Please enable JavaScript.
Coggle requires JavaScript to display documents.
Reasoning - Y2 - Coggle Diagram
Reasoning - Y2
How should we reason?
Normative (deductive) reasoning:
- Conditional reasoning (propositional) - 'if-then' evaluations (If A then B)
-> logical reasoning - conclusions are necessarily true, a logically valid inference; 96.8% of people get these right (Schroyens, Schaeken & d'Ydewalle, 2001)
- In other situations, the conclusion is probable but not necessarily true
-> the statement implies that all dogs are animals, but not all animals are dogs
-> So logically, this is an invalid inference - harder problem which only 64% of people get right
- Diagnostic reasoning - using the symptoms provided to arrive at a conclusion means diagnosis is probable but not necessarily true (Flores et al, 2014)
Logic - normative system of calculating which conclusions follow from a set of facts and connections between them
- In psychology - typically focuses on necessary inferences
-> Inferences must be true, no situation exists in which they might not be true e.g. given the premises, the number eight is always even
- Fasts and conclusions typically have two truth values - true and not true
- However, many interesting problems aren't either true or false, they are uncertain
-> The answer lies somewhere between these two extremes
-> the conclusion is probably true but we are not 100% certain
Probability -
- Likelihood of something being true can be expressed as a probability
- My initial belief that any random person I meet has OCD might simple be the base rate for the population
- 12 in 1000 people have OCD
- If I learn more information about that person, such as specific symptoms, I increase my belief
- Using new information to update existing beliefs is probabilistic reasoning
-> Bayes theorem describes the normative method in completely novel settings
Cognitive process of deriving new information from old information
- Deductive reasoning - drawing logically necessary conclusions from given information
-> Premises - assumptions about the problem; conclusions are only true if these are true and valid
-> Propositional reasoning - simple statements linked by logical statements ‘if’, ‘and’ or ‘not’
Inference rules that can be used to derive correct conclusions from patterns of propositions such that different patterns trigger different rules
- Modus ponens - mode of affirming - states that given ‘if p and q’ if p is true, it follows that q is true
-> Mental logic approaches - however, these schemas may have fallacious inferences and may not include others, and do not match the formal inference rules discussed
- Modus tollens - mode of denying - states that given ‘if p then q’ and given ‘not q’ therefore, not p follows
- Double negation - therefore
Conditional rule has attracted much research - two valid inference patterns involving such rules, and thus two common mistakes - affirming consequence or denying antecedent
- These fallacies could come from misinterpretations of the premises - suppression effects
Mental model theory of reasoning -
- We rarely reason using logical rules or probability equations
- We generate possibilities (models) of different situations described by the facts
-> Integrate the facts into one or more models
-> If a conclusion is true in every model, then it must be a valid conclusion (Johnson-Laird, 1983)
-> If multiple models are valid, one is favoured on grounds of coherence
--> Completeness
--> Consistency
--> Plausibility
Errors in causal mental models -
- If the mental model is flawed, so will the inference be
- Gaps in the model -
-> Children's mental models of the Earth (Vosniadou & Brewer, 1992)
-> Adult experience allows us to deduce the shape of the earth, but children give varied answers based on the assumption of their sensory and semantic models both being true but contradictory
- Normal errors (Perrow, 1984) -
-> Wrong decision is unavoidable - we are built to make error
-> Reasoning is normatively appropriate - decision is reasonable given that mental model
-> the information input is generally faulty - sensation or perception error
-> Cuyahoga Coast Guard Disaster (1987) - Burns (2000)
- Naive physics - novices attempt to apply well-defined everyday knowledge to more complex (expert scenarios), but fail due to not having expertise - experts do not reason, they act (automaticity)
Relational reasoning - logic -
- Can see the relation between two things by looking at the model created - we also update our models if we answer incorrectly
- Hippos are bigger than toads
-> Hippos are bigger than sheep
-> Sheep are bigger than toads
- Sometimes our logic is accurate but the outcome is wrong
Syllogistic inference - Johnson-Laird & Bara, 1984:
- Logical competence depends on the ability to interpret premises and mental models, which cope with deductive reasoning
- Logic provides a systematic formal procedure that guarantees validity
-> Not mental logic for quantifiers - ordinary reasoning appreciates the overrides needing to search for counterexamples for conclusions, but there is no systematic method and thus error
- Syllogistic - deductive reasoning
- Categorical syllogism comes from deductive reasoning in which assumptions about a problem are made because of assumptions of category membership
-> These kind of statements are used to test reasoning often
On the conflict between logic and belief in syllogistic reasoning - Evans, Barston & Pollard, 2013:
- Belief biases observed despite controls - substantial effect for logic
- Interaction between belief and logic - belief bias was more marked on invalid than on valid syllogisms
- Even those focused on the conclusion are influenced to an extent by logic - belief bias is always present
Syllogistic reasoning - statements about groups related to terms such as ‘all’ and ‘some’
- Atmosphere effect - a tendency to draw conclusions in syllogisms that are over-influenced by the form of the premises rather than the logic of the argument
-> Compared to conversion errors (illicit conversion) and probabilistic inference - atmosphere effect is weaker with familiar as opposed to abstract or unfamiliar material, and it does not explain this difference
- Figural bias - the effect of figure on preferred conclusions
-
Intuition and analysis
Dual process theory - we are not simply logical; many psychological theories propose two types of thinking
- System 1 and system 2
- DPT used to explain reasoning, judgement and decision making
- Many cognitive biases and other behavioural phenomena can be explained as S1 or S2 processes, or the interaction of the two (Evans, 2008; Frank, 2008)
System 1 -
- Implicit, automatic, low effort, rapid, evolutionarily old, associative, contextualised, independent of intelligence, independent of working memory and hot emotion
- Fast, intuitive responses learned through experience
- Concepts that cover system 1 - heuristics, emotions (somatic markers) and expert intuition)
System 2 -
- Explicit, controlled, high effort, slow, evolutionarily new, rule based, abstract, limited by intelligence, limited by working memory, cold cognition
- These are slow, analytic responses that are calculated from information presented
- Concepts that cover system 2 - expected utility model, using mental models to find logically valid solutions
Dual process theory proposes we have types of thinking, such as long standing distinctions between intuitive and deliberate thought
- S1 is low in cognitive demand, as it is fast and intuitive
- S2 is slow and analytic, thus has high cognitive demand
- S2 can also override S1 thinking if needed, as S1 usually acts first
When do you engage system 2 and inhibit system 1 - Fluency heuristic (Whittlesea & Leboe, 2003):
- The option that is processed faster or more easily is the best
- Fluent options seem truer, more likeable, more frequent etc
- Lack of fluency used as a metacognitive cu to switch on analytic thought (Thompson et al, 2013) - we are more likely to use S2 when we are forced too, and are less likely to believe ‘bullshit’
-
-
-
Moral reasoning
Trolley problem - variations (non-fatal, asymmetric stakes, driver, person responsible with remote control and child)
- Doctrine of doubles - intended outcomes v known side effects (more likely to choose a known side effect)
Dual process model of moral reasoning:
- Deontological moral judgements are based on rights and duties - Kant’s moral philosophy
- Utilitarian or consequential moral judgements are based on weighing up the overall greater good - Bentham’s moral philosophy
Development of moral reasoning - Kohlberg’s Heinz dilemma - It is not what we do that matters, but why we do that is most important
1958 - argues we progress through a 6 stage model of moral development from childhood to adulthood
Characterised into 3 age related levels:
- Preconventional level - obey the authority
-> Stage 1 - avoid punishment - moral reasoning is based on direct punishment
-> Stage 2 - self-interest - based on reward
- Conventional level - social rules
-> Stage 3 - good boy attitude - social expectations
-> Stage 4 - law and order - based on the legal laws
- Postconventional - universal ethics
-> Stage 5 - social contract - moral reasoning is based on social agreements that can be changed when necessary
-> Stage 6 - universal principles - moral reasoning based on universal ethical principles and justice
-
-
Summary:
- Deontological judgements are automatic and emotional
- Utilitarian judgements are controlled and rational
- A dilemma occurs when these two evaluations conflict
- Personal moral violations (e.g. direct bodily force) trigger emotional responses
-> More deontological responses when pushing a person off a bridge
- Cognitive load interferes with utilitarian responses more than deontological (Greene et al, 2008)
-> Because it selectively hinders rational evaluation not emotional response
- Can an immoral decision be rational?