Chapter 18

Errors and Biases associated with Representativeness

*we think we have an idea of what random things look like -Gambler's Fallacy -"Hot Hand" -Regression effects

*4 Kahneman & Tversky Heuristics*

-Availability Heuristic -Representativeness Heuristic -Anchoring-and-Adjustment Heuristic -Simulation Heuristic

Groupthink

-Everyone clams up and doesn't express themselves or communicate in order to "keep peace," often with a leader they look up to -- and then it is assumed that everyone agrees since nobody dissented. Introduced by Irving Janis and studied during Watergate and the Challenger Launch. Found that consistently many of these groups that consisted of smart, knowledgeable people made regrettable decisions for very similar reasons. -Symptoms of Groupthink ○ Overestimation of the group; illusion of invulnerability, belief in inherent morality of group ○ Close-mindedness; collective rationalizations, stereotypes view of opponent ○ Strong pressures toward uniformity; self-censorship, illusion of unanimity, direct pressure exerted on dissenters, self-appointed mindguards.

Regression effects

-Just like we attribute meaning to random streaks, we also expect stability of performance attributed to skill. We don't anticipate that after extreme performance, subsequent performance might "suffer" as players regress back to their normal level of performance

How to facilitate Group Creativity

-Leader should encourage discussion, directly ask people for new suggestions -Appoint a devil's advocate to find flaws in what is being discussed.

*Framing effects and Prospect Theory*

-Loss Aversion -Diminishing Sensitivity -Risk-averse vs risk-seeking

Loss Aversion

-Losses have a bigger impact on people relative to equivalent gains. The resulting value function is steeper for losses than for gains; losing $100 produces more pain than gaining $100 produces pleasure. Loss aversion often makes consumers reluctant to accept risk.

Gambler's fallacy

-People tend to persist in gambling even when they lose. The belief that "I'm due" or "My luck is going to change"

Risk-aversive vs risk-seeking

-Risk averse when it comes to gains. Basically, we get whatever we can. But willing to risk a bigger loss (risk-seeking) if it means we might lose nothing.

Common knowledge effect

-Stasser has shown that in group discussions people focus on shared qualities as opposed to unshared information. In creating group harmony, we neglect to take advantage of individual expertise & minority opinions.

'Hot Hand"

-Streaks are assumed to have some special meaning or cause, since they don't fit our perceptions of "random". Representativeness leads us to see a "hot hand" effect when it isn't there

Dilution effect

-The judgement impact of diagnostic information is often diluted by the large amount of information. It is even more pronounced when subjects expect to justify their predictions to others. -Even if given a piece of unrelated, bogus piece of information, we will still try to relate it. Dilutes the effect of the important information. Ex) Saying someone is a vegetarian is irrelevant to considering them for a job - dilutes relevant characteristics

Representativeness Heuristic

-We categorize objects based on how well an instance matches the typical member of a category (we say someone who is studious is more likely to be a librarian than a farmer, ignoring the fact that there are many more farmers than librarians)

Group Decision Making

-We tend to assume that groups will make better decisions than individuals. We subscribe to this philosophy in most important decisions, tend to value harmony more than progress. Ex. Juries, committees, President's cabinet

Availability Heuristic

-We use the ease with which instances can be brought to mind to determine the frequency or probability of an event (ex: shark attacks, plane crashes)

Simulation heuristic

-When judging the likelihood of a possible event, we use the ease with which we can mentally simulate the occurrence of that event, different possibilities -Events easy to imagine → judged more likely -Events difficult to imagine → judged less likely -Process of mentally imagining stuff that could happen (Different from availability because that refers to things that are already stored in memory)

Anchoring-and-adjustment heuristic

-When making judgments based on several pieces of information, we tend to let the starting value serve as an anchor and then do not adjust sufficiently from that anchor with later information. -This heuristic means the earliest pieces of information receives greater weight than any subsequently received information

Perseverance Effect

-after people are told something about themselves, they come up with reasons to justify it. Even if they are told the initial fact was totally wrong, people still tend to believe what they were originally told (because they now have self-generated reasons for it to be true)

Group polarization

-group decisions tend to be more extreme than individual decisions. Explained by common knowledge effect.

Subtraction-based contrast effect

-information is excluded from the representation of the target. As the amount and extremity of the favorable information excluded from the representation of the target increases, less favorable judgments of the target are formed.

Comparison-based contrast effect

-information is included in the representation of the reference point or standard of comparison. As the amount or extremity of the favorable information included in the representation of the standard increases, less favorable judgments of the target are formed

Premature Cognitive Commitment

-once we make a choice we tend to go with it and not want to over-complicate things. High need for closure

Diagnosticity

-our sense of how important/useful a piece of information is

Contrast effect

-shift in judgment of the target away from the reference point (ambiguity, typicality, and similarity are low). Information that is excluded from the representation of the target results in contrast effects.

Assimilation effect

-shift in judgment of the target toward the reference point (ambiguity, typicality, and similarity are high). -Information used to form a representation of the target results in assimilation effects: as the amount of favorable information used increases, more favorable judgments of the target are formed.

Biased assimilation

-tendency to interpret info in a way that supports a desired conclusion -Supporting facts seem stronger and negating facts automatically seem weaker -Similar to confirmation bias -- but we also are critical of opposing viewpoints

*Correspondence bias*

-we almost always make personal attributions for behavior even when we realize that there were strong situational reasons for acting that way -fundamental attribution error -people in collectivist cultures don't do this as much and are able to take into account situational constraints

Diminishing Sensitivity

-we are more responsive to changes when the amount was already small Ex/ $0 to $100 seems to be a greater change than than $1000 to $1100

overcorrection

-we don't know how much we should correct for biases

Pseudodiagnosticity

-we tend not to think about alternatives beyond our current hypothesis Ex/ "if you do X, business will improve" but people don't ask if business would improve if they did nothing

Selective Hypothesis Testing

-we tend to focus on one hypothesis and not think about alternatives or other possibilities. A problem many of us have in our decision making

Preference Reversal

-when people are inconsistent, -i.e., sometimes risk averse and sometimes risk seeking. People prefer one option at one point in time, but can prefer another option at another point in time. Contrary to standard economic theory (violates principles of transitivity, dominance, invariance) -It depends a lot on word choice/framing. Ex) Which is more valuable? Vs. Which is more attractive?