Test 3 Knowledge Flashcards
Thinking with emotions
Involves many processes and therefor we cannot lose the ability to think we just lose one aspect of thinking.
We often think emotion disrupts thinking and we react in the heat of the moment with our heart not our head.
However we overstate separation between heart and head.
Judgement
the process through we make conclusions from the evidence we encounter.
- we value experience- why we trust an old experienced coach. Or we value situation we won’t take marriage advice from someone married 5 times.
Attribute Substitution
Frequency estimate- how often various events occur
However we often don’t have access to this information
Availability Heuristic- strategy where you rely on easily accessed information as a proxy for information actually needed. If we have one easily recalled example we think it is a common event, a lack of example means it is rare. Example of attribute substitution
Representiveness Heuristic- rely on resemblances to known cases, you look like my ex ew can’t day. Type of attribute substitution.
Availability Heuristic and representativeness errors
Rely on easy access
Usually things in categories are homogenous enough that you can rely on resemblance for category membership
-however strategies can cause error.
Do more words start with R or have R in 3rd place? Easy to find examples of words that start with R since our brain is organized like a dictionary. Makes us think more words in R first place
Wide range of frequency effects
People overestimate frequency of some rare effects. Play part in willingness to buy lottery tickets, overestimate the likely hood of rare diseases
Ignore events that happen a lot but notice rare (especially emotional) ones.
Participants asked to recall 6 incidents, other group asked to recall 12. Then asked to rate how assertive they are. Those who recalled 6 ranked more assertive than those who recalled 12.
-obvious easy example of being assertive/not
Representative Heuristic and Gamblers Fallacy
Categories you encounter are usually homogenous
We expect each individual to resemble other category memories, so if it looks like a bird we can conclude it is a bird
Gamblers Fallacy- if coin is heads 6 times, it is due for a tail. The toss of a coin is independent to previous tosses. However according to category homogeneity a fair coin should be heads and tail 50/50 so we assume that this coin should represent this category
Reasoning from 1 case to whole population
Assume homogenity creates “man who” argument
- I knew a man whose iphone broke means all iphones break
These arguments are persuasive but only through representative heuristic not logically
Detecting covariation
People rely on heuristics for small and big choices. Heuristic error can trigger judgement of covariation.
-if x tends to be present when y is and x tends to be absent when y is.
Exercise and stamina- covary. Education and salary covary. Covary can be strong, weak, positive (exercise and stamina) or negative (exercise and risk for heart disease)
Need to be consider if cause and effect is suspected
Illusion of covariation
Often detect covariation when there is none
-relationship of star signs and personality, social stereotypes, etc.
In these judgements people only consider subset of facts and this subset is skewed by expectation.
-Even a 100% fair judgement is still based on a biased input.
Likely guided by confirmation bias-tendency to be more alert to things that confirm beliefs. This creates a biased sample that causes wrong estimation of covariation.
Base rates
Base-rate information- how frequently something occurs. A neglect of base rate can cause errors in covariation.
-Test new drug for cold, 70% felt better after 48 hours. Can’t interpret unless we know the base rate of time to get over a cold
-Participants told a group of 70 lawyers and 30 engineers- when asked who was more likely to be pulled they said lawyer. However, when a stereotyped individual who met descriptions of engineer, more likely to be an engineer guessed. People ignore base rates when stereotype is offered
Attribute Substitution produces neglect of base rates and can turn question into how well does it represent category.
Dual Process Models
Errors occur even with motivation, clear instructions and rewards to be right.
Human judgement rises over heuristics sometimes, sometimes we seek accurate bases other times inaccurate.
Type 1 and Type 2
Type 1 is fast and easy (heuristics belong here). Occurs in time pressure. Doesn’t necessarily mean sloppy if the environment has good triggers. Can still be sensitive to base rates, but more likely to be neglected if probabilities instead of frequencies. Intuition, association-driven. Assumes evidence is the complete data set.
Type 2- Slow effortful, and more accurate. Occurs when triggered by cues and right circumstances (can focus). Accounts for base rates and frequency, and understands evidence is a smple.
People choose when to rely on what system, however even with incentive people rely on type 1 more.
Role for chance
Fast-but-accurate judgements tend to be more likely if chance is involved. We are likely to realize evidence could be a fluke and people pay more attention to the quantity of evidence. We understand larger observations are less vulnerable to chance.
Participants asked to judge a restuarant based on a review, people accepted that review more than a review that chose their meal by random. By saying the meal was random it reminded them it was only 1 meal and shouldn’t be held responsible for whole meal
Education
Quality of thinking is effected by background info (amount of education)
Participants from week 1 of a stats class were asked to judge a baseball players career based on one season, by the end of the class the number of people who considered and asked about sample size doubled
Cognitive Reflection Task
People make judgment errors a lot, and this task offers an explanation
-consists of 3 questions and each has an obvious answer that is wrong, to pass the test you must bypass obvious and reflect. People who rely on type 2 thinking and avoid most errors.
-higher CRT score correlates with better scores in science, more skeptical, and analytical morally
Confirmation and Disconfirmation
Induction- make predictions for new based on old
Deduction- start with a “given” claim and find it’s consequences
Allows for us to ask what implication do these tasks have, and asks you to keep in touch with reality- if prediction based on beliefs is wrong than there is an error in your beliefs.
Confirmation bias
Umbrella term for tendency to protect your beliefs.
Participants asked to find rule for 2,4,6 by asking if a trio of numbers fit the pattern. Rule was it had to be in ascending order. Only asked for info that confirmed their rule not opposed it.
Asked sports gamblers if they had a good strategy, majority said yes and that their losses were a fluke.
Built in, unavoidable, tendency to prove yourself right.
Belief perseverance
Ignore evidence that disconfirms their belief.
Asked to tell which suicide letter was real, they were given feedback saying they were really good. Then told the feedback was meaningless and all letters were fake. Those who got good feedback still rated themselves highly accurate at finding authenticity.
Perils of balance
Climate change presentation offered balanced presentation, 2 speakers with 2 different view points. But speaker who doesn’t believe represents only 10% of the people. However, people thought the opinions were 50/50 and therefor not ones argument was enough to sway them.
Syllogisms
Thought was theorized to follow the laws of logic. Therefore reasoning error was a result of carelessness or misinterpretation
However, logic errors occur all the time.
Categorical Syllogisms- logical argument with 2 premises that contain a statement.
The syllogism is solved with conclusion made from the premises, valid means it follows them invalid means it doesn’t.
All M are B. All D are M. Therefor all D are B.
Participants asked to reason about syllogisms do bad, in the original study 9% got right, now 10-30%.
Belief Bias
Errors in logical reasoning are systematic.
Belief bias- if syllogism conclusion happens to be something believed to be true they will likely accept premises. Same if false they will reject premises.
-fail to distinguish good arguments from bad ones. Endorse illogical arguments if leads to conclusions they have doubts about.
We look at conclusion only and decide if its logical.
4 Card task
Conclusions derive from research about conditional statements show similar results.
If x, then y.
-selection task/4 card task. Told each card has a number on one side and a letter on the other. Asked to evaluate if card has a vowel it must have an even number. The cards are A, B, 6, 7.
33% flip over A and 6, only a few flip over A and 7. this causes more than 70% to give wrong answers.
-performance better in variation, if person is drinking a beer the person must be 21. 73% flip over drink beer and 16, not drink coke or 22
Both issues ahve the same logical structure but different performance. They both use inductive judgement and deductive reasoning
Both can document high quality of thinkin gif in right circumstances.
Decision making costs and benefits
We all have our own values- things we prize or hope to avoid, and goals- things we hope to accomplish and see. Each decision has costs (carry farther from goals and values) or benefits move closer to goals and values). When we decide we weigh costs against the benefits.
Utilitary Maximization- utility- value you palce on an outcome. Want as much utility as you can.
Framing outcomes
Easy to find we base decisions little on utility maximation.
Asked “option a 200/600 save or B: 33% chance all are saved or 66% all die” 72% select option A
Asked “option A: 400 will die, or 33% none die/ 66% everyone dies.
78% choose option B
some people even contradicted previous answer
If frame casts choice in gains we show risk aversion and hold tight to what we already have
Framing of questions and evidence
If we ask which parent do we award custody to we get a different outcome than which do we deny.
We rate a player with a 75% success higher than a 25% failure. More likely to endorse a drug of 50% success than 50% failure. `
Opt-in vs. Opt out
If ask to opt into donor program (12% registered)
If asked to opt out of program (99% registered)
Endowment Effect- you put higher value on your current status since it is your own.
