Lecture 19 Flashcards
What’s the difference between judgement and decision making?
Judgements involve using cues to draw inferences about a situation.
Decision making involves choosing what course of action to take in an uncertain world.
Logical (conditional) reasoning is a type of decision making, discuss this
This is the normative theory of human reasoning and there are many types.
Deductive reasoning is when you’re dealing with certainty, it argues from the general to the specific by saying that the conclusion follows from initial hypotheses. The argument is only sound if the premise is true, however, it can still be valid. For example, everyone who eats steak plays rugby. I like steak but I don’t play rugby. An example of abstract deductive reasoning is the Wason selection task to test the rules of the cards.
Inductive reasoning is when you use your common sense/probabilities, which suggests truth as it’s an educated guess, but it doesn’t ensure it. This can produce a true conclusion but it doesn’t always follow from the information in the premises, for example, over-estimations.
Syllogistic reasoning is when you conclude something by combining 2 statements you’ve been given. These conclusions can be correct or incorrect. Therefore, it’s important to detect inaccuracies. Deductive reasoning is kind of the opposite as you work backwards from the conclusion.
Discuss some errors when you do conditional reasoning
Forming incorrect conclusions, search errors like confirmation bias (positive test strategy) and memory-related errors (increased load on working memory can cause reasoning errors). People can often neglect base rates as well.
Why do people often neglect base rates/estimate frequencies wrong when making judgements
Because of heuristics; cognitively undemanding shortcuts that produce approximate answers. They simplify things and are usually correct. They’re fast and frugal and use the phrase ‘take the best, ignore the rest’. They place few demands on memory and cognition.
Representative heuristics; determination of the event based on how similar the event is to previous ones. It ignores base rates/how common things are. For example, thinking a ‘random’ looking order is more likely when tossing a coin but it could be in a ‘structured’ looking order as well. This also accounts for Gambler’s fallacy.
Conjunction fallacy; the mistaken belief that the probability of two events in much greater than one event. For example, a female is running for president. Is she more likely to win or is she more likely to win and have feminist campaigns. People select the latter as more likely even though there is no evidence for this. This happens because they assume she is a female and therefore will have feminist campaigns.
Availability; deciding the likelihood of an event based on how much information is available in their memory. However, our own memory isn’t always perfectly matched with the objective frequencies. For example, saying something is more common because it comes to mind more easily. Also, media coverage can cause this heuristic to occur. So basically this heuristic occurs because of general world knowledge, familiarity and vividness (mundane deaths aren’t considered frequent).
Anchoring; adjusting/using the previous information you have to make a decision, this adjustment is usually bad. Also, when you refuse to let go of previous beliefs and just adapt them to fit. Therefore, the primacy of information given is important.
Hindsight bias; thinking you knew something before being told the fact when you didn’t, e.g. “I knew it all along”.
Are humans always rational?
No because there’s too much information and too many calculations to be rational all the time. This is why we have heuristics.
What are the stages of rational decision making?
- Define the problem
- Search for alternatives
- Estimate costs and benefits
- Optimise/maximise (get the greatest value as a result)
However, people often just do satisficers which are ‘good enough’ and are more optimistic. They naturally limit their information intake. They aren’t always optimistic though, e.g. if errors have considerable consequences.
When decision making, how do you calculate the expected result (this is the expected value theory)?
What’s the expected utility theory?
Find out all of the potential results, calculate the probability of each and times the probability by the value of the result. However, behaviour often violates this because the calculation doesn’t account for the utility of things psychologically (heuristics), for example the difference between 1 and 10 seems greater than 10000 and 10010. Also, it explains why people buy lottery tickets even though they’re guaranteed a loss according to the expected value.
The expected utility theory; this is similar to the above theory except it accounts for the decision maker’s goals affecting the value of the result. It calculates the highest psychological value which can be different to the objective value. This can be affected by certainty effects, preferences reversals, contexts effects, the framing effect and sunk costs.
What is the prospect theory?
This is the idea that when decision making, people are more sensitive to loss than gain (loss aversion) and people create reference points to represent their current state. So with loss aversion, people are bias and weigh the loss as greater (overestimation) than the gain (underestimation). This can be highlighted with the mug experiment; when people are told they can have a mug, they value the mug as greater because of the weight of the loss (if they didn’t take the mug).
What are the two invariances that should be in place when decision making?
Why does this not occur?
Description invariance; making the same choice no matter how the problem is described to them.
Procedure invariance; making the same choice no matter how the choice is measured, e.g. pairwise vs ranking.
This doesn’t always occur though, for example with preference reversals. Your preference for A over B can change if a similar scenario is presented differently, like Shafir’s lecture study. People chose to take B (interesting but early) a lot more but had no preference when told which one they would drop. Therefore, context effects should be considered. Also, there’s the attraction effect. You pick a favourite out of two; option B. However, if a third option is introduced that is dominated by option A. Then option A seems more attractive than both of the other choices. There’s also the framing effect, e.g. 90% success rate sounds better than 10% mortality rate. Sunk costs are when you pursue a failed project because of the amount of effort/cost you’ve put in.
What are the problems with the majority of decision making research?
It usually only looks at one action and one choice. However, this is unrealistic. Bounded rationality accounts for complex decision making as it understands that decisions are bounded by environmental and psychological constraints.
Discuss reducing the influences of being bias when decision making
You need to identify the influence and know the proper alternative strategy. However, being biased is often an unconscious mental process so this makes it difficult to overcome.
