PSYC333 TEST 1 Flashcards
What is Anchoring?
People estimate probabilities based on the anchor: initial value provided to them.
(A) Anchors exert “drag”
(B) Leads to insufficient adjustment
Describe Kahamen and Tversky’s (1974) study and what their findings tell us?
Participants were asked: What is the percentage of African countries in the UN?
First participants spun a “wheel of fortune” which would either land on the number “65” or “10”. Once it stopped spinning participants were asked if they though the correct answer was greater or less than this random value.
They found that if it landed on 10 people tended to guess 25% and if it landed on 65 people tended to guess 65%. The difference in estimates highlights the anchoring effect where a numerical value that was chosen at random and couldn’t tell us anything useful still exerts drag over peoples judgements causes insufficient adjustments.
What did Waskin, Kent and Hoch (1998) study tell us?
Waskin et al.’s (1998) study took the anchoring effect found in the laboratory and wanted to test if this effect is applicable to consumer behaviour. Their study looked at consumer behaviour in a supermarket looking at the effects of sale signs on consumer behaviour compared to weeks where the same items were not on sale,
(A) 1 for $1.00
(B) 4 for $4.00
They found in both conditions consumer behaviour went up when items were on sale however, sales went up more when the sign said 4 for $4.00 even though the number 4 doesn’t mean anything or give us information about the product but the mere presence of the number exerts influence on our purchasing decisions.
Does the anchoring effect only apply to numerical values?
No, a study by Oppenheimer, Lebeuf and Brewer (2008) found that the same anchoring effect in line sizes. In their study they asked participants to trace a series of lines that were (all) either long or short. Then participants were asked to estimate the length of the Mississippi river. They found that the average estimate when asked to trace long lines was 1224 miles and 720 miles for short lines. Highlighting that anchoring is an effects that occurs across modalities- I know it’s a weird study.
How can we get rid of anchoring?
You cant….
> Incentives i.e. you will earn more money the more accurate you are=doesn’t work.
>Forewarning people of anchoring effect and biased judgements=doesn’t work
>Size of anchor, giving estimates that are so extreme that they couldn’t possibly be accurate=doesn’t work judgments are still biased albeit not as much as when people are given plausible anchors.
Which study looked at the size of the anchor or extreme initial values?
Quattrone et al. (1984) where participants were asked: Is the number of Beatles records which made it into the top 10 greater than or less than 100,025?
People still showed an anchoring effect- insufficient adjustment from the initial value. Even though the answer was so extreme it couldn’t possibly be real.
Anchoring as a Process: Name the (3) explanations proposed in-class to account for this phenomenon.
(1) Confirmation Bias
(2) Satisficing
(3) Self-Generated Anchors
How could the confirmation bias account for anchoring?
Mussweiler & Strack (1999) theorised that people tend to insufficiently adjust from initial value provided to them because they believe it may be correct and have a tendency to seek information which confirms their belief and ignore any contradictory information.
For example:
Participants were told each of the following cards will have a number on one side and a letter on the other. They were then told a rule “If the card has a vowel on one side the other will be an even number”. What cards would you need to turn over to tell if they were lying?
Note cards shown: E, K, 7, & 4.
Most people answer E & 4. This is the incorrect answer because these 2x cards can only confirm the rule, not disprove it.
The correct answer is E & 7. These are the only 2x cards which could disconfirm the rule.
i.e. the rule mentions vowels and even numbers…
If we flip the vowel over and there is an odd number on it the rule is not true.
Similarly, if we flip the 7 over and there is a vowel on it the rule is not true.
Evidence of confirmatory hypothesis testing, that people choose or weight information which confirms their hypothesis and ignore information which disconfirms it.
What is the Satisficing theory of anchoring?
Satisficing theory was used to account foe Quattrone et al.’s study for which confirmation bias could not account for.
Satisficing argues that people adjust their estimate from the anchor until the hit a value which falls within their range of acceptable answers.
Answer is good enough.
e.g. Was Thomas Aquinas born before or after 800 AD?
What study of anchoring can confirmation bias not account for?
Satisficing theory was used to account for Quattrone et al.’s study for which confirmation bias could not account for.
What is the self-generated theory of the anchoring effect?
That insufficient adjustments occur because people are provided initial values by the experimenter. Epley and Gilovich argued that in real life people spontaneously generate anchor’s themselves to guide their estimates which can be beneficial when we need to make decisions under conditions of uncertainty.
e.g. What is the freezing point of Vodka?
People the freezing point of water which is 0 celsius as a reference and adjusts their estimates from this initial value. People tended to be quite accurate and close to the correct answer which was -20 celsius but their was still a small effect of anchoring.
What theory accounts for why self-generated anchors can still produce anchoring effects?
Epley & Gilovich claim that adjustments from the anchor require EFFORT. Thus, people tend to use SATISFICING to reach an acceptable value and optimise their limited capacity.
Two studies which proved adjusting from self-generated anchors require effort?
sober > drunk No load > cognitive load and S=D NL=L
Do adjustments from experimenter-provided anchors require effort?
No, people do not put effort in they tend to use satisficing regardless of their cognitive load or capacity.
Epley & Gilovich: removing anchoring effect on probability estimates with self-generated anchors:
Incentives > No Incentives Forewarning > No Forewarning and I=NI F=NF
Anchoring effects is larger when:
(A) Higher Ambiguity (uncertainty)
(B) Lower familiarity (of knowledge)
(C) More trustworthy source (conversational maxim)
(D) Plausible anchors (confirmation bias)
Self-Generated anchors are effective when:
(A) People use effort
(B) People are motivated
(C) Aware of biases
People show anchoring even when:
(A) Anchor is chosen at random
(B) Anchor size is extreme
(C) People use incentives or forewarning
What is the Homo Economicus model?
An economic model which describes hypothetically what an economic man would be:
> Self-Interested
> Maximises utility and profit
> Is rational
What is the Expected Utility Theory?
Model of rationality, outlines (6) principles on what an economic man should do when choosing between alternatives in order to maximise utility:
- Ordering alternatives (preference and indifference)
- Cancellation (decision-based on differences)
- Dominance (pick the dominating option)
- Transitivity (If A>B, and B>C then A>C)
- Continuity (prefer a gamble of best and worst outcome over a guaranteed mediocre outcome)
- Invariance (framing shouldn’t matter)
What model opposes the Homo Erectus model?
Bounded Rationality. argues that rarely do people have all the information and capacity to be fully rational when making decisions. Thus, they argued a more descriptive model should be used to explain how people actually make decisions not how the should.
Key aspect of this model is “satisficing”-people tend to be content with finding an answer which is good enough i.e. optimising their limited capacity.
i.e. evidence of this is heuristics-where people violate the 6 principle of the expected utility theory.
Examples of Heuristics that violate rational decision making:
(A) Representativeness
e.g. conjunction fallacy, hot hand phenomena and is jack an engineer?
(B) Availability Heuristic
e.g. famous males & less famous females
(C) Simulation Heuristic
e.g. Easy to imagine symptoms and hard to imagine symptoms.
Steps of the PATH Model:
Problem
Analysis
Test
Help
Mental Accounting
How people code, categories, and evaluate possible outcomes.
There are (3) non-fungible mental accounts:
> Current Incomes
Current Wealth
Future Income