who have thought about being healthy and survived, but it ignores the people who thought about being healthy and died, because they are not available as data points.
(4) motivational determinants of belief,
(4a) empirical support for the wish to believe, After the Nixon / Kennedy debates, supporters for each side thought their side had one. They interpreted the same thing in different ways.
(4b) mechanisms of self-serving beliefs, If you want to believe it you ask "Can I believe it" while if you don't want to believe it you ask "Must I bnelieve it".
(4c) optimistic self-assessment The vast majority of people believe they are above average in intelligence and beauty.
(5) the biasing effect of second hand information
(5a) sharpening and leveling, In relaying situational information, descriptions of peoples' behavior tends to be 'sharpenned' or emphasized, while descriptions of their surroundings tend to be 'leveled' or de-emphasized.
(5b) the corrupting effect of increasingly indirect evidence, The game of 'telephone' is a great example.
(5c) telling a good story, In order to make the story interesting to the audience, distortions are often introduced. The 'historical movies' that come out of hollywood are examples of how telling a good story often distorts facts in favor of 'flavor'.
(5d) distortions in the name of informativeness, Stories are often told with exagurations of the fact to make a point. A little girl down the block did than and she was never seen again...
(5e) distortions in the name of entertainment, 'There is one example of...' becomes 'I had a friend who...' and the audience misinterprets it as if their own friends probably... Inquiring minds want to know... The media is notorious for this.
(5f) distortions in the name of self interest, Look at the statements of political parties.
(5g) distortions due to plausibility, So-called urban legends are good examples of this - for example the non-existent US patent agent who supposedly resigned because he thought that nothing else could be invented.
(6) exaggurated impressions of social support,
(6a) social projection and the false consensus effect, Most people think that most other people agree with them about their views on things.
(6b) inadequate feedback from others. People may agree out of politeness or not indicate that they disagree because of a desire not to offend. Children show less of this than adults.

Mechanism	Example
(0) Effects should resemble their causes,
(0a) Instances should resemble their categories	Similar looking animals must be more closely related genetically than different looking ones.
(0b) Like resembles like	Measles come from germs with spotted coatings.
(0.5) Tendency toward oversimplification,
(0.5a) Occum's Razor	When a simple explanation will do, choose it over the more complicated one.
(0.5b) Black and White	Tends to be preferred over shades of gray.
(0.5b) Rule of 3s	Lists of three things are better accepted in some cultures.
(1) the misperception of random events,
(1a) the clustering illusion,	Events appear to be correlated even when they are not correlated
(1b) overapplication of representativeness,	The 'law of small numbers' - a few examples are taken as more signficant than they really are.
(1c) misperceptions of random dispersions,	Various random events are seen as 'shooting streaks' because randomness is not well understood by most observers.
(1d) the creation of casual theories,	People have a tendency to create thearies to explain what they see and adopt them regardless of evidence.
(1e) the regression fallacy,	People underestimate the effect of regresson. For example, if you usually average two sales a day and make five sales for each of three days in a row, people will think you are in a slump when you only make one or two sales a day for the next week.
(2) misinterpretation of incomplete or unrepresentative data,
(2a) the excessive impact of confirmatory information,	A small number of confirmations are treated as proof, while an occasional refutation may be dismissed as invalid for some posteriori reason (perhaps generated 1d above).
(2b) the tendency to seek confirmatory data,	If you are looking for red in fires you will tend to count orange as red, and not discount the presence of blue along with red.
(2c) the problem of hidden or absent data,	If you justify the quality of your hiring process by tracking only the success rates of people you hire, you are ignoring the missing data on how successful the people you didn't hire might have been.
(2d) self-fulfilling prophecies,	If people believe the markets are crashing, they will pull their money out, and thus the markets will crash.
(3) the biased evaluation of ambiguous and inconsistent data,
(3a) ambiguous information is interpreted in context,	We tend to interpret ambiguous data in the context of what we are looking for.
(3b) unambiguous data is shaded,	An explanation for the invalidity of data that is inconsistent with theories is often found.
(3c) multiple endpoints,	If the data is ambiguous we will tend to associate it with our expectations for outcomes, thus biasing the result. For example, some element of a baby's face looks like anyone and will be associated with the parents face even if the child is adopted.
(3d) confirmations and non-confirmations,	Non confirmations are often ignored rather than treated as refutations. Selective memory is an example where people will tend to remember predictions that come true over time and forget those that do not come true.
(3e) focused and unfocused expectations,	If we believe that bad things come together in threes but don't set a time limit on what it is to come together, we will wait till the count hits three and declare that we were right. If we are trying to associate a dream of a sunny day with events of the day, we will find the moment that the sun broke through the clouds as a confirmation.
(3f) outcome asymmetries and one-sided events,
(3f-i) hedonic asymmetries,	There is a tendency to overemphasize things that are more striking to us. For example, it may seem like you almost always get splashed by a passing car on wet days, when in fact you just remember being splashed more than not being splashed.
(3f-ii) pattern asymmetries,	You remember when you wake up and see 1:11 or 2:22 on the clock better than when you see 1:52 or 2:17
(3f-iii) definitional asymmetries,	Things won't get better till you have hit rock bottom - but since 'rock bottom' is not pre-defined, it is always able to be true since we can call wherever you turned around, 'rock bootom'.
(3f-iv) base rate departures,	"Thinking about being healthy will help you cure cancer" is supported by people who have thought about being healthy and survived, but it ignores the people who thought about being healthy and died, because they are not available as data points.
(4) motivational determinants of belief,
(4a) empirical support for the wish to believe,	After the Nixon / Kennedy debates, supporters for each side thought their side had one. They interpreted the same thing in different ways.
(4b) mechanisms of self-serving beliefs,	If you want to believe it you ask "Can I believe it" while if you don't want to believe it you ask "Must I bnelieve it".
(4c) optimistic self-assessment	The vast majority of people believe they are above average in intelligence and beauty.
(5) the biasing effect of second hand information
(5a) sharpening and leveling,	In relaying situational information, descriptions of peoples' behavior tends to be 'sharpenned' or emphasized, while descriptions of their surroundings tend to be 'leveled' or de-emphasized.
(5b) the corrupting effect of increasingly indirect evidence,	The game of 'telephone' is a great example.
(5c) telling a good story,	In order to make the story interesting to the audience, distortions are often introduced. The 'historical movies' that come out of hollywood are examples of how telling a good story often distorts facts in favor of 'flavor'.
(5d) distortions in the name of informativeness,	Stories are often told with exagurations of the fact to make a point. A little girl down the block did than and she was never seen again...
(5e) distortions in the name of entertainment,	'There is one example of...' becomes 'I had a friend who...' and the audience misinterprets it as if their own friends probably... Inquiring minds want to know... The media is notorious for this.
(5f) distortions in the name of self interest,	Look at the statements of political parties.
(5g) distortions due to plausibility,	So-called urban legends are good examples of this - for example the non-existent US patent agent who supposedly resigned because he thought that nothing else could be invented.
(6) exaggurated impressions of social support,
(6a) social projection and the false consensus effect,	Most people think that most other people agree with them about their views on things.
(6b) inadequate feedback from others.	People may agree out of politeness or not indicate that they disagree because of a desire not to offend. Children show less of this than adults.

These mechanisms are detailed and supported by substantial evidence and most of them are believed to be common to most individuals in all human societies.