Reasoning & Decision-making Lecture 6

Content and believability ....

'AFFECT" Content and believability effect Content effects: Whereas abstract propositional logic (i.e., if P then Q) is hard, when meaningful content is used (e.g., if a patron is drinking, she is over 18 years of age) it is very straightforward. Believability effects: People are likely to judge a conclusion that reinforces their initial assumptions, regardless of whether the conclusion follows from the premises, e.g., Some university professors are intellectuals Some university professors are liberals

People have a _____ Bias:

'Comfirmation Bias, in hypothesis testing Specifically, people: Favour confirming evidence to support their hypotheses rather than non- or disconfirming evidence. Often don't use available disconfirming evidence. Forget disconfirming evidence (our memory is biased to confirm hypotheses). Fail to consider better available hypotheses.

A cognitive illusion exists when there is:

A "correct" way of answering a question or making a decision (as indicated by the rules of formal logic or probability calculus) An intuitive or reflexive judgement or decision A discrepancy between the logical and the intuitive judgement or decision that always goes in the same direction (i.e., not random)

Are humans rational or irrational

A good deal of the research into human thinking and reasoning has led many researchers to conclude that people are biased in their thinking and frequently irrational, employing heuristics rather than the rules of formal logic . Gerd Gigerenzer (pictured) argues that evidence for the use of heuristics should not lead us to conceive of human thinking as riddled with irrational cognitive biases, but rather to conceive rationality as an adaptive tool that is not identical to the rules of formal logic or the probability calculus

Factors that influence reasoning also inlcude

Alteration of premise meaning People often make assumptions or alter the meaning of certain terms such that their interpretation of what the premises mean is at odds with what the problem actually states. Failure to consider all possibilities. People often to consider the range of meanings attached to premise information, e.g.; All A's are B's Some B's are C's

Cognitive biases

Basically, in logic tasks, people's intuitions seem to deviate from the logical requirements of the task in systematic ways (e.g., the confirmation biases found by Wason in his reasoning studies)

Syllogistic reasoning: true or false?

Categorical syllogisms with negative premises (such as "No A are B" or "Some A are not B") have no conclusion that necessarily follows. All polar bears are animals Some animals are white Some polar bears are white No union members are fearful No children are fearful Nothing follows Categorical syllogisms in which both premises are quantified by "some" have no valid conclusion

Factors that influence reasoning

Effects of premise phrasing. Compared with premises that contain positive phrasing, premises that contain negatives (e.g., no or not) are: More difficult to work with. Result in more errors. Take longer to comprehend. e.g., "You can't not believe what you read" Quantifiers (i.e., in syllogistic reasoning) differ in the ease with which we make sense of them: "All" or "none" are easy to understand. "Some" is more difficult. The order in which a propositions are presented also affects comprehension.

Reasoning about risk

Eight out of every 1,000 women have breast cancer. Of these 8 women with breast cancer, 7 will have a positive mammogram. Of the remaining 992 women who don't have breast cancer, some 70 will still have a positive mammogram. Imagine a sample of women who have positive mammograms in screening. How many of these women actually have breast cancer?

Propositional reasoning....

Evaluating the truth value of arguments can be difficult Doing so requires that we assess: 1) whether the premises are true or false 2) whether the argument is valid or invalid An argument that is valid and has true premises is said to be a sound argument. If an argument is sound, we can be certain that its conclusion is true. Logicians have specified various rules of proof to evaluate deductive arguments

Types of reasoning....Deductive reasoning

Goes from the general to the specific or particular. Conclusion is true if the premises are true. Any conclusions drawn are implicit in the premises. No new knowledge is added by deduction . Types of deductive reasoning include: 1)Propositional reasoning 2) Syllogistic reasoning

Types of reasoning... Inductive reasoning

Goes from the specific or particular to the general. Conclusions are possible, but not guaranteed. Inferential processes expand knowledge in the face of uncertainty. Therefore, inductive reasoning adds knowledge.

What is reasoning?....

Historically, construed as the type of focused, goal-directed thinking that is performed when solving puzzles. Often involves drawing logical inferences or conclusions drawn from other information (i.e., from premises or propositions).

Judgement under uncertainty: heuristics & biases

How do we cope with the myriad uncertainties encountered in our lives? We tend not to rely too much on formal, deliberative judgement and decision making processes. Instead, we use a variety of "fast and frugal" heuristics - rules of thumb - which involve rapid processing of relatively little information. Depending on one's perspective, these heuristics imply either rationality or irrationality.

Hypothesis testing

Hypothesis testing involves going from the particular (e.g., a sample) to the general (e.g., a population) Hypotheses cannot be confirmed as true, but they can be shown not to be true (i.e., disconfirmed). Every hypothesis has indefinitely many theoretical rivals that provide alternative explanations of the same data.

The Anchoring & Adjustment heuristic

In such instances, we might start with an idea or standard in mind (i.e., our anchor) and adjust our estimate from that starting point e.g., Suppose two people are asked to offer a numerical estimate of the population of Melbourne Imagine further that these two people are given a "starting value," obtained by spinning a roulette wheel, giving starting values of 3 million and 5 million, respectively The person assigned the starting value of 3 million might, following some adjustments from this anchor, offer an estimate of 3.5 million Similarly, the person assigned the starting value of 5 million might anchor their initial estimate at that value and, following some adjustments, offer an estimate of 4.5 million

Propositional reasoning....

Involves drawing conclusions from premises that are in the form of propositions. "The population of Melbourne is four million" . "Bachelors are unmarried". "Mary is psychology major". For the sake of convenience, propositions are abbreviated to single letters (e.g., P, Q). Compound propositions are created using logical connectives (i.e., &, ∨, ﹁, and ) e.g., P Q, P & Q, P ∨ Q, etc. In compound propositions, such as P Q, P is called the antecedent and Q is called the consequent. Either True of False.

Inductive reasoning: Reasoning under uncertainty

Involves reasoning about conclusions that are likely to be true, but not certain or guaranteed If we start with limited details, we can only hypothesise rather than deduce (cf. deductive reasoning) Inferential processes that expand knowledge in the face of uncertainty Examples of inductive reasoning include: Analogical reasoning Hypothesis testing

What are judgements & decisions?

Judgments refers to the processes by which individuals make use of various cues (which may be ambiguous) to draw inferences about situations. Decision making refers to the processes by which an individual selects one course of action from among alternatives. In practice, there is considerable overlap between judgement and decision making. This is reflected in judgement and decision making (JDM) research.

Foci of early jdm research

One group of psychologists took notice of the efforts of economists and statisticians to account for and advise people about their decision making. For these psychologists, the central questions were: How do people decide on a course of action? How do people choose what to do next, especially in the face of uncertain consequences and conflicting goals? Do people make their decisions rationally? If not, by what psychological processes do people make decisions, and can decision making be improved? Research into utility models of decision making reflects this approach to decision making See Galotti (2008) for details (these models are examinable)

Framing effects

People evaluate outcomes as changes from a reference point their current state. Depending on current state, outcomes are perceived as gains or losses. Driving down the road, you notice that your car is running low on petrol and you see two service stations, both advertising their petrol prices. Station A's price is $1.50 per litre and station B's price is $1.40. Station A's sign also announces "10 cents/litre discount for cash" By contrast, Station B's sign announces "10 cents/litre surcharge for credit cards. Which do you choose? All else being equal, most people report a preference for Station A; the one that offers the discount

Improving syllogistic reasoning when...

Performance improves when Venn diagrams are used. All A are B All B are C All A are C

Improving syllogistic reasoning...

Performance improves when specific examples are generated. All A are B All B are C All A are C All marsupials are brown All wombats are marsupials All wombats are brown

Wason's (1968) selection task Participant were told: If a card has a vowel on one side, then it has an even number on the other side....

Results: 46% chose A and 4 33% chose A Only 4% correctly chose A & 7 Reasoning on this task necessitates that people seek both confirming and disconfirming evidence. Deductive proof requires both kinds of evidence By and large, people are biased toward confirmation (i.e., A & 4)

Reasoning about risk

Risks are often communicated as probabilities: Degree of belief (subjective probability) Propensities (determined by physical properties of some object). Frequencies (based on long-run observations of event frequency), Three forms of risk communication invite miscommunication: 1) Single-event probabilities 2) Relative risks (versus absolute risks and number 3) needed to treat) 3)Conditional probabilities Conditional probabilities Express the probability of some event conditional on the occurrence of some other event. e.g., what is the probability that a person has cancer, given a positive test result for cancer?

The Availability heuristic

The availability heuristic proposes that we evaluate the probability or likelihood of an event by judging the ease with which relevant instances come to mind. When availability is highly correlated with actual frequency, estimates should be accurate. When uncorrelated, then our estimates may be way off e.g., we may assess the divorce rate in a community by recalling divorces among our acquaintances - who are all divorced! This raises the issue of "kind" versus "wicked" learning environments Factors that affect availability include: Biased encoding (e.g., vividness) Biased retrieval (e.g., primacy effects, recency effects

Two major cognitive systems

The deliberative system Conscious, deliberative, effortful, slow Processes 'information' in a successive or sequential manner Has been variously described as: The logicoscientific mode (Bruner); the rational system (Epstein); analytic cognition (Hammond); the deliberative system (Hogarth) The tacit system Unconscious, automatic, effortless, fast Processes information in a simultaneous manner Has been variously described as: the narrative mode (Bruner); the experiential system (Epstein); intuitive cognition (Hammond); the tacit system (Hogarth)

The Representativeness heuristic...

The representativeness heuristic is operant when we assess the degree to which the object of our attention is similar to, or represents, our basic idea (or stereotype) of that object . Consider, for example, of two sequences of coin tosses: heads, heads, heads, tails, tails, tails tails, tails, heads, tails, heads, heads Most people think series 2 is more representative of the (random) process that generated it

Hypothesis testing: Wason's 2-4-6 task...

The task: You are given the numbers 2, 4 and 6 and are told that this triplet of numbers follows a rule. Determine what the rule is, within these guidelines: You may not ask direct questions about the rule Offer your own examples of triplets (e.g., 4-6-8; 1-3-5; 9-7-5; 2-3-4; etc.). For each one you give, you'll be told whether it follows the rule. You must not guess. Announce a rule only when you are confident you know what it is. That is, formulate and test hypotheses about what the rule is. Results: Only 20% of participants solved the problem without error. People generally attempted to confirm rather than disconfirm their hypotheses.

Syllogistic reasoning...

This type of problem presents two or more premises and requires that we determine whether the conclusion is true or false. Categorical syllogisms present premises that deal with classes of entities. Examples of premises are: "All golden retrievers are dogs" "No polar bears are inanimate objects" Unlike the propositions considered earlier, the propositions considered in syllogistic reasoning cannot be regarded as composites of smaller propositions.

Analogical reasoning...

Using a relationship between two variables as the basis of a relationship between two other variables A is to B as C is to _______________ DOG is to COCKER SPANIEL as CAT is to ___________ a) Sennenhund b) Persian SPACE is to POINT as TIME is to ___________ a) location b) clock c) standard d) eternal e) moment Performance depends: Complication of individual terms Knowledge of the reasoner Ease of finding a relationship between A and B Number of possible answers Load on memory

Propositional reasoning: psychological studies.... Lay person's terms

Wason's (1968) selection task Participant were told: If a card has a vowel on one side, then it has an even number on the other side.

The research of such psychologists as Daniel Kahneman and Amos Tversky highlighted

the reflexive mental operations people use to make complex problems manageable. This research illuminated how these same processes can lead to both accurate and flawed judgements. These reflexive or intuitive processes are described variously as heuristics, cognitive biases or cognitive illusions. People tend to appear quite irrational in this program.